Datasets:

mimiklee

/

masterthesis_clean_text_20k

like 0

INTRODUCTION Crime is one of the major problems in many metropolitan areas across countries. Although city crime rates have dropped globally since the mid-1990s, 1 there are still large variations and dramatic fluctuations across cities. 2 Past criminological studies revealed that variations in crime rates were explained by characteristics of metropolitan areas, such as population size, ethnic heterogeneity, geographic mobility, economic segregation, unemployment rate, poverty level and degree of social integration and control. [2][3][4] Thus, many governments have made great efforts to reduce the crime rates especially in metropolitan areas by intervening in those characteristics to ensure the safety of their residents. Safety from crime is not only an essential human need in daily life, but also a prerequisite to human health. 5 A body of past studies has reported the relationship between residents' perceptions of neighbourhood safety and their health outcomes. 6 7 For example, one UK survey with 407 adults reported that fear of crime was significantly associated with self-rated health and mental well-being. 8 Ziersch et al 9 showed that perceived neighbourhood safety was related to physical and mental health among 2400 residents in western suburbs of Adelaide, Australia. However, these earlier studies suffer from the following limitations. First, most of the previous studies used individual-level neighbourhood perceived safety as an exposure variable, which could be influenced by several factors such as the prior individual experience of victimisation or individual health conditions other than neighbourhood-level safety. 10 11 Strengths and limitations of this study ▪ A multilevel analytic frame was used to examine the association between district-level perceived safety and residents' self-rated health using a representative sample of metropolis, Seoul, South Korea. ▪ We succeed in adjusting for potential confounders such as the district-level crime rate in our analytic model, which past studies have failed. ▪ Causal relationships cannot be inferred from the cross-sectional data of this study. This could be particularly a critical issue in previous crosssectional studies because of the potential reverse causation, meaning that the sick people are more likely to perceive their neighbourhood as unsafe. 7 12 The second limitation is a lack of representativeness of samples within the operationalised definition of neighbourhood. Few studies had enough sample size or the sample size within neighbourhood to be representative of each neighbourhood. 6 8 13 Unless the responses are obtained from a representative sample of participants within each neighbourhood, aggregated perceived neighbourhood measures can potentially be prone to measurement errors. The final limitation is that previous studies did not adjust for the district-level crime rate as a potential confounder although the crime rate has been reported to influence perception of neighbourhood safety as well as residents' health outcomes. In this study, we assessed the district-level perceived safety, which was obtained by aggregating responses from the residents that are representative samples for each administrative district in Seoul, the capital of South Korea. Then, using multilevel analysis to account for the hierarchical structure of the data, we examined the association between district-level perceived safety and selfrated health after adjusting for potential confounders including the district-level crime rate. --- METHODS --- Study population Data were obtained from the Seoul Welfare Panel Study , which tracked a representative sample of households residing in 25 administrative districts in Seoul, South Korea. The SWPS was launched in 2008 by the Seoul Welfare Foundation. The first wave of the survey was conducted in 2008 and its supplementary survey targeting the low-income households was implemented in 2009. The SWPS was suspended after the second wave of the survey was conducted in September 2010. The survey employed a two-stage stratified cluster sampling approach where a representative sample of census tracts for each district was first drawn, and then households were randomly selected within those sampled census tracts at the baseline. A household representative answered the household survey and all members of a household whose age is 15 or older were interviewed. A total of 7761 individuals completed the interviews in wave 1. The SWPS has been publicly released . Because all respondents answered on questionnaire items we used in this study, we were able to conduct our analyses based on the entire sample participated in the first wave of SWPS without listwise deletion or missing value imputation for handling missing data. The final sample used in the data analysis of this research consists of 7761 individuals from 3665 households from 25 administrative districts in Seoul. The number of households in each district was 146.6 on average, ranging from 108 to 198. This research received Institutional Review Board exemption from the Division of Research Affairs at the San Diego State University. Exposure: district-level perceived safety District-level perceived safety was assessed through the household survey using a question about how much a household representative agrees with the following statement: "my current residential environment is unsafe." Respondents answered in a five level ordinal scale from 'strongly agree' to 'strongly disagree' . The answer was then dichotomised into 'unsafe' for the response, 1 and 2 and 'safe' for the response, 3-5. The binary responses from household representatives were aggregated to calculate administrative district-level perceived neighbourhood safety by taking a weighted average of household-specific perceived safety within each district with the household size used as weight. Such aggregation results in that the district-level perceived neighbourhood safety is essentially the sample proportion of individuals who answered 'safe' within each district. Outcome: poor self-rated health Poor self-rated health was assessed through the individual interview using the question 'how would you rate your overall health?' This question is on the ordinal level, ranging from 'very good' to 'very poor' . The response was then dichotomised into 'good health' for the response, 1-3 and 'poor health' for the response, 4 and 5. Although self-rated health cannot assess multidimensional aspects of health conditions, it is known to be a reliable predictor of life expectancy after adjusting for other health indicators. 14 --- Covariates We included several confounders in the data analysis. For individual-level confounders, we have sex, age group , education level , marital status and job status , household income with six categories and individual perception of district safety . Because neighbourhood safety was assessed solely from the household survey, we assigned the value of perceived neighbourhood safety measured from each household representative to all members of the household. We considered the district-level crime rate as a potential covariate at the district level because it can influence residents' health as well as perceived safety. District-level crime rates for each of the 25 administrative districts in Seoul were collected from the 'analytical report on crimes, 2008' that was published by supreme prosecutors' office in South Korea. 15 The district-level crime rate was calculated by dividing the total number of crimes by the total number of residents in each district . Using an identifier for each administrative district in the SWPS, we linked the official crime rate of each administrative district to our final dataset of the SWPS. --- Data analysis Mixed effect logistic regression was used to investigate the association between district-level perceived safety and self-rated health. Because of the hierarchical structure in our data , within-household and within-district correlations were incorporated using household-specific and districtspecific random intercepts. We made stepwise adjustments of potential confounders in the data analysis. First, we adjusted for potential confounders including sex, age, education level, job status, marital status and household income. Second, we additionally adjusted for the district-level crime rate. Finally, we examined the association after adjusting for individual perception of district safety in addition to all of the previously mentioned confounders. All of the confounders were included as categorical variables, and the district-level perceived safety was included after standardisation for simple interpretation in the model. All computations were performed using the R statistical software. --- RESULTS Table 1 presents the distribution of the study population and the prevalence of poor self-rated health by each of the individual-level, household-level and district-level characteristics. Overall, poor self-rated health was reported at 20.9% . The proportion was higher for women and showed an increasing pattern with age. Higher prevalence of poor self-rated health were observed for participants with lower education levels. The unemployed and people living in an unsafe neighbourhood exhibited higher prevalence of poor self-rated health than the employed and those living in a safe neighbourhood. The household income was fairly equally distributed in the SWPS. As to the district-level perceived safety and crime rate, given the overall mean of each variable, the relative size of each SD shows that there were considerable variations among the 25 districts. District-level perceived safety was significantly associated with poor self-rated health while different sets of confounders being step-wisely adjusted . Living in a district where its safety level is 1 SD higher resulted in 13% lower odds of reporting self-rated poor health status after adjusting for sex, age, education level, job status, marital status and household income. When we additionally adjusted for the district-level crime rate, this association was slightly attenuated but still significant . When we adjusted for individual perception of district safety in addition to previously mentioned potential confounders, the magnitude of this association remained significant . --- DISCUSSION Evidence from our study indicated that district-level perceived safety, which was assessed by aggregating responses from residents in each district, was significantly associated with poor self-rated health even after controlling for demographic information, socioeconomic status and district-level crime rate. Notably, this association was robust also when we additionally adjusted for individual perception of district safety. Our findings are in line with previous research that showed association between perceived neighbourhood safety and health outcomes. Past studies have also reported that residents who perceived that their neighbourhood had more severe problems were more likely to experience greater anxiety, stress and depression. 6 16 The studies, which sampled women, children and the elderly, also provided consistent evidence of a relationship between perceived crime risk and physical health. 7 13 In this study, district-level crime rate was not associated with self-rated health in the fully adjusted model. Furthermore, our post-hoc analysis showed that there was no statistically significant relationship between district crime rate and residents' self-rated health regardless of adjustment of confounders, although the district-level crime rate could have a major influence on district-level perceived safety. This finding is different from past studies that reported a significant relationship between district crime rate and residents' health such as coronary heart disease 17 and low birth weight. 18 The differential association between district-level perceived safety and crime rate in relation to self-rated health could be explained in three ways. First, mass media may increase individual-level perceived neighbourhood insecurity regardless of their neighbourhood crime rates, especially when they reported the crime in ways of exaggeration. 19 20 The mass media tend to emphasise criminal stories which can draw attention from the audience. 21 Previous studies called this phenomenon the 'cultivation effect' meaning that exposure to the world of television cultivates exaggerated perceptions of viewers and magnifies viewers' fear about crime. 22 The residents who watched news about neighbourhood crimes are more likely to perceive their neighbourhood as more vulnerable regions to crime regardless of the regional crime rate. 20 Second, different types of crimes would have different effects on the perceived risk or fear of crime. For example, murder, rape and personal theft may have stronger effects on the fear of crime than larceny and auto theft. Hence, the total crime rate that was used in this research might not be sophisticated enough to examine the association between the prevalence of crime in the district and the residents' health condition. 23 However, when we conducted a post-hoc analysis using a different measure, '5 index crime rate', which includes five major serious crimes that has been adopted by the Korean police to indicate violent crime rate, still we could not find association with residents' self-rated health. Finally, if social and physical resources of neighbourhood are deteriorated or deprived, residents tend to perceive neighbourhood safety more irrespective of the objective neighbourhood crime rate. 24 The poor quality of the social and physical environment, such as dilapidated houses or having a lack of formal or informal neighbourhood networks, may work as a trigger to make residents perceive their neighbourhood as dangerous. 25 26 There could be several pathways linking district-level perceived neighbourhood safety to residents' self-rated health. First, poor district-level perceived safety can cause less outdoor physical activities 7 27 leading to poor health. Second, poor district-level perceived safety may aggravate social supports which may harm mental and physical outcomes. 28 Lastly, poor district-level perceived safety may be a latent stressor causing chronic stress status that could undermine residents' mental health. 29 Our study has several limitations. First, potential reverse causation is of concern due to its cross-sectional study design, implying that people with poor self-rated health are more likely to perceive their neighbourhood as unsafe. Future studies are required to examine the causal association between district-level perceived safety and health outcomes. Second, this study assessed perceived neighbourhood safety through a single-item measure. This item may not reflect multidimensional aspects of the neighbourhood safety. Despite these limitations, our study has the strength in that we used representative samples for each operationalised administrative district, which enabled multilevel analysis using a district-level aggregate measure of perceived safety whereas most of the previous studies used individual reporting of perceived safety as an exposure variable. Furthermore, to the best of our knowledge, this is one of the first studies to examine the association between district-level perceived neighbourhood safety and health outcome after adjusting for the district-level crime rate. In summary, our study showed that district-level perceived safety was associated with residents' poor selfrated health even after controlling for demographic information, SES and district-level crime rate. Our study results evoke the importance of local authorities to make efforts toward improving neighbourhood perceived safety to enhance residents' health.

Objectives: Several studies have reported the relationship between residents' perceived neighbourhood safety and their health outcomes. However, those studies suffered from unreliability of neighbourhood safety measure and potential residual confounding related to crime rates. In this study, using multilevel analysis to account for the hierarchical structure of the data, we examined associations between district-level perceived safety and self-rated health after adjusting for potential confounders including the district-level crime rate.

Introduction England has an assessment regime in which all state school children are tested at ages 5 , 7 , 11 and 16 ; Key Stage 3 assessments at age 14 were discontinued in 2010. Results nationally can be compared year on year for evidence of system improvement and to identify so-called poorly performing schools. Full national data sets available over many years foreground and quantify a number of national educational policy challenges. One challenge is low attainment which is unequally located in specific groups; 'closing the gap' between affluent and poor pupils is a concern across Europe . A second challenge, revealed in the data, is recurrent low average attainment of specific ethnic minority groups ). Gender is the third area of policy challenge with girls out-performing boys by a margin of 10% at every assessment point, though this advantage does not persist beyond school into wage levels. This paper returns to a tradition of political arithmetic in which Stevens points to a progression over time in the focus of academic research on educational inequalities, which has moved successively from 'the deficient child through the deficient family to the deficient school'; this could be extended to the deficient society, an extension which demands attention be given to policy decisions and political resistance to addressing inequalities. Nearly two decades ago, Giddens, architect and intellectual champion of The Third Way, insisted that this third way 'must reduce inequality', and, if it does not, 'is a betrayal of the social democratic ideals of collective provision for the poor and needy' . In 2015, policy in the UK is still about 'removing barriers', adjusting in-school factors rather than interventions to lift people out of poverty. The evidence is strong that the causes of low attainment lie largely outside school and could be better tackled if the poverty argument were accepted and addressed . The Spirit Level , UNICEF report cards and OECD reports show just how poorly placed the United Kingdom is to embrace redistributive policies. --- Large datasets and the analysis of inequalities in England Strand has analysed national datasets and the Longitudinal Study of Young People to unpick and weight factors associated with low attainment. He sets out the challenges for researchers and politicians both in terms of grasping the statistics, theorising inequalities and devising viable policy proposals. Strand's analysis of the educational progress of an entire national cohort between age 7 and 11 showed that 'Black Caribbean boys not entitled to free school meals, and particularly the more able pupils, made significantly less progress than their White British peers' . Strand's 'The limits of social class in explaining ethnic gaps' draws on LSYPE data covering an interview survey of over 15,000 young people who were aged 14 in 2004. Raw scores produce an ordering where Pakistani, Bangladeshi, Black Caribbean and Black African groups perform below the levels of White British. Introducing family background, parental attitudes and student risk and protective factors reduces the apparent inequality, but Black Caribbean students, stood out as attaining at lower levels than might be expected . It also contains the following summary across gender, ethnicity and social class: 'The gender gap was just 0.8 points with boys scoring lower than girls ... the ethnic gap was … 3.3 points. … the social class gap … 9.6 points' 1 . Strand's 2012 paper looks at a specific claim of institutional racism whereby Black Caribbean students are disproportionately allocated to the lower tiers of a course at age 14 and cannot then achieve the top GCSE grades. He concludes that 'this under-representation in the higher tiers [which] is specific to one ethnic group and persists even after taking account of extensive additional explanatory variables, suggests a significant cause for concern' . Elsewhere, Strand reports that Black Caribbean and those designated Black Other perform worst at age 11 with a White British-Black Caribbean gap of 0.45 of a standard deviation, and across socio-economic status groups, it is 0.57 . 'Low SES White pupils were the lowest achieving group' . He finds that, at 16 , 'The only group of students to make significantly less progress than White British students was Black Caribbean boys with high prior attainment and those of average prior attainment at medium and high SES' . Strand's studies, and others using data from the Youth Cohort Studies and the National Pupil Database are important but dated. There is an urgent need to clarify, reconcile and update competing claims about ethnicity, deprivation and gender correlates with low attainment, which this paper sets out to do. --- Methodology applied to school national attainment data for England 2009 - --- 2014 The rest of this paper is the author's update of the ethnicity-gender-deprivation debate in relation to school attainment in England. It pays particular attention to the way the three major demographic variables -gender, ethnicity and deprivation -combine with different weight to affect low attainment, examining whether they combine in different ways at different assessment points or ages in pupils' school careers. Deprivation may exert a stronger influence on some ethnic groups than others, or affect one gender more than another within these ethnic groups, and this may be more marked at one key stage than another. 1. An overview of the population under discussion showing the sizes of the ethnic minority sections of this population compared with those classified as White or White British and the different Free School Meal rates. 2. An examination of the performance by ethnicity for the Early Years Foundation Stage Profile , Key Stage 2 and KS 4 over four years. . 3. An analysis of mean attainment scores for the same three assessment points for 2014 by ethnicity separated into FSMnon-FSM and within that divided by gender. 4. An extension of the poverty argument is extended through IDACI which allows attainment levels to be identified from the poorest to the most affluent. 5. An international context on welfare regimes and child poverty in different countries and the impact of welfare transfers on poverty/inequality levels. 6. A consideration of ideological/theoretical explanations of racism and inequality in English education. --- Results --- Overview of the school population in England Table 1 shows the relative sizes of ethnic groups in the contemporary school population of England, as a basis for an understanding of the scale and locations of inequality. Black, Asian and Mixed groups are disaggregated into their main sub-groups as it is misleading to deal with them as homogenous entities. Groups could be broken down further and, within the Black African category, children from some countries of origin countries achieve greater success, most notably Nigerians and Ghanaians compared with Somalis. Demie points to language diversity as important, with Black African speakers of Yoruba and Igbo achieving significantly higher than those speaking Somali and Lingala, to pick examples amongst the more numerous groups . This level of detail is not available outside the National Pupil Database and is not part of this study. Nearly 80% of pupils in schools in England are White and over 90% in that group are White British. The largest single minority ethnic group is Pakistani, followed by Black Africans; the latter have more than doubled in number over the last ten years. No individual minority group constitutes more than 4% of the total school population but they are not evenly spread. Some local authorities have over 50% ethnic minority pupils in their schools. pointed out how limited the use of free school meals is in analyses which require a full grasp of the relationship between gradations of poverty and attainment. --- Ethnicity Tables 6 and7 apply IDACI deciles to attainment. --- Attainment by ethnicity at three assessment points In the analyses of national data and charts presented below, the level of attainment of White pupils is not included in the graphs because their position is so close to the national mean it would barely show. Separating White British from other White groups, which are under 4% of the White total, also makes minimal difference 3 . --- Mean attainment ethnicity, deprivation and gender Tables 2 -4 represent a further breakdown of these groups at three assessment points to ascertain whether, within categories of ethnicity and deprivation, the performance of boys differs from girls. National percentage data for these groups by gender and FSM displays a constancy over time though only 2014 figures are shown here; data from 2013 in Appendix 1 show a very similar picture for all three assessment points. Text accompanying the tables below summarises the highest attainers, lowest attainers and the 'gap'. Table 2 demonstrates regularities in the differences, represented also for the most part in Tables 3 and4. White FSM pupils do consistently poorly and at every age level from age 5 onwards, with both boys and girls, falling increasingly further behind. Poverty affects education outcomes more for some ethnic groups than others, and at KS2 the poverty impact is greater for boys than girls though fairly equal at KS 4 as shown by the 'Gap' columns in Tables 3 and4. . Caribbean and Pakistani heritage pupils reach the national average. This is evident from Tables 3 and4, from Figure 5 and Appendix 3, Tables A6 andA7. --- FSM Non-FSM There is, therefore, a 'benefit' associated with some ethnic groups and 'deficit' for others; and they are the same groups over the last five to 10 years. Taking ethnicity, gender and deprivation serially, one can calculate the 'negative impact' on groups in relation to these attributes. These are shown for Key Stage 2 and 4 for 2014 in Appendix 3. At KS 4, ethnicity is related to an 'impact' range from a mean 'benefit' of 17 per cent points or a mean 'deficit' of 10 per cent points and it is not sensible to average this out. Being a boy brings with it a mean 'negative impact' of about five percentage points. Being a pupil categorised as eligible for free school meals brings a further 'deficit' of 20 percentage points. It is, as Connelly et al , in an unhelpful, triumphalist tone, state, 'Socioeconomic differences in educational attainment trump both race and gender' (p. --- 62). Translated into odds, Table 5 Table 6 shows the differences between the least and most deprived deciles in all four curriculum areas at KS 1. The differences are big, although the 'gaps' reduced over the five year period, by one third in reading and writing. Appendix 2 Tables A4 andA5 show mean attainment scores for each decile at KS 2 and 4. Table A5 shows a relentless climb by one to two percentage points for each step up the IDACI decile scale for almost every year. Efforts to reduce this gap have had some success reducing the percentage points gap by a third. Similarly, Table A5 shows, for 16 year-olds, a rise of two, three or four percentage points rising up the IDACI decile scale. The gap between top and bottom reduced by nearly a quarter over the years displayed. One can see greater improvements in the levels of attainment of the poorer deciles, as for Key Stage 2, but the improvements possibly attributable to schools' efforts are dwarfed by the enduring poverty-related difference. The case for there being a substantial and enduring link between attainment and straightforward 'income deprivation' is strong. --- An international context to poverty and attainment England's income redistribution is relatively ungenerous, leaving a big gap between rich and poor, as represented by the Gini coefficient 6 . Only Spain has a bigger Gini Coefficient gap after taxes and transfers have been taken into account. UNICEF reports on comparisons of child poverty in affluent countries shows the UK as having a higher proportion of children in poverty than most of the countries with which it would want to be compared. The --- Scandinavian countries and The Netherlands have enviably low levels and France and Germany do significantly better than the UK. Jerrim reports that, 'The association between family background and high achievement is found to be stronger in England than in most other developed countries' . Wilkinson and Pickett's diverse list of measures, where wellbeing is better where inequality is low, is persuasive . In more equal societies, child well-being measures are higher, mental illness rates lower, use of illegal drugs is lower, the teenage birth rate is lower and women's status is higher -to list but a few of the areas associated with greater equality. OECD sets out the reasons why inequality is bad, morally and economically, and how the state needs to do more to address it and has a section on 'Reducing inequality in educational outcomes ' . --- Discussion This section summarises the quantitative evidence on low attainment by specific groups, the roles of ethnicity, gender and poverty and the effect which might be attributed to the school in countering inequalities. It finally considers a society's responsibility for the production and maintenance of poverty and the deficient society. It is appropriate to begin with the assertion that the race effect can be positive or negative, and that large variation must be acknowledged in theoretical and policy discussions. Ethnicity accounts for about one quarter of the gender FSM effect. Gender affects the mean score at KS 2 and KS 4 by around one third of a standard deviation, FSM by over one standard deviation and the gender/FSM effect is one and half standard deviations. --- Poverty and low attainment White FSM pupils, both boys and girls, are the largest FSM group and the lowest attainers at age 5, 11 and 16 with unbroken regularity. A negative 'deprivation impact' of around 17 percentage points at KS 2 and 22 percentage points at KS 4 is calculated from the 2014 national data . The number of White FSM KS 4 pupils was 63, 370. This is the largest group and they achieve the lowest mean scores There have been calls for family poverty to be addressed and government reports press for this in relation to health, education and social mobility . Poor communities suffer from a range of social ills in inner cities, coastal areas, ex-mining and other deindustrialised areas and their engagement with education has been documented . A much overlooked publication, Parenting in Poor Environments , gives detail and causal links on what makes support for children and young people in deprived circumstances necessary. The Joseph Rowntree Foundation similarly explicates the mechanics of poverty and the struggles it poses for relationships, parenting and lone parenthood. Cooper and Stewart , in their earlier Joseph Rowntree publication, identified 'significant effects of household financial resources on wider outcomes for children, including cognitive, social-behavioural and health outcomes …. Money itself makes a difference to children's outcomes' . Pickett and Vanderbloemen write relatively optimistically about the effect of the Pupil Premium, but reduce confidence about in-school prevention and intervention initiatives by the use of the cliff edge metaphor -erecting a fence, having an ambulance waiting at the bottom or even stopping people being motivated to rush to the cliff's edge. 'The educational parallel to the ambulance and the cliff analogy is that educationally focused policies and interventions cannot deal with the structural issues of poverty and inequality which are the root causes of educational inequality' . --- The Social Mobility and Child Poverty Commission 'holds Government's feet to the fire', focusing on 'what the UK governments …. are doing to tackle poverty and improve mobility' . Amongst its suggestions are to 'unlock social progress' and to ensure that 'family incomes that are supported by decent levels of pay and the right incentives to find employment and work enough hours' . The Commission's record to date is not impressive. The policy implications from these calculations, if there is authentic commitment to child protection writ large and social justice for children, is for adjustment to the distribution of wealth and income, partly through social transfers to bring the UK nearer to some of its European neighbours. This area is one where change can be effected in ways it cannot be with gender and ethnicity. --- Gender Girls do better on average than boys at every stage in education, whatever ethnic or deprived sub-group they are in. A negative 'boy impact' of 0.34 at KS 2 and 0.39 of a standard deviation at KS 4 can be calculated. Girls comprise slightly under half of the each of the age cohorts from Foundation Stage to GCSE , but a slight majority at A Level , where they continue to achieve better average results in average points scores, though not in the proportion achieving 3 A*-A grades or in the Russell Group of Universities' 'facilitating subjects' . There is still the marked difference in subject choice at A Level. Eighteen year-old women were also one third more likely to go to university in 2014 . The debate about girls' experience of education has moved beyond concern over curriculum content and teaching styles to broader concerns of progress beyond formal education whether school or higher education. Skelton and Francis provide an excellent account of the phases of feminist agendas on girls' education. The focus has swung to boys ' under-achievement with explicit attention to so-called working class boys. Yet amongst girls, the same degree of low attainment is shown relative to other girls. White FSM girls are the lowest attaining girl group at 16 . Dillabrough et al. with a global and historical span write of girls' and women's identities and seek to 'reposition' the debate in this wider ecological and life-span context. The policy implications here are complex and multi-faceted. The double negative impact of being both a boy and poor is around 20 percentage points at KS 2 and 30 percentage points at KS 4. There is considerable variety in the impact of poverty as indicated by FSM eligibility on different ethnic groups . --- Ethnicity Analyses of national DfE data, and repeated 'snapshot' measures over time, in whole age-group populations, show trends and constancy in attainment levels for different groups and it is the disaggregating of the 'Black' and 'Mixed' ethnic categories which reveals specifically where sustained low average attainment is found. One cannot talk about Black or Asian pupils under-performing when we see that at every level, but increasingly towards the later stages in education, it is specifically Black Caribbean, other Black background and Mixed White and Black Caribbean pupils who perform poorly along with Pakistani pupils; arguably social and educational system inputs over decades have not worked sufficiently well ). 'Passive racism' draws attention to ethnic inequalities that are recognised, year on year, yet the focus and resources devoted to correcting these inequalities are insufficient. Action cannot be limited to the school alone, but one has to note the strong performance at Key Stage 4 of Black African, Mixed White and Black African and Bangladeshi students of both sexes whether of FSM or non-FSM status . Indian and Chinese pupils consistently achieve higher. These differences should prompt a sensitivity to the cultures of the different groups which Archer and Francis on Chinese pupils and Richardson and Wood on Pakistani pupils amply display. The label 'institutional racism' applies poorly to the situation where Black groups vary greatly in the benefits gained from education. Critical Race Theory, imported from the USA, where it may well apply , has little useful to say in the English context. Notions of 'conspiracy' or 'white supremacy' do not contribute to explaining the long-standing low attainment of Black Caribbean, Other Black background, Mixed White/Black Caribbean or Pakistani pupils nor the high attainment of Black African, Mixed White/African, Bangladeshis or Indians. The picture is complex and solutions need to be tailored to national and local conditions. The lower than expected attainment levels of more affluent Black Caribbean and Mixed White/Black Caribbean pupils, both boys and girls is a particular case to study. Strand draws attention to this and Rollock et al. have pursued this. Some ideological stances are psychologistic and lack a structural dimension. 'Intersectionality' is not pursued effectively in quantitative terms and Gillborn's 'conspiracy' and 'white supremacy' do not intellectually connect with the UK's increasingly neo-liberal politics, reduction in the size and limiting of the state's role, reducing taxation, negative redistribution and creating poverty and inequality. We lack a longitudinal approach which would follow children through from the Foundation Stage Profile to 16, analysing the attainment data by gender, ethnicity and deprivation to see at what points, and with which groups, attainment levels alter. A report which addresses this lack analyses longitudinal cohorts and indicates, that some ethnic minorities eventually exceed and others approach the national mean attainment level at 16, improving at each assessment point as they progress through the school. The complexity is increased where we note the significant change in how numbers in each ethnic group have changed over the most recent five years with notable increases of over 20% in Black African, Bangladeshi, Other Asian background, Mixed White and Asian and Mixed White and Black African pupils. White pupils declined by 30,000 in the same period. In addition, second generation of newly arrived minorities, many of whom would wish to be recognised as Black British, bring a different inheritance to the educational experience. --- The limited power of the school to increase equality One 'official' judgement is that schools can do the job: 'London's educational performance suggests that the problem of white working class underachievement in education can be tackled' . This related to the London Challenge scheme and the laudable cooperation between schools it promoted. However, a more convincing explanation of London's improved attainment levels is the changed ethnic make up of London's school populations 7 particularly the increases in higher achieving ethnic minorities in London over that period . There are various estimates from large datasets about the proportion of attainment that can be attributed to 'the school effect'. Goldstein, a renowned educational statistician, has said in a media interview that the school effect is about 10% . Drawing on the work of Peter Mortimore, Gorard reinforces the 10% point with great clarity writing, 'Of the 30 to 40% that can be explained, the vast majority of this is attributable to the prior and individual characteristics of the pupils' . This parallels Ball's estimate of between 5 and 18 per cent . Thrupp claims similarly that, 'educational quality in low-SES [Socio-Economic Status] settings will not be able to be substantially improved without redistributive policies of various kinds ' . Berliner , in his Our impoverished view of educational research, had similarly pointed out that 'small reductions in family poverty lead to increases in positive school behaviour and better academic performance …. Poverty places severe limits on what can be accomplished through school reform efforts' (p. --- 949). Those arguing whether it is racism, sexism or poverty/inequality that underlies the tragedy and malevolence of low attainment are misguided. They divert attention from the evidence in correlational analyses, which consistently show the primacy of income poverty -not having resources to reliably and stably run a family life. The school improvement/effectiveness movement has attracted many leading researchers and considerable funding. The major limiting role of poverty is largely acknowledged by these authors but in tokenistic form. Ainscow and colleagues, for example, argue that equitable developments in education will ultimately depend on government pro-equity policy frameworks, a statement then regrettably softened by the ultimately misleading sentence: 'In the meantime, it is also the case that much can be achieved by school change' . --- The deficient society Economic inequalities are sustained by national policies and income deprivation is a resolute correlate of low attainment. Epidemiological research shows powerful correlates between infection and mortality and poverty in the same way. Marmot laments the lack of political attention given to this link which is mirrored in the English education establishment. An economically advanced nation does not lack the resources to tackle poverty with more direct interventions. Piketty internationally and Dorling on the UK make plain the macro-economic trends governments passively allow or covertly promote. Amongst Pickett and Vanderbloemen's conclusions in their Mind the Gap is the statement, 'The most important influence on … how well a child develops in the early years, performs in school, in later education and in adulthood, is family background… Children do better if their parents have higher incomes [and] Inequalities in educational attainment and outcomes have a social gradient' . The deficient society will not be corrected by the public availability of the detailed data presented here 8 . England's political decision-making would need to take seriously the individual damage to children resulting from poverty and respond in some measure to the challenges set out in OECD reports showing that education and life chances are diminished and 'less inequality benefits us all' . The evidence that poverty is a major factor in low attainment and other social ills should make politicians heed Atkinson's entreaty that, 'It is imperative that the EU should prioritise measures to ensure the achievement of the Europe 2020 target for reducing poverty and social exclusion ' and indeed the UK's own eradication of child poverty target as expressed in the Social Mobility and Child Poverty Commission reports . Marmot writes, 'What makes these health inequalities unjust is that evidence from round the world shows we know what to do to make them smaller. This new evidence is compelling. It has the potential to change radically the way we think about health, and indeed society' . This same ecological perspective applies to education and teacher-led, a London Challenge collaborative style, with 'some' extra money, cannot solve low attainment problems which are so strongly correlated with poverty. Addressing educational inequalities requires also that educational professionals at all levels understand and believe the data on the poverty/inequality link and campaign for it to be seriously addressed for the benefit of those identified as most vulnerable to, and most obviously bearing the undeserved consequences of, their disadvantage. One might question whether the costs of collecting, organizing and publishing detailed annual population education attainment data and the burden on teachers and children to supply it. The data are used for accountability but could be used by front line professionals to point to the poverty, gender and ethnicity correlates of low attainment and other social justice goals. There are moral choices which require collective professional confidence and a wider public to turn the spotlight away from schools as the 'saviours', expected to succeed against the odds, to influence realistic political commitment and action targeted at child poverty reduction in the UK. The 'deficient society' is not a natural occurrence but one in which the contracting state serves the interests of those with wealth, paying insufficient attention to inequalities in relation to race, gender or poverty. 1 This summary is misleading because: gender splits the whole population into two; but ethnicity compares the top and bottom groups -9,406 White British with 558 Black Caribbean ; social class compares the 1,378 in the top of eight social classes to the 1,019 in the bottom class . It is not as simple as ethnicity being three times more powerful than gender in explaining the gap nor that social class is three times more powerful than ethnicity. However, there is a strong signal here about the range and relative strengths of different factors. 2 The FSM percentage is calculated from DfE reports on KS2 and KS4 assessments for 2013. No published DfE statistics provided FSM status by ethnicity. 3 The 'other' White groups are made up of Travellers of Irish Heritage, fairly constant at about 130 in any year group and Gypsy/Roma whose numbers shown in the GCSE results statistics have risen from 480 to 820 in the 2009 -2013 period. Both groups score very low percentages of 5A*-C . The Irish, constant at each year group at 1,900 score about eight percentage points higher at KS4. The largest group is the 'any other white background' which has increased from 2009 and at KS4 registers about five percentage points below the national average. Together constituting 3.9% of the national total, they make little difference to the overall White attainment percentages and are not presented separately; White British GCSE results are only 0.2% higher than the percentage level for All Whites. 4 OAS are included here but in no other graphs. They are as numerous as Bangladeshi and other groups included in most analyses. They are amongst the best FSM performers and third amongst non-FSM pupils. Other Asian background children are included in Appendix tables A7 and A8. 5 The Income Deprivation Affecting Children Index is calculated from: • Children in Income Support households • Children in Job Seekers Allowance households • Children in Working Families Tax Credit or Disabled Person's Tax Credit households whose income is below 60% of median before housing costs • National Asylum Support Service supported asylum seekers in receipt of subsistence only and accommodation support. IDACI deciles are simply taking rankings in order of deprivation divided into 10 equal bands from the most to the least affluent. Each band contains between 47,000 and 70,000 children in a year group nationally. 6 A Gini coefficient of zero indicates perfect equality, where all incomes are the same; a coefficient of one expresses maximal inequality where one person has all the income. 7 In 2006, both the national and London percentage 5A*-C grades including English and mathematics at 16 was 45%. In 2013, it was 61% nationally and 65% for London. In the same period, overall numbers of 16 year -olds had fallen and the ethnic minority percentages in London schools increased from 60% to 67% compared with a 21% to 25% increase nationally. The Asian proportion rose to be nearly as large as the Black proportion in London and the Asian pupils as a group score very much higher than Asians nationally and 7.2% higher than London White children. 8 Though a boon to researchers, one questions the costs of collecting, organizing and publishing detailed annual population education attainment data. There is a burden on teachers and children to supply them and the use to which they are put by subsequent informed policy action for improvement, social justice and accountability is not readily evident. --- Source: SFR50_2014_KS2_National Tables ~ Table 10a. National mean = 79 Ethnicity variation is from Black Caribbean 73 to Chinese 88/Indian 86, a range from 6 below mean to 9 above. Source: SFR06_2015_KS4_National_ and_LA~KS4 Tables 1 and2a. National mean = 56.6 Ethnicity variation is from Black Caribbean 47 to Chinese 74/Indian 73 a range from 10 below mean to 17 above.

Attainment data on England's school pupils is more extensive in coverage, detail, quantity, accessibility and of higher quality than monitoring statistics routinely available in other European countries. These data facilitate investigation of low attainment in England's schools, and its relationship to ethnicity, gender and poverty. This paper reviews longitudinal sample studies and extends this with simpler presentations of England's national attainment statistics for education over five years up to 2014. The analyses show recurrent low attainment within specific ethnic minority groups, with gender and most strongly with low income sections of society. There is a strong case, from these data and from other's research, that these inequalities are rooted in social and economic factors outside the school, created and sustained by neo-liberal economic practices and elitist structures. It is argued that reducing the proportion of children growing up in poverty will have a bigger impact on raising average attainment levels than focussing on in-school factors.

Introduction In recent decades, the use of anabolic androgenic steroids has attracted attention as a growing phenomenon among recreational athletes, such as body builders . AAS is one of several performance-enhancing drugs and is a doping substance if not used with a physician's prescription for medical purposes. In international competitive sports, the use of AAS and other substances aimed to increase athletes' performance is regulated by the World Anti-Doping Agency in accordance with a list of substances that are not allowed in the body of participants in elite sports contests . The regulation of doping substances in recreational contexts, however, differs between countriesspecifically, the criminalization of their manufacturing, handling and use, and which doping substances are deemed illegal . Users of doping substances are at risk of negative physical and mental health effects. Specifically, the use of AAS is associated with aggression, depression, infertility, liver and muscle damage, cardiac injuries, arrhythmias, and sudden cardiac death . In addition to individual health problems, AAS is associated with users' violence and crimes against other . The prevalence of doping outside the elite sports context is difficult to measure; however, estimates from Western countries suggest that approximately 1% of national populations use doping, comprised largely of men . Self-reported data from recreational athletes and other subpopulations, however, show higher rates . In a Nordic meta-analysis on the lifetime prevalence of non-medical use of AAS, the authors found a lifetime prevalence of 59.2% among drug users and 26.2% among prisoners and arrestees . Furthermore, prevalence estimates among recreational athletes in eight European countries indicated that 3.1% of men and no women reported doping in 2019 . Finally, self-reported data on AAS use among gym-goers in Stockholm County showed a lifetime prevalence of 3.9% among men, while 1.4% of them had used AAS during the past 12 months . To counteract health and social problems associated with doping substances, Sweden adopted its first doping law in 1991 . The act on the prohibition of certain doping substances covers synthetic anabolic steroids and testosterone, and its derivatives; growth hormone and chemical substances that increase the production and release of either testosterone and its derivatives or growth hormone . In Sweden, in contrast to most other countries, not only are the manufacturing, sale, supply, and possession of doping substances criminalized, but the use of them outside the healthcare sector is also criminalized, which implies that detection in the human body can lead to punishment . The implementation of the Swedish doping law amplified anti-doping work in the recreational sports context from the 1990s onwards . Despite these efforts, the ban has proven insufficient to prevent the use of doping substances in recreational sports . To increase the possibility of preventing doping among recreational gym-goers in Sweden, a multi-component program-100% Pure Hard Training was developed in 2007 . The program has since then been disseminated nationwide and is currently used in around 600 gyms across Sweden. Within the program, the police officers are crucial actors who can enforce the law through supervision at gyms. Owing to concerns about recreational athletes' use of doping, the European Union Commission published a review of the evidence base for policies to combat doping in recreational sports in 2014 . Along with legislative measures and controls, this review concluded that doping prevention in recreational sports relies primarily on education and information. The authors explained that published studies examining the effects of antidoping education programs are rare. Some experts who contributed to the review claimed that educational media campaigns could have the intended effects; however, since most legal, administrative, and political arrangements regarding doping prevention in recreational sports were fairly new in 2014, the report by the EU Commission could not present a consensus on ideal practices . Later, a single study obtained support for combined educational programs and practical strength training, targeting adolescents and students . Another exploratory study suggested that prevention efforts should involve key stakeholder groups, such as coaches . In 2021, Bates and Vinther conducted a literature review on doping prevention interventions and concluded that the evidence base remained underdeveloped and the effects unclear . Given the important role that the police plays in preventing doping in recreational sports in a Swedish context, particularly at gyms, this study examines police officers' perceptions of doping, as it is defined in Swedish law, related problems, and suggestions for effective prevention of doping in the society. This study can contribute to development of the knowledge base concerning recreational doping and how to prevent it, as national regulations are continuously changing based on research and practice. --- Methods --- Design This study used a cross-sectional approach, based on a web survey of employees within the police authority in Sweden. --- Participants and procedure Participants were recruited in three different ways. Initially, three police officers with a coordinating function regarding substance use-related issues and crime prevention invited police officers who had previously attended a two-hour lecture on doping and colleagues who had not attended any known training on doping. The second sample consisted of police officers who signed up for a half day of digital training in police-related doping prevention work, which was provided within the 100% PHT program. Finally, the possibility of participating in this study was communicated through the police authority's own internal channels: the Swedish Narcotics Police Association's magazine and the 100% PHT website. Therefore, the sample was a convenience sample. Potential participants were invited via e-mail and informed about the study purpose and other relevant aspects connected to the Declaration of Helsinki. The e-mail also contained a link to the web survey , which the participants could complete after providing informed consent to participate. The survey was based on previous prevention and implementation research , and it was developed by STADs' researchers in consultation with selected police officers and doping prevention coordinators. The questions covered participants' backgrounds, doping as a societal problem, availability of doping substances, and police's doping prevention activities, along with related barriers and possibilities. The questions had fixed answer options; however, comments and elaborate answers were left in the free-text area following several questions. In addition, free-text questions were included. The survey was pilot tested by police officers and took about 15-20 min to complete. A total of 810 police officers were invited to answer the survey, and data were collected between October 2020 and December 2021. To increase the response rate, several reminders were sent to those who had not yet completed the survey. Data were analyzed using SPSS 27.0 and presented descriptively. The free-text answers were analyzed using qualitative content analysis. The study was performed in accordance with the Declaration of Helsinki, and the protocol was sent to the Swedish Ethical Review Authority . --- Analysis Quantitative data were analyzed using SPSS 27.0 and presented descriptively. Some questions had minor internal non-responses as reported in the respective tables. Mann-Whitney U-tests were used for comparative analyses based on median values and Chi-square tests were used to compare proportions. For all analyses, a p-value of < .05 was considered a statistically significant difference. The qualitative material was analyzed using directed qualitative content analysis , focusing on the literal meaning of words One author repeatedly read the free-text answers to group the content into meaningful categories. Four categories were generated and further discussed with a second author , who agreed on the number of categories, which were named "general action," "action on gyms," "the police's work to curb doping," and "barriers to police's doping prevention activities." --- Results --- --- Doping as a societal problem More than 60 percent of the respondents believed that doping was a large or very large societal problem . Almost all respondents thought that it was their duty to prevent doping, which some respondents underlined by commenting that doping is de facto illegal. Moreover, most reported that they met people who use, or whom they suspect use, doping substances once a month or more often, and almost half reported similar figures for the handling of doping substances . However, several respondents commented that the unrecorded number was probably large because of the lack of targeted efforts against doping and a lack of knowledge. Several respondents commented that doping occurs in connection with other crimes and has become a very common feature in the lifestyle of gang criminals. Almost all participants thought that doping occurred in connection with violent crimes. The corresponding figures for intimate partner violence and drug-related crimes were each more than 80 percent . The perception that doping is overlooked several times, especially in intimate partner violence, was also reported by several participants. --- Availability of doping substances Approximately one-fourth of the participants perceived that the availability of doping substances had increased to a certain extent during the last three years . The respondents commented that this was mainly owing to the increased availability on the Internet, in which substances are sold on encrypted sites. Further, respondents commented that the detection of postal items with doping has increased, which they indicated might be owing to postal companies becoming more attentive. When asked if they experienced a change in the incidence of dopingrelated problems during the last three years, approximately onefourth of respondents answered that they felt there had been an increase to some or a large extent . Comments from respondents revealed the perception that doping has become more common among younger people and that it is commonly used by criminals, not least by gang criminals. However, some participants emphasized in comments that increased knowledge about doping also contributes to the perception that doping has increased. Several also indicated that they perceived that a correct assessment of potential users is difficult to carry out because many who use doping do not necessarily have large muscles; thus, the police may not have reason to suspect them. Just over half the respondents reported that they thought the Internet was the most common way for doping users to obtain doping substances for the first time . Approximately one in five respondents reported that they believed first-time users are introduced to doping through friends and at gyms. --- Prerequisites for police work against doping Most respondents thought that the issue of doping was not a priority in their local police district . Further, just over half of the respondents stated that they were not given the necessary prerequisites to work toward decreasing doping, and just over onethird stated that they were partially given these prerequisites. Reasons for this, such as resource and knowledge gaps in the organization, were also provided in comments. Almost three out of four respondents stated that they lacked further training in their anti-doping work, about 60 stated that there was a lack of human resources, and approximately 30 stated that supervision was missing . They also commented on the need to clarify the link between doping and other crimes and opined that many police officers lack knowledge about how to present evidence for offences that can lead to prosecution. Accordingly, the importance of raising the level of knowledge among investigation leaders so they can make confident decisions that succeed in court was proposed by some participants. Several participants also commented that police officers, including themselves, had to attend doping training to highlight the doping issue. Some said they have trained officers in their local police district who act as liaison officers for gyms. Just over half the respondents requested a simplified procedure in doping tests, stating that blood tests or other rapid tests could be used. This suggestion was put forward due to the consideration that the collection of urine samples is sometimes difficult, especially if necessary permission from the investigation leaders is not given or if the suspect refuses to provide the sample . Many respondents reported that their local police districts have targeted efforts directed at gyms once or twice per year. Some respondents commented that these interventions have been successful but also perceived as insufficient . Just over one-third of respondents reported that such targeted efforts happen once a year, and approximately one-fourth reported that it never happens. About one in ten respondents, however, reported that it occurred only once a month . --- Cooperation with other actors Almost one-third of respondents stated cooperation was lacking in their doping work , while one-third reported that they collaborated with 15 other actors regarding doping issues . By far, the most common actors that the police collaborate with are gyms and training facilities, but they also collaborate with municipal coordinators and the County Administrative Board to some extent. Further, 40 mentioned that they believed that doping is included in the citizens' pledge between the police authority and the municipality, while just over one-third did not think so . In addition to the police authority, almost all respondents reported that they believed that gyms and training facilities should engage in anti-doping work to obtain success . A large proportion of respondents further suggested that several other actors, such as the healthcare workers, school officials, post office workers, youth clinic workers, social service employees, customs officers, and municipal coordinators, should engage in doping prevention. Other actors that were mentioned in the comments were sports association and other nongovernmental organizations. Regarding postal agents, some participants mentioned that the current legislation is a barrier to doping prevention, as the penalty value is not high enough for secrecy in the Postal Act to be breached. The Postal Act contains provisions concerning postal operations and universal postal service. --- Reflections on doping prevention work Finally, the respondents were asked to reflect on what they thought could be done to prevent doping in society. Over half the respondents contributed with reflections, which were qualitatively analyzed and interpreted, and subsequently divided into the following themes: general actions, actions at gyms, police work to curb doping, and barriers to police doping prevention activities. --- General actions Several respondents referred to the fact that young men today obtain their masculine ideals from the media, not least from social media, in which they are given an image of a perfect male body with six-pack abs. Thus, they meant that many men strive for this ideal body and search for quicker ways to attain it. Society as a whole has a twisted image of the male ideal where the man is portrayed as a 25-year-old with a six-pack and big arms/ shoulders in every other commercial for all kinds of products. Visible muscles indicate success. Some respondents suggested that information efforts in schools and training facilities could be an appropriate action to counteract the influence of social media by informing students and the public about the consequences of doping, physically and legally. Inform schools and training facilities about the risks. Today, there is an incredible obsession with how you should look and that you should have quick results. Remind young people that you cannot aspire to look like celebrities and others on social media. One way is to involve social media profiles/influencers in the work against doping. We also need more inspections on training facilities and more effective prosecution. Similarly, it was suggested that sports associations inform young members about doping. Talk about it in schools and sports associations and highlight the consequences of doping. According to respondents, if suspicion exists about doping use, teachers or leaders should talk to the person individually. Schools should also talk about the effects of doping and, if suspected, talk privately with the concerned person. Some respondents also mentioned that the level of knowledge among staff in healthcare settings, including psychiatry, needs to be raised to become better at detecting and dealing with people subject to doping. More people in healthcare and psychiatry who are trained and perceptive to the side effects of doping, such as depression, potency problems, aggression, heart problems [should be engaged]. --- Actions at gyms Several respondents highlighted the importance of more gyms becoming involved in and actively working against doping. Gyms [should] clearly distance themselves from doping. It was further mentioned that gyms, to manifest their policy, need to communicate with their members that they take an active stance against doping; there is visible information in gyms about the consequences of doping; the gyms work with policies and clear procedures for how they manage various dopingrelated situations; they train their staff in the doping issue; and they cooperate with the police when doping is suspected . Suggestions were also made that gyms could have mandatory talks with young people younger than 18years-old who buy a gym membership card, informing them Gyms should possibly "force" young people under the age of 18 who buy membership from them to participate in an information meeting about anabolic androgenic steroids. Further, it was suggested that the membership agreement between the customer and the gym should state that as a member, one approves testing in the event of suspected doping or routine check, and that if one fails to take samples or test positive, they will be suspended from the gym until clean results are provided. They [the gyms] should be able to demand a urine sample if they have their own suspicion. It should be stated in the agreement between the customer and the gym that it is 100% Pure Hard Training and in case of suspicion, you have signed in the agreement/membership that you accept this. Several participants also mentioned the doping prevention method of 100% PHT, and that it is a well-established method that gyms should work with. Further, several participants believed there was a need for good collaboration with gyms to facilitate more frequent and targeted supervisory visits. It was also mentioned that the local police district should appoint liaison officers who are trained in the doping issue and work more closely with the gyms, for example through agreements, to promote cooperation between gyms and the police and enable undercover visits. --- Police's work to curb doping In addition to what was mentioned about how the police can work by targeting more gyms and other such facilities, other proposals for more effective doping prevention were mentioned by the respondents. Primarily, these are about increasing the knowledge level among police officers because doping is often connected to other crimes. Some mentioned that doping should be a compulsory element in basic education at police academies. Make the debriefing clearer, introduce it as a mandatory part of basic [police] education, and increase knowledge internally [in the police organization]. The importance of increasing knowledge among investigation leaders was highlighted as critical because their decisions is needed to collect and present evidence of doping offenses. Respondents also mentioned the need for the police to work more systematically with targeted efforts in the same way as it did for the campain "traffic weeks" in an attempt to counteract vehicle accidents. Finally, participants mentioned that obtaining doping substances often takes place on the Internet, and that it is desirable to have targeted efforts with systematic Internet spying to access suppliers. The police authority must invest resources in targeting [doping] suppliers via Internet surveillance. --- Barriers to the police's doping prevention activities A recurring barrier to doping prevention activities that was reported by the respondents, and outlined above, is the low level of knowledge about doping within the police force, prosecutors, and the "judicial ombudsmen" -an authority that examines that other authorities work in accordance with the laws and regulations that govern their work. One respondent described a case that began with suspected drunk driving, which was then extended to doping offenses after taking the offender into custody because of clear signs of intoxication. This was judged to be a reasonable basis for the decision to search the offender's house. During the search, doping substances were found. However, the case was, according to the respondent, reported to the JO, who assessed that the circumstances leading to the decision to search the house were insufficient for a reasonable suspicion that the suspect had doping substances at home, which meant that there was no basis for the decision to search the house. Even though we had found doping [in the house] the JO said, ": The circumstances that the decision-maker has reported [decision on house search] in this case cannot be considered sufficient grounds for reasonable suspicion that [X] had doping agents in the home at the time in question." According to the respondent, this circumstance could have resulted in police officers not making this type of decision in future cases. Another respondent reported a barrier to testing urine samples, claiming that since individuals can refuse testing there is a need to implement mandatory testing, like with a catheter. Expand our coercive measures on those who are suspected, such as locking up or catheterizing those who refuse to urinate! Simplified tests were requested; for example, through blood tests or other rapid tests. Further, the relatively low penalty for doping offenses was mentioned as a barrier to the preventiontion efforts. Suggestions were made to change this rule to address the issue of doping. Moreover, participants found it difficult to calculate doping doses when reporting the use or handling of doping in connection with doping-related crimes because there are less-known substances found during house searches, which makes reporting of doping crimes time-consuming. According to the respondents, this, in combination with a relatively low penalty value compared with that for crimes related to narcotics and low knowledge of legislation around doping risks, can impede doping prevention activities. Penalty values are low. These factors, combined with the fact that far from all police officers have knowledge of, or routine use of doping legislation, risk creating a "vicious circle". You do not report because you are unsure, you are unsure because you do not have sufficient knowledge or training. The Postal Act was also highlighted as a barrier to the detection of doping substances in postal items because the post staff are not allowed to open items owing to the low penalty value. --- Discussion This study examined police officers' perceptions of doping, as it is defined in Swedish law; related problems; and suggestions for effective prevention of doping. Results showed that most respondents consider doping a societal problem and that doping occurs in combination with several other crimes, such as violence, illicit drug use, and drug dealing. These results are supported by a recent interview study with police officers in Sweden . Almost all respondents also believe it to be their duty to prevent doping but lack proper prerequisites to do so, emphasizing the opinion that there is a lack of knowledge in the police organization. This opinion was supported by the fact that approximately 75% of respondents reported that they lacked appropriate training to work against doping. More knowledge is, according to the respondents, perceived to be needed at all levels in the police organization; for example, municipalities police officers or intervention police officers must know how to recognize doping users and how to calculate doses when reporting suspected crimes. Further, investigation leaders, who give mandates for sampling and arrest, need to be properly informed about what is needed to facilitate the collection of evidence connected to doping offenses, according to the respondents. Additionally, respondents believe that managers in the police organization need to be better informed about doping to prioritize preventive activities. As previous implementation research supports the idea that sufficient knowledge is crucial for effective implementation , the strengthening of basic education for police officers and further training within the police organization could abolish the barrier of ignorance . Respondents further stated that quicker doping tests could accelerate the collection of evidence when building a doping offense case. Previous studies have with promising results assessed alternative test procedures, such as hair, saliva, or breath tests, which could make the collection of evidence more effective , however, to our knowledge, there are no commercially available instruments for testing via hair, saliva, or breath. Using a catheter to obtain urine samples from suspected individuals or using blood tests was also suggested by the respondents. However, this method is currently not allowed, at least not if coercion is needed, highlighting the relatively low penalty value for the doping use offense . Good collaboration with gyms to facilitate police work was also emphasized by respondents as important for effective doping prevention. This coincides with a previous European study that emphasized collaboration between key actors to counteract doping among recreational athletes . A success factor in the current context, noted by respondents, was that local police districts appoint special liaison officers for gyms to communicate about doping-related issues. Collaboration agreements between gyms and the police, including enabling police officers to exercise undercover in the training premise to detect doping-related crimes, could also, according to the respondents, constitute facilitating factors. This view was previously highlighted in interviews with Swedish police officers . Several respondents also highlighted the importance of more gyms working actively against doping, preferably in line with the 100% PHT doping prevention method, which aligns with suggestions of field experts . This can include systematic policy work with clear routines for how gym staff handle different doping-related situations, staff communication with members that the gym takes a stand against doping, making information about the negative consequences of doping visible, and cooperation with the police . Another barrier reported by respondents was the lack of economic and personnel resources to prevent doping. This barrier was previously highlighted in a study among key European actors in recreational sports , general implementation research , and a recent Swedish interview study among police officers . Thus, additional resources appear to be needed to increase the effectiveness of doping prevention. As previously indicated, increased knowledge about doping among managers in the police organization could amplify this priority . To decrease the availability of doping in society, several respondents highlighted the need to access postal items to detect doping substances. Currently, this is not permitted because doping offenses have relatively low penalty rates . Finally, suggestions for decreasing the doping demand in society; i.e., to convince people not to use doping substances, were put forward by respondents. These suggestions included engagement from other authorities and organizations, which coincides with previous research on the prevention of illicit substance use . Since doping has negative mental and physical health effects , healthcare sector professionals can play an important role in doping prevention in their encounters with users who could end up in psychiatry or other departments owing to doping side effects. Further, it is widely recognized that efforts to prevent harmful behaviors should be performed early in life. Therefore, schools and sports organizations with children and adolescents as pupils or members could be suitable areas for doping prevention interventions . Doping prevention in schools was recently supported by a systematic study of methods for preventing doping among young people ; however, it has been difficult to indicate that the interventions changed students' attitudes toward doping . This study has several strengths that should be highlighted. It is based on responses from police officers from a wide range of regions in Sweden with various experience as police officers, different functions, and representing both men and women. An additional advantage is that the response rate is relatively high. The survey is also unique from a national and international perspective, as little has been published about police officers and their views on doping prevention. This study also has certain limitations. The most prominent is the selection of respondents within the police authority or registered for a police training course on anti-doping work. This means that there could be a higher interest in addressing doping and related issues among respondents in this study than among police officers in general. However, only half the respondents reported knowing about the nationwide disseminated doping prevention method for 100% PHT, which indicates that several of them had low engagement in doping issues. Further, the structure of some of the response options was unbalanced with unequal numbers of positive and negative answer options, which could have contributed to response bias. Additionally, the instrument development, data collection, data interpretation, and data analysis were conducted by researchers who operate at a research and development unit whose practitioner group have been involved in the development of an anti-doping program. The respondents could also have fallen prey to social desirability bias . Although the proportion of women in our sample did not differ from the police force as a whole, the generalizability to the Swedish police force as a whole can be questioned as our sample was overrepresented in some regions and underrepresented in other regions. Moreover, the generalizability to the international context is partly limited since the Swedish context differs from most other countries regarding the prohibition of the use of doping substances. Finally, data collection took place during the COVID-19 pandemic, which affected the gyms' operations and could have influenced respondents' answers. --- Conclusion Police officers perceive doping to be a societal problem and are motivated to work against it. Those that participated in this study highlighted several areas for enhancing doping prevention, such as increasing knowledge in the police organization, legislative changes to enable simplified routines for doping tests, permission to control postal items, intensified doping prevention at gyms, and commitment from other authorities and organization to engage in doping prevention. --- Data availability statement --- The data are available from the Centre for Psychiatry Research, a collaboration between the Karolinska Institutet and Region Stockholm. However, restrictions apply to their availability, as they were used with ethical permission for this study and, therefore, are not publicly available. The data are available from the authors upon reasonable request and with permission from the Centre for Psychiatry Research. --- Ethics statement The studies involving humans were approved by the Swedish Ethical Review Authority. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Introduction: The use of anabolic androgenic steroids among recreational athletes has received growing attention in recent decades. Several countries have implemented bans on doping; however, recreational athletes and other subpopulations continue to use doping substances. Recognizing that the police play a crucial role in preventing the use and dealing of doping substances in Sweden, efforts have been made to intensify police interventions and enhance collaboration with other key actors. This study examined police officers' perceptions of doping as defined in Swedish law, related problems, and suggestions for effective prevention of doping in the society. Methods: A cross-sectional survey study was conducted using a web survey of police officers (N = 597). Data were analyzed using descriptive statistics and free-form text responses were analyzed using content analysis. Results: Participant responses to the survey (73.7% response rate) indicated that approximately 62.6% thought that doping is a societal problem, and approximately 26% perceived that the availability of doping substances has increased over the past three years. A total of 95.6% of respondents believed that doping occurred in connection with other crimes such as intimate partner violence (88.2%) and drug-related crimes (88.0%). Further, 96.3% of respondents perceived that it was their duty to prevent doping, but 63.8% indicated that doping-related work was not prioritized within their local police district. Discussion: Police officers perceived doping as a societal problem and expressed motivation to counteract it, highlighting increased knowledge, legislative changes, intensified doping prevention in gyms, and commitment from other societal actors to increase the effectiveness of doping prevention. Suggestions for increasing the efficiency of doping prevention included education and increased knowledge at all levels in the police organization, intensified prevention efforts at gyms, legislative changes to permit simplified doping test procedures, and breach of secrecy for postal items. There was also a suggestion for further engagement from other actors, such as healthcare workers, school officials, and non-governmental organizations.

INTRODUCTION Alcohol use can cause much harm to the individual user as well as to society ). Hence, identifying determinants of alcohol use is of public interest. Previous studies suggest that disclosure of alcoholrelated content on social media indicates concurrent alcohol use , while exposure have in some studies been found to predict later alcohol use . Social media could thus be an arena for detecting and preventing problematic alcohol use . The association between disclosure and exposure and alcohol use may depend on type of content shared or seen , but the relationship between different types of alcohol-related content and alcohol use has not been investigated. The causal mechanisms underlying the relationship between alcohol-related disclosure and exposure and alcohol use are not fully understood although it has been suggested that common factors may be at play. It has further been suggested that disclosure and exposure of alcohol-related content may be a mere reflection of alcohol use or that disclosure and exposure may instigate alcohol use . Certain characteristics seem to increase the likelihood of both disclosure and exposure of alcohol-related content on social media and high alcohol consumption . Status as single and extroversion are both positively associated with disclosure of alcohol-related content and alcohol consumption . Certain aspects of social media use, like the number of online-friends, seem also to be positively associated with disclosure and exposure and alcohol use . Associations between different aspects of social media use and alcohol use are, however, not consistently found . Although studies on the relationship between alcohol disclosure and exposure on social media and alcohol use have controlled for factors such as gender, age, and number of online-friends , no previous study, thus far, has controlled for the wider range of demographic, personality, and social media factors that may influence this relationship. Disclosure of alcohol-related content may be a direct reflection of the sharer's alcohol use . Disclosures could also be a reflection of alcohol-related social cognitions , which are known to predict alcohol use . Moreover, exposure to alcohol-related content could be associated with alcohol-related social cognitions as well . Individuals with high alcohol intake and positive attitudes toward alcohol may be more attentive to alcohol-related content on social media, and hence more frequently exposed to such content as studies have shown stronger attention bias toward alcohol among heavy social drinkers than light social drinkers . Frequent exposure to alcohol-related content may also indicate high alcohol intake in the individual's social network , and peers' alcohol use is commonly known as a strong predictor of own alcohol use . According to this line of reasoning, some studies have indicated that controlling for alcohol-related cognitions and/or alcohol use could weaken or eliminate the association between disclosure and exposure to alcohol-related content and future alcohol use . Disclosure and exposure of alcohol-related content on social media have also been suggested to cause an increase in alcohol use . Disclosing content on social media could lead to a stronger commitment to continue to act in accordance with the attitudes and behaviors that were displayed in order to maintain a coherent self-image . Receiving positive feedback on alcohol-related posts have been suggested to further enhance drinking through positive reinforcement mechanisms . Exposure to alcohol-related content has also been suggested to have direct causal effects on alcohol use through mechanisms such as mere exposure effects , and indirectly through altering attitudes and perceived norms for alcohol use . The latter claim is supported by experimental studies showing that exposure to alcohol-related content on social media strengthened the receivers' perception of drinking norms among peers . Previous studies have linked disclosure and exposure to alcohol use, but few of these studies have been based on longitudinal or experimental designs . The directionality between disclosure and exposure and alcohol use is therefore unclear. The current study sought to examine the longitudinal relationship between disclosure and exposure of different types of alcohol-related content on social media and later alcohol use, and to identify covariates that may explain these relationships. --- MATERIALS AND METHODS --- Procedures and Sample All students registered at the four largest institutions of higher education in Bergen municipality, Norway, were during fall 2015 invited to participate in an online survey. A total of 11,236 agreed to participate. Participants from the first wave were invited to participate in a follow-up online survey during fall 2016. A total of 5,217 agreed to participate in the follow-up survey. The majority of the former participants were contacted by their student e-mails, while some were contacted by their private e-mails. Approximately 40% of the students ended their education between the first and the second wave of the survey of some reason , according to the institutions from which the sample is recruited. Based on the assumption that 25% of the former participants which were contacted by their student e-mail did not receive the invitation to participate in the follow-up survey, we estimated that about 61.2% of the ones who received an invitation agreed to participation. A rate of ∼40% ending their education yearly may be a somewhat high rate in an international perspective. This might be related to the wide availability of higher education in Norway which may cause Norwegian students more likely to quit their education before completing a degree, as quitting will be less associated with financial loss compared to the cause in many other countries. The study protocol was approved by the Regional Committee for Medical and Health Related Ethics, Western Norway , and the Norwegian Data Protection Authority . The study was conducted according to the principles expressed in the Declaration of Helsinki. In the beginning of each survey the participants were met with an informed content page, were they could choose whether or not they agreed to participation. --- Measurement Demographics Demographic variables were measured by questions about birth year, sex, place of birth, and religious identification at the first wave . The participants were asked about relationship status, and parental status in both waves . At T1 the participants were asked "Have you changed educational institution the last year?" . --- Personality At T1 the five factor model's personality traits were assessed by Mini-International Personality Item Pool . The Mini-IPIP consists of 20 items , where the respondents are asked to rate the degree specific statements regarding behavior are typical for them. Response alternatives range from "very wrong" to "very right" , some items are reversed. Total scores range between 5 and 20 for each trait, where higher scores indicate higher levels of the personality trait in question. The internal reliability of the measurements in the current study was acceptable. The items measuring extroversion, agreeableness, conscientiousness, neuroticism and openness had Cronbach's alphas of 0.83, 0.77, 0.69, 0.75, and 0.74, respectively. Selfmonitoring was measured by the revised Self-Monitoring Scale, comprising 13 statements in which respondents are asked to rate the applicability of the statements to own behavior . Response alternatives range from "certainly always false" to "certainly always true" , some items are reversed. Composite scores range between 0 and 65, where higher scores indicate higher levels of self-monitoring. The internal reliability of the revised Self-Monitoring Scale was acceptable in the current study, with a Cronbach's alpha of 0.82. --- Social Media Use In the first wave participants were asked several questions regarding social media use; if they had an account on a social media site, which sites they used , number of online-friends, and frequency of logins to social media . Disclosure and exposure of alcohol-related content on social media were assessed by the following questions: "How often do you post content on social media that": "Refers to positive consequences of alcohol use ?, " and "Refers to negative consequences of alcohol use ?" . Similar questions were asked regarding the frequency of exposure to alcohol-related content. The participants were asked to think only of alcohol-related content that were visible to more than two persons. For the questions regarding exposure to alcoholrelated content, the participants were instructed to think only of posts etc. from online-friends or individuals they follow on social media. --- Alcohol Use and Alcohol-Related Cognitions Alcohol use was assessed at T1 and T2 by the Alcohol Use Disorders Identification Test , comprising 10 items . The respondents are asked to assess their own alcohol use the past year and indicate how often they consume alcohol, how many alcohol units they drink on a typical drinking occasion, how often they drink more than six alcohol units, and how often they experience different adverse consequences related to their alcohol use . The response alternatives vary somewhat, but answers to the different items are given a value between 0 and 4. Total scores on AUDIT range from 0 to 40, where higher scores indicate higher alcohol consumption and more frequent occurrence of alcohol-related harm. Scores over seven are considered to indicate risky drinking . The Cronbach's alpha for AUDIT at T1 and T2 was 0.78 and 0.79, respectively. The measurement of alcohol use at T1 had a strong correlation to the measurements at T2, with a Pearson correlation of 0.80. Descriptive norms for alcohol use and the participants' prototypic apprehension of the typical heavy drinker and the typical sharer of alcohol-related content on social media were assessed at T1. Prototypic apprehensions were measured by the following questions: "What is your overall impression of the typical student that drinks six alcohol units or more on a regular drinking night?, " and "What is your overall impression of the typical student posting alcohol-related content on social media?" Response alternatives ranged from one to 10 . Descriptive norms were assessed by the questions: "Think about the five students you know best. How many of them do you think drink": "alcohol a couple of times a week?, " "10 alcohol units or more on a typical drinking occasion?, " and "6 alcohol units or more a couple of times a week?" . Similar questions were asked to assess descriptive norms for drinking among online-friends, but for these questions the participants were asked to think about the five individuals of which they see most posts from on social media. The answers to the three questions regarding descriptive norms for alcohol use among co-students and among online-friends were summarized. Total scores on descriptive norms for alcohol use among co-students and online-friends thus ranged between 0 and 15, respectively. In the current study the Cronbach's alpha for descriptive norms for alcohol use among co-students was 0.69 and the Cronbach's alpha for descriptive norms for alcohol use among online-friends was 0.72. --- Analysis Data analyses were conducted with IBM SPSS Statistics 23, R, and Mplus 7. Missing data were deleted listwise. Descriptive analyses were conducted to identify the sample's central tendencies on the study variables. To check for dropout bias, the current sample was compared to the participants that only participated at T1 on a range of variables by the use of independent sample t-tests and chi-square tests. Cohen's d or phi coefficients were calculated as an indicator of effect-sizes. By conventional standards Cohen's ds of 0.20, 0.50, and 0.80 represent small, moderate, and large effect sizes, respectively . For phi coefficients, 0.10, 0.30, and 0.50 represent small, moderate, and large effect sizes, respectively . Significance tests of difference were conducted for both the independent sample t-tests and the chi-square comparisons. Further, we conducted equivalence tests for the independent sample t-tests and chi-square tests. Equivalence tests of group difference are based on the assumption that Cohen's ds or correlations which are significantly smaller than a given value indicate the absence of any practical meaningful differences . In the current study effect size cut-offs were based on power analyses with 90% power. Alcohol use among those participants who reported to frequently disclose or frequently were exposed to alcohol-related content referring to positive and negative aspects of alcohol were compared to the alcohol use of those who reported low-frequent disclosure or exposure. Baseline alcohol use refers to the participants AUDIT-scores at T1, and later alcohol use refers to AUDIT-scores at T2. Frequent disclosure/exposure was defined as reporting disclosure/exposure monthly or more often. Crude, partly adjusted, and fully adjusted linear regressions were conducted to investigate the association between disclosure and exposure of alcohol-related content and later alcohol use and changes in alcohol use. Covariates were controlled for to investigate which factors could explain the relationship between disclosure and exposure and later alcohol use. The dependent variable was AUDIT-score at T2 and change in AUDIT-score . The four main independent variables of interest in the current study were: frequency of disclosure of alcohol-related content depicting positive aspects of alcohol, frequency of disclosure of alcohol-related content depicting negative aspects of alcohol use, frequency of exposure to alcohol-related content depicting positive aspects of alcohol, and frequency to exposure of alcohol-related content depicting negative aspects of alcohol use. The disclosure/exposure variables were dichotomized into frequent disclosure/exposure vs. low-frequent disclosure/exposure. The first regression-models were crude, where no covariates were controlled for. The second models were adjusted for demographics and personality factors . The third models were adjusted for other aspects of social media use . The fourth models were adjusted for T1 alcohol use . The fifth models were adjusted alcoholrelated cognitions . Finally, fully adjusted regressions models were run. All the independent variables, with the exception of changes in relationship-, childcare-, or student status between the first and the second wave, were based on measurements from the first wave. Completely standardized betas are reported for the different regression models as an indicator of effect size; completely standardized betas of 0.10, 0.30, and 0.50 represent small, moderate, and large effect sizes, respectively . --- RESULTS Table 1 illustrates the sample's characteristics, and a dropout analysis comparing the sample to the participants that only partook at T1. The sample's mean age at T2 was 25.8 years, 64.8% were women, 92.7% were born in Norway, and 83.7% were still students at T2. There were only few significant differences between the sample in the present study , and those that only participated at T1. Equivalence could not be established for all variables which were included in the dropout analysis. The effect sizes regarding the differences between the group that participated at both T1 and T2 and the group that only participated at T1 were, however, within the range of what is considered as very small. The sample had a mean reduction in AUDIT-score of 0.6 from T1 to T2. The frequent disclosure and exposure groups' alcohol use compared to the low frequent disclosure and exposure groups' use are shown in Tables 2, 3. The frequent disclosure/exposure groups had significantly higher AUDIT-scores at T1 and T2 compared to the respective lowfrequent disclosure/exposure groups. The frequent disclosure Number of 5 closest online-friends that drink 6 units or more a couple of times a week 1.1 1.0 Cohen's d = 0.058 N.E,**,2 --- SOCIAL MEDIA USE AND ALCOHOL-RELATED DISCLOSURE AND EXPOSURE Frequency of login to social media e 6.6 6.7 Cohen's d = 0.042 Frequent disclosure of content reflecting positive aspects of alcohol f 9.4% 9.6% Phi = 0.004 E., N.S. and exposure groups had a higher distribution of risky drinking at T1 and T2 , compared to the low-frequent disclosure and exposure groups. Disclosing alcohol-related content at T1 was particularly indicative of risky drinking both at T1 and T2. --- Disclosure and Exposure and Alcohol Use When Controlling for Covariates The relationship between frequent disclosure and exposure and later alcohol use, when controlling for different covariates, are shown in Table 4. Controlling for demographic and personality factors reduced the strength of the relationship between frequent disclosure and exposure and later alcohol use, but the reduction was small compared to the reduction seen when the other covariates were controlled for. The association between frequent disclosure and exposure was further reduced when other aspects of social media use was controlled for. Controlling for baseline alcohol use involved the largest weakening of the association between all types of disclosure and exposure and later alcohol use. The association between frequent disclosure and exposure of content referring to positive aspects of alcohol and later alcohol use were, however, still significant even when baseline alcohol use was controlled for. Controlling for alcohol-related cognitions resulted in a weakening of the association between frequent disclosure and exposure and later alcohol use as well. Only the association between frequent disclosure of content referring to positive aspects of alcohol and later alcohol use remained significant when all covariates were controlled for. Participants which frequently disclosed content referring to negative aspects of alcohol or were frequently exposed to content referring to positive or negative aspects of alcohol had a significant reduction in their AUDIT-scores from T1 to T2, compared to the respective low-frequent disclosure and exposure groups. This reduction was eliminated when baseline alcohol use was controlled for. Participants who reported frequent disclosure of content referring to positive aspects of alcohol experienced an increase in later alcohol use, compared to the participants that reported low-frequent disclosure of such content. Regression models with large sample sizes and single items measurements could involve an increased risk of conducting type I errors . The regression models were subsequently conducted using structural equation models to ensure that the results found in the regression models were robust. Alcohol use at T1, the five factor model's personality traits, self-monitoring, descriptive norms for alcohol use among co-students and among online-friends, and alcohol use at T2 were latent variables in the structural equation models. These models yielded similar results as the reported regression models. --- DISCUSSION Our findings suggest that frequent disclosure and exposure to alcohol-related content on social media is positively associated with both baseline and later alcohol use. This supports the notion that social media can be a suitable arena for detecting problematic alcohol usage. In particular, disclosure of alcohol-related content was related to high alcohol use, which is in line with previous research . It should be noted that the whole sample demonstrated a reduction in alcohol use over time. The frequent disclosure and exposuregroups had higher alcohol consumption compared to the lowfrequent disclosure and exposure groups. Some of the disclosure and exposure-groups did, however, also have a larger reduction in alcohol use , compared to the respective low-frequent disclosure and exposure groups. We speculate that the larger reduction in alcohol use observed among some of the disclosure and exposure-groups could be explained by their initial high alcohol use, as individuals tend to regress toward group means over time . In line with this, the disclosure and exposure-groups relative reduction in alcohol use disappeared when baseline alcohol use was controlled for. Frequent disclosure of content referring to positive aspects of alcohol use was associated with stable or slightly increasing alcohol use over time , compared to the reduction in alcohol use found among participants with low-frequent disclosure of such content. The relationship between disclosure and exposure and baseline and later alcohol use differed based on the type of alcohol-related content shared or seen. Disclosure of content referring to negative aspects of alcohol use was a stronger indicator of alcohol use than disclosure of content referring to positive aspects of alcohol use, when no covariates were controlled for. The experience of adverse effects is considered to be an important indicator of problematic alcohol use . Accordingly, disclosing content related to negative aspects of alcohol may serve as a measure of problematic alcohol usage. However, disclosing content referring to positive aspects of alcohol use were identified as a stronger indicator of later alcohol use when the different covariates were controlled for, which may suggest that disclosure of such content predicts stabile or increasing alcohol use when other factors are hold constant. In addition, exposure to alcohol indicated later alcohol use; here the relationship was strongest for content reflecting positive aspects of alcohol use. However, the relationship between exposure and later alcohol use was not significant when all covariates were controlled for. The association between disclosure and exposure and later alcohol use was mitigated when controlling for demographics and personality factors, and for other aspects of social media use. Disclosure and exposure have been or may be linked to certain demographic, personality, and social media use factors which again have been linked to increased alcohol use . The current results nevertheless suggest that the relationship between disclosure and exposure and later alcohol use cannot be fully attributed to such confounding factors. Disclosure and Exposure Was Strongly Associated with Baseline Alcohol Use and Alcohol-Related Cognitions The association between disclosure and exposure and later alcohol use was substantially reduced when adjusting for alcohol use at baseline. Few studies have controlled for baseline alcohol use, but the current findings are similar to the findings from comparable studies . Disclosure as an indicator of later alcohol use has not been investigated longitudinal before, while controlling for baseline alcohol use. The current findings suggest that disclosure and exposure primarily reflects baseline alcohol use, that high alcohol use predicts both disclosure and exposure to alcohol as well as further alcohol use. On the other hand, it should be noted that the measurements in the current study were separated by 1 year. It is possible that the temporal relationships between disclosure and exposure and later alcohol use are either too short-lived or take longer time to become established than the one-year follow-up time used in present study. In addition, the strong correlation between alcohol use at T1 and T2 and the clear association between alcohol disclosure/exposure on social media and high alcohol use, suggest that these behaviors might be intertwined and reciprocally reinforcing. Thus, the issue of directionality seems hard to determine with the current design. Individuals with high alcohol use may display and be exposed to more alcohol use in both online and offline settings and continue to have a high alcohol use over time. Hence, the potential separate effects of online alcohol disclosure/exposure on further alcohol use might be hard to detect. The current results nonetheless suggest that alcohol disclosure/exposure's potential effects on further alcohol are likely to be small in populations were alcohol habits are pre-established, e.g., college/university student populations. Adjusting for baseline norm perceptions and attitudes also resulted in a weakening of the association between disclosure and exposure and later alcohol use. The current finding suggest as such that the association between alcohol disclosure/exposure on social media and later alcohol use, could in part be explained by an association between disclosure/exposure and lenient alcoholrelated cognitions . --- Disclosing Positive Alcohol Content May Influence Later Alcohol Use Frequent disclosure of content reflecting positive aspects of alcohol use was the only independent variable that predicted later alcohol use, when all covariates were controlled for. Disclosure of content related to positive aspects of alcohol use may reflect attitudes or experiences regarding alcohol use not measured in the current study. Disclosure of positive alcohol-related content could suggest that the individuals have positive attitudes toward alcohol or alcohol-related cognitions, which were not detected by the alcohol-related cognitions questions included in the present study . Disclosures of alcohol-related content referring to positive aspects of alcohol could also indicate that the individual is experiencing pleasurable effects of alcohol. Positive alcohol-related cognitions and the experience of positive consequences are further considered as important motivational determinants of further alcohol use . Disclosing content related to positive aspects of alcohol might, however, also represent a causal influence on later alcohol use. Posting alcohol-related content referring to positive aspects of alcohol use may make the senders' attitudes toward alcohol more positive, through mechanisms such as self-identification with own postings, potential self-fulfilling prophecies, and cognitive dissonance . Disclosures of positive alcohol-related content may also cause a later increase in alcohol use through likes and other types of virtual appraisals acting as positive reinforces . A previous study found disclosure of content referring to positive aspects of alcohol to yield more likes than disclosure of content referring to negative aspects of alcohol . The increased number of likes associated with alcohol-related content reflecting positive aspects of alcohol may explain why disclosure of this type of content predicted later alcohol use, even when all covariates were adjusted for. --- Limitations and Strengths The present study is not without limitations. The measurements are based on self-report, and responses to self-report questions may be affected by social desirability and recall biases . Another limitation with the current study is that some concepts were measured by single-items , which could make the results more likely to be affected by measurements errors . In addition, some variables that may explain the relationship between disclosure/exposure and alcohol use were not included in the study. Finally, the study's sample consisted of Norwegian students, which may limit the generalizability of the current findings. A major strength with the present study is the longitudinal design, and the study can hence give an indication of directionality and causality . The high correlation between alcohol use at T1 and T2, and the strong association between concurrent alcohol use and alcohol disclosure/exposure do, however, hamper conclusions regarding directionality. Further, it is important to note that causality cannot be established by the current research design . The study provides new knowledge regarding the relationship between different types of alcohol-related content and later alcohol use, and the importance of different covariates in this relationship. The comprehensive set of covariates included and controlled for, and the large sample size are further strengths of the present study, which clearly distinguishes the current study from any previous studies in the research field. --- CONCLUSIONS Frequent disclosure and/or exposure to alcohol-related content indicate consecutive high alcohol intake. The association between disclosure/exposure and subsequent alcohol use was considerably weakened when baseline alcohol use was adjusted for. This finding might suggest that disclosure/exposure primarily reflects baseline alcohol use. Alcohol disclosure/exposure on social media and high alcohol use might, however, be intertwined and reciprocally reinforcing behaviors making the separate effects of online alcohol disclosure/exposure on further alcohol use hard to detect. More research and experimental studies are required to make final conclusions regarding directionality and causality in the relationship between alcohol disclosure/exposure on social media and alcohol use. Disclosing content reflecting positive aspects of alcohol may have a self-enhancing effect on the sharers' alcohol use and can predict stable or slightly increasing alcohol consumption over time when other factors are hold constant. The relationship between disclosure of content referring to positive aspects of alcohol and later alcohol use may, however, also be explained by potential third variables not included in the current study, e.g., positive alcohol experiences. --- --- FUNDING Research reported in this publication was funded by the University of Bergen and the Bergen municipality. --- Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Copyright © 2017 Erevik, Torsheim, Andreassen, Vedaa and Pallesen. This is an open-access article distributed under the terms of the Creative Commons Attribution License . The use, distribution or reproduction in other forums is permitted, provided the original author or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

This article aims to investigate whether alcohol-related disclosure and exposure on social media can predict later alcohol use, and to identify covariates in these relationships. Data were collected by online surveys (two waves) among students in Bergen, Norway. The first survey was administered in fall 2015. The follow-up took place during fall 2016. A total of 5,217 students participated in both waves. The surveys included questions about demographics, personality, alcohol use, alcohol-related cognitions (e.g., attitudes and norms), social media use, and disclosure and exposure of alcohol on social media. Bivariate comparisons were conducted to assess differences in alcohol use between the frequent (i.e., monthly or more often) disclosure and exposure groups and low-frequent disclosure and exposure groups. Crude and adjusted linear regressions were employed to investigate if disclosure and exposure of alcohol could predict later alcohol use, when controlling for a range of covariates. Compared to the low-frequent disclosure and exposure groups, participants which frequently disclosed or were frequently exposed to alcohol-related content had higher alcohol use at baseline and 1 year later (p < 0.001), when no covariates were controlled for. Frequent disclosure of content reflecting positive aspects of alcohol predicted stable or slightly increased alcohol use at Time 2 (p < 0.01), even when all covariates (i.e., demographics, personality, alcohol use, alcohol-related cognitions, and social media use) were controlled for. In conclusion, frequent disclosure and/or exposure to alcohol-related content predicted alcohol use over time. Alcohol disclosure/exposure on social media could for the most part not predict later alcohol use when baseline alcohol use was controlled for. High alcohol use and alcohol disclosure/exposure on social media appear to be strongly intertwined, which hampers identification of directionality between alcohol use and disclosure/exposure. Disclosing content reflecting positive aspects of alcohol was the only independent variable that could predict further alcohol use when other factors, like baseline alcohol use, were held constant. This finding suggests that disclosure of alcohol content reflecting positive aspects of alcohol might have a self-enhancing effect on the sharers' further alcohol consumption, or that disclosing such content could indicate lenient alcohol-related cognitions not detected by the current measurements.

Globally, HIV incidence and associated morbidity and mortality have been on the decline in heterosexual populations . However, HIV epidemics among men who have sex with men continue to expand, and despite being a small minority group , MSM now account for more than 50% of new HIV cases in the majority of higher-and middle-income countries, including Thailand and the United States . The high HIV incidence among young MSM was documented in multiple locations, including Southeast Asia . This development is especially alarming given that, globally, MSM accounted for 12% of new infections in 2015. Data reported by countries across the world show that HIV prevalence among gay men and other men who have sex with men often is substantially higher than it is among than the general population . The HIV prevalence in the general population in Thailand has declined to ~1% since peaking at slightly over 2% during the mid-1990s . In contrast, the HIV prevalence among MSM recruited from entertainment venues and sampled in greater Bangkok has almost doubled, from 17.3% in 2003to almost 30% in 2012 without any sign of abating. This increase in HIV prevalence largely was accounted for by high HIV incidence in young MSM ≤ 21 years of age. Data from clinic and cohort settings in Bangkok have confirmed a high ongoing HIV incidence among YMSM varying from 8.8 to 12.2 per 100 person-years during 2005-2012 , much higher than among their older counterparts. There are several behavioral risk factors for HIV infection that may help explain the high incidence among YMSM globally. YMSM express significant variability in terms of risk behavior, including alcohol use, drug use, and sexual risk . Other studies have implicated drug use, connectedness to the gay community, peers as agents of behavioral change, gay sex partnering apps, and Internet sex-seeking as correlates of sexual risk-taking among YMSM . Substance use and social environmental context have also been associated with high-risk sexual behavior among African American and Latino HIV-infected YMSM in New York City . The studies about behavioral factors affecting elevated HIV incidence among the youngest group of MSM in Thailand were conducted but not sufficient. Some quantitative studies found that alcohol and drug use are important factors among MSM in Thailand and the United States . Moreover, YMSM who attend sex parties are more likely to have more sexual partners and to use drugs than those who do not . Finally, social media is frequently mentioned by YMSM population as having both positive and negative roles influencing their behaviors . However, it is not known how these factors have changed over time and to what extent they contribute to the likelihood for HIV infection among YMSM today. We conducted this study to understand and contextualize the HIV epidemic in YMSM in Bangkok, Thailand. --- Methods This qualitative study was conducted within the framework of the Bangkok MSM Cohort Study , a longitudinal investigation of prevalence, incidence, and risk factors for HIV infection among 1,744 men. It is a cohort study with follow-up visits every 4 months conducted between April 2006 and July 2012. Participants were homosexual Thai men, at least 18 years old and residents of Bangkok . We conducted eight focus group discussions consisting of four to eight participants per group and 10 key informant interviews in two phases. Of all the FGD participants, 10 participants were selected to participate in KIIs. In phase one, five FGDs and five KIIs happened between June and August 2012. In phase two, the remaining three FGDs and five KIIs occurred between March and June 2013. FGDs and KIIs were used to explore factors contributing to the elevated risk for HIV infection among YMSM. Topics discussed included peer pressure, Internet use, use of methamphetamine and sexual enhancement drugs, sex parties, sex on drugs, and related issues . No personal identifiers were collected. Participants were asked to use a fictitious nickname to address each other throughout the focus group and key informant interview. FGDs and KIIs were audio-recorded, transcribed, and analyzed using qualitative software. Sociodemographic and behavioral data were collected using a paper-based self-administered short questionnaire, and data were key-entered into the database. No personal identifiers were used. Focus group discussions and key informant interview participants were reimbursed 500 baht for their time and cost of travel. --- Study population Participants were selected using a purposive sampling method. Inclusion criteria were Thai nationality, age 18-24 years, uninterrupted availability for two hours during the time of the FGD or KII, and a history of oral or anal sexual intercourse with another man in the preceding six months. Seventy-three participants were recruited from the BMCS and from individuals presenting for HIV testing services at the Silom Community Clinic. During the FGD, the moderator carefully observed the potential YMSM who reported more sexual risk experiences but could not fully describe them in the FGD. YMSM were selected from FGDs for participation in KIIs based on their involvement with HIV risk activities and settings . Of 73 recruited participants, 21 were unavailable during scheduled FGD times, and five were unable to be reached prior to FGDs being conducted. In total, 47 young men participated in the FGDs, and 10 participants were selected from FGDs to share more in-depth experience in the KIIs. Eligible participants were asked to provide available dates and time. Study staff assigned them to a prescheduled focus group discussion upon their availability. The study staff did not stratify the groups by any particular characteristics to make sure a wide variety of YMSM of different backgrounds was represented in the group. --- Ethical issues As part of the consent procedure, participants were informed about the purposes of the focus group discussion and key informant interview, procedures, risks, and benefits, and their rights as a study participant. A study staff answered all questions asked. All participants were asked to sign the informed consent form prior to their participation in the study. Participants were given a copy of the informed consent form to keep. The study protocol was approved by the Institutional Review Board at the United States Centers for Disease Control and Prevention and the Thai Ministry of Public Health Ethical Review Committee for Research in Human Subjects. --- Setting We conducted FGDs and KIIs in a private meeting room at the Silom Community Clinic in central Bangkok on weekdays and weekends, between 16:00 and 22:00 hours. --- Data collection A short self-administered questionnaire was administered before the FGD to assess key sociodemographic characteristics and sexual and drug-use risk behaviors. We developed a FGD guide and KII guide to streamline and facilitate discussion and to ensure that all predetermined topic areas were addressed. A moderator, a note taker, and an interviewer were well trained and experienced in working with the YMSM and MSM community. --- Data analysis Professional typists transcribed audio files from FGDs and KIIs in Thai. The verbatim transcripts were coded and analyzed in the original language . The study staff analyzed the data by themes corresponding with research questions, and categories of findings were developed. Standard computer software for qualitative data analysis was used for coding Thai language scripts and categorization . Content analysis was used to identify key categories, such as certain risk behaviors, risk situations, and other factors that may contribute to high HIV incidence among YMSM. The findings with verbatim quotes in Thai were translated to English during manuscript development. --- Results --- Participant characteristics Among the 47 participants, the mean age was 22.5 years. Most participants were full-time employed , graduated bachelor's degree or higher , and self-identified as gay . All participants were sexually active, with the largest proportion engaging in both oral and anal sex . Many participants reported that they had had receptive anal sex only and had been to clubs, bars, or saunas in just one area in Bangkok . Among YMSM who participated in this study, 76.6% reported ever meeting a partner from the Internet, 29.8% having used club drugs, and 23.4% having had sex while high on drugs. A similar percentage had used Viagra or its generics. Participation in sex parties was reported by 27.7% of YMSM, high parties by 12.8%, and having sex in exchange for money or goods by 17.0%. Overall, the average number of risk behaviors a participant experienced is 2.19 . Regarding anal sex experience in the past six months, 91.3% of YMSM reported having had anal sex with a steady partner, and among these 53.5% reported always using a condom; 74.5% YMSM reported having had anal sex with a casual partner, and among those 54.3% reported always using a condom . --- Factors influencing HIV risk behavior We classified emerging categories in focus group discussions and key informant interviews related to factors influencing HIV risk behaviors among YMSM into six categories: the use of social networks for seeking sexual partners and the marketing promotions of MSM entertainment venues, social influence by peer and older MSM, easy access to high parties and group sex, easy access to club drugs, conceptions related to HIV risk, and sexual preferences of YMSM. --- The use of social networks for seeking sexual partners and the marketing promotions of MSM entertainment venues Social networking and MSM entertainment venues were perceived by most participants to encourage casual partner seeking. Participants described that several mobile applications were made for MSM to seek casual partners and were free to download. Participants explained how easy it is to browse photos and find the nearest partners with specific preferences and to identify a meeting spot through private chat rooms: Sexual partners from the Internet, especially Camfrog, are the type of people I actually have sexual intercourse with. For those partners from public places, I don't take them back to my room or go to their places. Only hand jobs or oral sex are enough. I paid $2.99 a month for a privilege to see more cute guys than ordinary members and more features added. I added a list of favorite guys to get notifications if they came close to my location. I didn't have to travel far to have sex with them. Some participants discussed that MSM entertainment venues such as saunas and clubs presented a sexually challenging and stimulating environment for YMSM. Participants expressed that they knew about marketing promotions for YMSM or special events through word of mouth, Web sites, Facebook, MSM magazines, and sauna posters. First of all, I call my friends asking where they are. They update the promotions that free alcohol and close at 4 a.m. I will follow my friends. Some people like to go to saunas so they go to saunas after the clubs close. Some MSM clubs also close at 9 a.m. I usually go to the saunas during weekdays because they have nice promotions. On weekends, I go to the clubs because they open until dawn. After the clubs, I go home because I cover with sweat and smell of alcohol. --- Social influence by peer and older MSM Participants expressed that peer pressure was the most influential factor accelerating risk behaviors, especially for those MSM who lived away from family. Most YMSM who participated in the discussions had relocated to Bangkok for further education and, therefore, did not live with family members or relatives. Instead, peer groups had become a main part of their social life. Participants stated that sexual orientation and sexual experiences with casual partners were revealed and often discussed among friends. Peer groups introduced them to MSM entertainment venues and social networks and supported their sexual identity disclosure. Living without family, participants have more freedom to express their sexuality, and they are more willing to disclose their sexual orientation among peers than among family members. In return, they gain connections and the ability to fit in with a new society. We found evidence that peers from university and older MSM from clubs were a main factor increasing risk behavior in YMSM in this study. Participants reported that in order to recruit YMSM into private sex events, older MSM searched online for attractive members to participate in group sex. Free drugs and alcohol were offered for guests. Sex occurred among guests and sometimes with the hosts' steady partners, while some older men often watched these activities for pleasure. Emotionally, both same-aged peers and older MSM provided a sense of belonging, security, and acceptance, as well as feelings of value and excitement. My friends took me to saunas for the first time. I never go there by myself even I have all my freedom because I live alone. I don't have a lot of free time but if my friends ask me to, I will definitely find time for them. Some participants mentioned that older MSM were one of the most powerful peer groups for YMSM due to Thai culture's respect for age and seniority. Older MSM were 30 years old or higher and had long-standing MSM connections and more money to influence YMSM who wanted to be protected and welcomed into MSM society. As mutual benefits between YMSM and older MSM, YMSM gained protection and financial support, while older MSM had the power to influence sexual risk by urging YMSM to have sexual intercourse with multiple sexual partners, consume alcohol, and use drugs. When I hang out with older MSM, I don't have to pay for anything because I'm a young one. From my own experience, I went out with older MSM and he put a pill in my mouth followed by a glass of water. I didn't know what it was and I swallowed it. It turned out to be ecstasy, I was high and I knew it is not a good thing to do but couldn't say no. He always takes care of me and supports me. Easy access to high parties and group sex "High parties" involve the consumption of recreational drugs to facilitate sexual activity. Participants reported that easy access to drugs was particularly true for high parties. Most high parties were several days long, and participants were solicited online, typically sponsored by older MSM who selected YMSM attendees based on profile photos or live video images. Some participants had experienced group sex, or swinging parties. They stated that the structure was similar to high parties, except high parties involved drug use and lasted over a few nights. Group sex could occur indoors and outdoors. There were various venues for young men to seek sexual pleasure, regardless of security concerns such as getting robbed, sexually abused, or raped by multiple sexual partners. YMSM gathered for adventurous casual sex acts in public places such as public toilets, parks, theaters, and fitness centers. Outdoor sex was freely advertised through Web sites that provided the meeting time and location for such activities. Participants explained that members were recruited from Web sites, social networks, and on site in sauna dark rooms. Sometimes, group sex involved up to 30 members and took place in settings such as bareback parties , sex-in-the-jungle parties , and full moon parties . Some participants reported using ice to increase sexual pleasure; however, it also had an effect on erectile dysfunction. This led to the use of erectile dysfunction medication together with club drugs during many sexual events. In high party, the main objective of the host is to have sex. Using drugs is not the main thing. For the guests, the objective is to get free ice . The host uses Ice to exchange with sex. Nothing comes for free. --- Easy access to club drugs According to Thai law, illicit drug use is prohibited. Illicit drug dealers and drug users can be sentenced to lifetime incarceration. However, access to club drugs has never been difficult for YMSM who seek them. Participants described finding club drugs through peers at university, casual sexual partners, older MSM at high parties, and through a government law officer at the club. Participants explained that the types of drugs varied and were used differently depending on the purpose of the event. Frequently, ice and Viagra were used together during high parties. Ecstasy and poppers were sometimes provided by party hosts. Participants reported that there were Internet posts for selling club drugs, but it was safer to buy from peer connections. Most participants reported having once used drugs for different purposes such as weight loss, hyperactivity, and sexual pleasure. I used ice when I wanted to have good sex with my boyfriends. It added so much pleasure and fun. Sometimes, I used it while I worked as a masseur… just to give pleasure when I had to have sex with clients. --- Conceptions related to HIV risk Participants discussed three categories of conceptions that often led to higher-risk behaviors and became barriers to effective HIV prevention: anal sex, HIV transmission, and HIV risk assessment. Anal sex.-Common conceptions that potentially reduced condom use included beliefs that anal sex did not transmit HIV because it was "not really sex." One conception of participants was that the definition of sex was a physical activity in which a man inserted his penis into the vagina of a woman. HIV risk level of anal intercourse was perceived to be low: "Everyone knows three ways of HIV transmission from health education class in school. But I don't get it how HIV can be transmitted through anus. It is totally different from vagina" . HIV transmission.-YMSM participants believed that the likelihood of HIV transmission in one episode of sex was low. Not using condoms once in a while could not cause HIV transmission because no one was that "unlucky." However, MSM who had multiple sexual partners at the same time were perceived to be at a greater risk than those with one sexual partner at a time, regardless of the frequency of changes in partners: "Most people think they are not that unlucky. This time may be fine " . Another participant conception about modes of HIV transmission was that HIV could not be transmitted unless ejaculation occurred. Participants believed that there was no need to use a condom as long as they could control ejaculation to be outside the anus: "It is risky if I ejaculate inside my partner's anus because HIV is in semen. The safe thing to do is I control ejaculation to be outside " . HIV risk assessment.-Participants believed that someone who looked "clean" or "young" could not be HIV-infected. Participants explained that young sexual partners represented naivety and innocence. They were desirable among older MSM and believed to be a "never been touched" type of sexual partner: "Because he is young, I think he is cleaner than I am" . Participants who had never personally known any HIV-infected persons believed that HIV infection could not happen among their circle of friends and that HIV commonly spread among other groups such as commercial sex workers and people who inject drugs: "AIDS is for people who don't know how to protect themselves, like hookers or MSM who sell sex at . We are upper-class educated MSM" . HIV testing.-A belief among YMSM was that it is better not to know HIV status. It only brought stress into their lives. Participants also believed that people their age should not go to the clinic regularly. Participants reported seeking clinical services for symptomatic STI treatment only: "I never wanted to know my HIV status. I would not come here if I didn't have to get herpes treatment. When you don't know about it, you don't have to think about it" . --- Sexual preferences of YMSM Condom use and sexual desire.-YMSM reported that they preferred not to use condoms, despite general knowledge of HIV risk. They did not like the feeling of having sex with a condom especially in an insertive anal role. Carelessness, laziness, overwhelming sexual desire, and curiosity were perceived to be causes of non-condom use: Mostly, craving for sex is more powerful than fear of HIV. They don't care about anything else at that moment, only focus on intercourse. The desire for sex is overwhelming. When I was having outdoor sex with many men, condoms were not enough and it would take time to go get more. I was lazy to waste my time on buying them. It is hard to carry a condom in sauna when you're in a small towel or underwear. Sometimes when I needed it, I already dropped it somewhere. I had to go downstairs to get a new one. Receptive anal role: A role of a giver.-Receptive anal role participants stated that they also performed oral sex with partners both before and after intercourse. They tried to please a partner by swallowing a sexual partner's semen after ejaculation. It was a gesture to show appreciation for manhood and represented affection for a man. This commonly accepted view was believed to cause difficulty in rejecting unsafe sex or negotiating condom use: "To be receptive role, you cannot choose to use a condom, nor choose a sexual partner who will use a condom" . --- Discussion Our study shows six categories in focus group discussions and key informant interviews related to HIV risk behaviors among YMSM: the use of social networks for seeking sexual partners and the marketing promotions of MSM entertainment venues, social influence by peers and older MSM, easy access to high parties and group sex, easy access to club drugs, conceptions related to HIV risk, and sexual preferences of YMSM. Our analysis indicates that nearly 75% of participants in our sample have had sexual intercourse with partners found via the Internet and mobile applications. Our findings were consistent with previous studies that found the Internet to be an effective tool for reaching numerous sexual partners, while anonymity and same-sex behavior were maintained . Men who seek sexual partners through the Internet were more likely to have multiple partners than those who did not use the Internet as a sex-seeking tool . Based on descriptions of sexual activities given by participants, MSM gathered for adventurous casual sex acts indoors and outdoors. Most YMSM fantasized about having group sex experiences with their boyfriends and others. MSM couples may have sexual agreements with mutually agreed-upon conditions involving group sex between friends, swinging parties, and group sex in saunas . We found that YMSM risk behaviors were associated with a combination of factors as also found in New York City . Findings from other studies also showed similar characteristics of YMSM who engaged in risk behavior and YMSM who were binge drinkers. Both groups were more likely to live without their families, frequent gay bars, have greater numbers of friends who engage in risk behaviors, and be sensation seekers . Conceptions about HIV risk behavior stemmed from discussions among peers. YMSM search for same-sex education from the Internet or learn how to react to situations by consulting with peers. The Internet contains inaccurate information that contributes to YMSM risk behavior. Sex education at school emphasizes heterosexual practices and neglects homosexual practices. Most of YMSM understood basic HIV risks such as sexual intercourse, mother-to-child transmission, and shared needles. However, YMSM underestimated how easily they could become infected with HIV from anal intercourse. Lack of MSM-specific HIV knowledge is worrisome, including a lack of knowledge about sexual sanitary practices such as anal douching for receptive anal role. --- HIV prevention methods for YMSM Given the high incidence of HIV among YMSM in Bangkok, prevention strategies aimed at this population are urgently needed. HIV and sex education that address the conceptions held by this group could help to increase condom use. Data from this study reveal the significant influence of the Internet on risk behaviors for YMSM in Bangkok. Therefore, prevention efforts may benefit from appropriate and targeted use of the Internet. This is consistent with other research that suggests that online and social network-based strategies for YMSM may be useful . In Los Angeles, YMSM suggested the use of technology, including the Internet and mobile devices, for HIV prevention strategies . Similarly, the Internet plays a significant role in HIV prevention messaging for YMSM 18-29 years of age in New York City . Participants in another study among YMSM in the United States also suggested use of technology including social networking and the Internet for HIV prevention programs. Men noted that, in the absence of formal sex education for YMSM, the Internet was a main source of sex education. They suggested that sex education including HIV prevention information relevant for YMSM at an early age would be beneficial . There is also evidence that Web-based HIV prevention strategies for MSM may be useful in Thailand, especially when MSM are involved in the design of prevention materials . Although the Internet can accelerate certain risk behaviors, it may also be a tool for successful HIV prevention among YMSM. This study provides evidence, consistent with other findings, that strategies using the Internet and social networking may be effective in Bangkok. In cooperation with the Thailand Ministry of Public Health, MSM entertainment venues openly support HIV prevention efforts and provide condoms and distribute information. Despite these attempts, venues such as saunas and clubs, which usually provide an irresistibly sexually stimulating environment for YMSM, remain a challenging setting for HIV prevention. For example, although condoms are provided onsite, it is not convenient to carry a condom into saunas. In order to provide convenience and promote condom use for MSM customers, a wristband with a small pocket for condoms could be helpful. Continued high HIV incidence and risk behaviors among YMSM showed the need for innovation and increased efforts to prevent HIV infection in this population . After three decades of HIV, it is obvious that condom use and abstinence alone are not enough to stop this epidemic. Additional effective ways to protect MSM from HIV infection are urgently needed . One clinical trial and its open label extension study have shown evidence that pre-exposure prophylaxis reduces the risk for HIV infection among MSM . However, acceptability and feasibility of PrEP use in YMSM should be further explored in order to find the most effective way for their sexual practices . Prevention approaches that combine behavioral, biomedical, and structural interventions also should be considered and tailored for vulnerable and high-risk populations globally . --- Limitations The topics discussed in this study are sensitive and personal. Some of the participants may have felt uneasy answering questions openly in front of other participants. This study showed a fair amount of risk behaviors and drug use, but it was possibly an understatement of actual risk. The small sample size might affect generalizability; nevertheless, we collected in-depth useful insights from a sufficient number of YMSM to reach data saturation. Participants from FGDs and KIIs may have been relatively well educated in terms of HIV risk as all participants were members of the Bangkok MSM Cohort Study and/or had been attending the Silom Community Clinic for VCT for 2-5 years. They may have had more knowledge about HIV/AIDS and fewer risk behaviors than MSM in the wider community. --- Conclusion Our qualitative data show that social networks and MSM entertainment promotions, peer groups and older MSM, high parties and group sex, drug use, conceptions related to HIV risk, and preferences of YMSM are influential factors of HIV risk behaviors among YMSM. While most YMSM had basic HIV knowledge, several conceptions were common that likely reduced condom use and increased HIV risk, particularly in the presence of events with multiple sexual partners and drug use. Evidence from our study suggests that YMSM-specific HIV prevention efforts are needed to reduce HIV and STI infection in this population. YMSM Voluntary Counseling and Testing and peer-group sex education related to same-sex practices are urgently needed for YMSM, with emphasis on consistent condom use, particularly in tandem with drug use and multiple partner situations. --- Table 1. Sociodemographic characteristics of young men who have sex with men who participated in qualitative component of Bangkok MSM Cohort Study, Bangkok, Thailand, 2012-2013 .

The Bangkok Men Who Have Sex With Men (MSM) Cohort Study has shown high HIV incidence (8-12/100 person-years) among 18-21-year-old MSM. These data led to a further study using qualitative methods among young (18-24 years old) MSM in order to understand the factors driving the HIV epidemic among YMSM. We conducted eight focus group discussions and 10 key informant interviews among YMSM in Bangkok, Thailand. Sociodemographic and behavioral data were collected using a questionnaire. We audio-recorded, transcribed, and analyzed qualitative and questionnaire data using computer software. The categories relating to risk behavior were (1) the use of social networks for seeking sexual partners and the marketing promotions of MSM entertainment venues, (2) social influence by peers and older MSM, (3) easy access to high parties and group sex, (4) easy access to club drugs, (5) conceptions related to HIV risk, and (6) sexual preferences of YMSM. Increased HIV testing, same-sex education, and YMSM-specific HIV prevention efforts are urgently needed for YMSM in Bangkok.

Background Social Network Analysis is a research approach consisting of a set of theories and methods for mapping communication, information flow and relationships between individuals or groups. SNA has been applied in a wide range of social and physical science fields [1][2][3]. While some of the seminal SNA articles are health related [4], and there is a growing body of research on health professional advice networks [5][6][7], little is known about networks among frontline health workers in low and middle-income countries [8][9][10]. Over the last 10 years, Ethiopia has expanded its health work force considerably, through the introduction of a community-based cadre, health extension workers . Yet skilled human resources for health remain a constraint in the equitable availability, accessibility and delivery of healthcare [11]. Ethiopia's impressive gains in achieving child health targets, with respect to the Millennium Development Goals, have been attributed in part to the Health Extension Programme [12,13]. Recent data suggest maternal and newborn mortality rates are starting to improve after a period of stagnation. Although the 2016 Demographic Health Survey indicated progress with a promising drop to 29 from 37 neonatal deaths per 1000 live births, deaths in the first month of life still account for a large proportion of overall child mortality [14]. Similarly, maternal mortality stagnated with no improvement between the 2005 and 2011 DHS reports [15]. The most recent DHS shows a decline to 442 from 676 deaths per 100,000 live births in 2016 [14]. These gains in health outcomes are encouraging and may be attributable to the health systems changes described above. Nevertheless, maternal and neonatal mortality rates are unacceptably high and well away from the Sustainable Development Goals targets for 2030. Ethiopia's previous maternal and newborn health targets were relative to their levels in 1990; the new SDGs set absolute targets for all countries to meet by 2030 [16,17]. To achieve these SDGs, Ethiopia must reduce neonatal deaths from 29 to 12 per 1000 live births and maternal deaths from 442 to 70 per 100,000 live births in an environment of increasingly constrained health resources [14,17,18]. Communication among healthcare workers could be a critical contextual factor affecting whether or not mothers and newborns receive life-saving interventions. Understanding professional advice networks provides foundational information for designing network-based interventions to improve health outcomes [19]. This study aimed to contribute to the understanding of professional advice networks of frontline health workers in Ethiopia. Specifically, this research explored the properties of professional advice networks; the content of advice exchange, the context in which advice exchange took place; who participated in advice exchange; and the extent to which advice networks met healthcare worker needs. Our study focused on healthcare workers implementing a government program known as Community Based Newborn Care , designed to address gaps in maternal and newborn care. This subset of healthcare workers were chosen as they were implementing a new health program, thus creating the possibility of advice exchange related to these services. --- Ethiopian health delivery context Ethiopia has a decentralised, structured model for the delivery of healthcare with the Federal Ministry of Health as the policy-setting body. The four tieredsystem provides services at specialised, zonal and district hospitals and primary healthcare units [11]. PHCUs serve approximately 25,000 people through a referral health centre and five health posts. In 2003, Ethiopia introduced the Health Extension Program , an ambitious plan to improve delivery of primary healthcare through the introduction of over 30,000 Health Extension Workers [11]. HEWs are government salaried women, 18 years or older, who completed 10th grade schooling [20]. They work in their kebele, the Ethiopian equivalent of a village, in pairs at a health post after receiving a one-year training to deliver a package of 17 essential preventative and curative health services [11,21]. HEWs have a formal reporting system whereby within a PHCU all HEWs report to a single designated staff member at the health centre. HEWs are supported by a network of volunteers. Studies of HEWs have found their full potential unrealised, calling for additional training and resources for them to maximise their impact on maternal and newborn health [11,[21][22][23][24]. In 2013, the Ethiopian government launched CBNC to build on the HEP platform [25]. CBNC aims to improve newborn health outcomes using a framework of "four Cs" and 9 components, see Additional file 1 [26]. CBNC depends on HEWs working with Woman Health Development Teams to identify and refer sick newborns, and when referral is not possible, treat in the community. Health systems are social systems, in Ethiopia linking the community to health services relies on the professional relationships of HEWs [27]. --- Social network analysis A social network is a set of socially relevant 'nodes' representing actors -either individuals or organisationsconnected by one or more relations [28]. In SNA, the patterns in the relationships between actors are studied. This is distinct from much quantitative public health research, which looks at the relationship between variables and outcomes of interest [29,30]. While there is a longstanding history of social network analysis in health, [3] recent reviews suggest that the potential has yet to be realised [10,31,32]. Much of the work to date focused on spread of disease, diffusion of ideas, impact of social networks on individual health behaviour and interorganisational structure of health systems [33]. There was little known about professional communication among health professionals [33]. Research on healthcare worker advice networks has mostly been descriptive and not related to the provision of primary healthcare [7,[34][35][36][37]. Applications of SNA methods in the community-based health contexts of low-and middle-income countries are similarly limited [38]. To date there have been few social network analysis studies in Ethiopia [39][40][41][42][43] and none in primary healthcare settings. Relational data can be collected through questionnaires, interviews, observations and analysis of existing records, diaries or other methods [44]. These data are then populated into matrices, or tables and uploaded into software designed to generate visualisations, knowns as sociograms, and calculate network properties. In SNA, visualising data is both a means of presenting findings as well as a tool for identifying patterns and generating findings. The most basic sociograms depict actors as points with connecting lines representing relationships. It is possible to overlay additional information to qualify the actors or their relationships. These "attributes" can be displayed by changing the colour, size or shape of the actor. The strength of the relationship can be represented by adjusting the thickness of lines connecting actors [45]. SNA studies are characterised by whether the network is directed or undirected, valued or unvalued, and by how the boundaries are defined [44]. Directed networks capture if the relationship is one-way, where one actor initiates and another receives or two-way where both initiate and receive. In contrast, in undirected networks, the relationship either exists or does not exist. Valued networks quantify the strength of the relationship between actors [46]. Network boundaries can be specified either through what is known as a 'realist approach' whereby study participants define their own network boundaries or through a 'nominalist approach', which uses formal criteria to determine the network or a hybrid combining the two [47]. The overall objective of this study was to contribute to our understanding of professional advice networks of PHCU staff in rural Ethiopia. --- Methods --- Aims This research aimed to describe the properties of professional advice networks, the content of advice exchange, the context in which advice exchange took place and the extent to which the existing advice networks met healthcare worker needs. --- Study design This was an observational, cross-sectional network study. There were two stages of data collection: a structured network survey, followed by semistructured qualitative interviews. The structured network survey captured valued, directed networks using a roster of PHCU staff from our selected PHCUs, but allowed for respondents to nominate, or name other "off-roster" healthcare professionals as either having provided or sought advice for each of the advice networks of interest. It was directed and valued, with data such as who provided or sought advice from whom, both within the PHCU and beyond it, and the frequency of their interaction were documented. --- Sampling PHCUs in zones implementing CBNC formed the sampling pool. Two PHCUs per agrarian region were purposively selected for diversity considering number of health posts, number of healthcare workers and coverage of key maternal and newborn health services. Semi-structured qualitative interviews were conducted with 20 network study participants to explore and gain an in-depth understanding of patterns emerging from the quantitative analysis. Participants were purposively selected to capture the range of PHCU network properties (densest networks, individuals with highest eigenvector centrality, least off-roster advice exchange, least ties and greatest distance, breadth of cadres and individual network characteristics. --- Tool design Preliminary interviews to inform study design took place in 2013, a network tool was pre-tested in April 2015 and again in August 2015, translated and back translated, revised and field tested in November 2015. It captured the frequency of seeking and giving advice among PHCU staff using a roster. The following respondent characteristics, also known as attribute data were also captured: gender, age, cadre and total years of experience. The semi-structured interview guide was initially developed in August 2015, revised following analysis of quantitative data and further refined after the initial qualitative interviews. Themes explored included who is sought for advice and why, reasons for advice exchange, if advice needs are being met, and barriers to advice exchange. --- Data collection Data were collected by two research teams after a threeday training. The first PHCU's data were collected together by the two research teams with the lead researcher overseeing and managing the daily review sessions. Research teams consisted of two interviewers and one supervisor who were native speakers of the languages involved . Data were collected over a period of 3 weeks in November and December 2015. Semi-structured interviews were conducted in June and August 2016 by two teams of trained qualitative researchers. Teams consisted of an interviewer and an interpreter. Interviews were conducted in private spaces in a mix of English, Amharic, Oromifya or Tigrinia. Interviews were sound recorded and detailed notes were expanded immediately following interviews. --- Data management Network survey data were collected using paper forms. All rosters and 160 network questionnaires were double entered into Excel [48]. Discrepancies were reviewed and reconciled. Data were imported into UCINET 6.0 for analysis [49]. Each PHCU had seven network matrices and one table of attribute data. Qualitative data included 20 sets of expanded field notes written in English. Sound recordings were used to spot check the translation and expanded field notes. These were imported into MAXQDA 10.0 for analysis. --- Data analysis For each PHCU, a valued adjacency matrix was prepared in Excel for each of the seven networks. Additional networks were created by collapsing data captured into themes: all ties, all advice seeking and all advice giving. Data were imported into UCINET and dichotomised for the calculation of network and actor-level metrics and then imported into Netdraw for visualisations. See Additional file 2 for definitions of network metrics and their calculations [50]. Final figures were regenerated in InDesign. Qualitative data were coded based on inductive and deductive approaches, building from grounded theory, but more applied, focusing less on theory development [51,52]. An initial coding tree, with code definitions based on review of literature, understanding of the subject area and initial readings of the expanded field notes was applied to a subset of interviews, reviewed, revised and then reapplied. Coded text fragments were reviewed by segment and intersections of codes were reviewed. Codes were then grouped by respondent cadre to analyse whether patterns emerged by cadre. --- Results --- Network survey The eight PHCUs each consisted of one health centre and an average of 4.38 Health posts . Approximately two thirds of the 160 participants were female, with health officers disproportionately male , and midwives disproportionately female . The average number of years of experience was 3.6, with 2.5 years at their current post. HEWs on average had the longest total experience and most years at their current post. Overall, 46% were trained or orientated in CBNC programme. All HEWs should be trained in CBNC and from this sample of 160, 78% reported being trained. --- Network metrics Across all networks and all PHCUs the following average network level metrics were observed: density .26 , degree centrality .45 , distance 1.94 , number of ties 95.63 , size of network 20.25 Table 3 presents the network level statistics by network metric, allowing easy comparison of each network metric across networks and PHCUs. Additional file 3 presents the same data, grouped by PHCU providing an overview of each PHCU by network type. Patterns emerge with typically ANC and maternity advice networks being denser and having more ties than PNC and Newborn care advice networks. Half of the PHCUs had this same pattern across the networks with degree centrality. There were some exceptions: for PHCU E and F, ANC advice exchange was noticeably denser, however maternity, PNC and newborn care advice exchange networks had similar density. The other clear pattern was that certain PHCUs had more advice exchanging than others . --- Actor-level network metrics In addition to calculating network level properties, actor-level metrics were calculated and the cadre of the actor with the highest value is reported in Additional file 4 for each of the networks. Midwives were far more likely to be the actor with the highest in-degree centrality, meaning the most people within the PHCU came to them for advice. This is highlighted in Additional file 5 and is true for all subject areas, although they were as equally sought as nurses for advice related to providing newborn care. --- Visualisations Each PHCU had their networks visualised both with dichotomised data, which facilitated aggregating the ties across networks, and valued data, which added a layer of understanding related to the frequency of interactions. To illustrate the variability within a PHCU across these networks, PHCU H was selected. Figure 1 visualises four dichotomised networks with advice seeking and giving for each care area aggregated into one. These sociograms show more ties and fewer isolates for ANC and maternity advice networks relative to PNC and newborn care advice networks. Valued data were visualised in the same way as for the other sociograms with the exception that the line widths reflect the frequency of interaction . PHCU A's PNC advice seeking and advice giving sociograms were selected to illustrate in Fig. 2 how there seems to be more individuals seeking advice than giving advice. To show the variability across the PHCUs for a given type of network, maternity advice seeking was selected to show across all 8 PHCUs in Figs. 3 and4. For all of these graphs midwives play a central role, as expected, despite some variability. In Fig. 3, PHCU D has only advice exchange happening at the HC with the exception of one HEW engaged, whereas in Fig. 4, PHCU F has many HEWs engaged frequently with HC staff and even some HEWs seeking advice from each other. The intra and intercadre advice exchange depicted by PHCU F was more typical of the findings across PHCUs and topics for advice exchange. The data on formal supervisory structures were available only from a subset of those PHCUs that were selected for the qualitative inquiry. Due to staff turnover, only one could be analysed along with the quantitative network data. This example confirmed what was observed across other PHCUs, a willingness to engage in informal advice exchange outside of formal supervisory structures. --- Off roster advice seeking Of the four cadres of healthcare workers, health officers reported the fewest number of individuals they either sought or gave advice to who were not working within their PHCU. However, after adjusting for the different number of HCWs per cadre those distinctions largely disappear. For these HCWs more advice is sought off roster than they are giving to those outside of the PHCU. This is particularly the case for Health officers and HEWs. By far the most off roster advice exchange occurred for nurses and HEWs seeking advice regarding providing ANC. --- Qualitative findings Who is sought for advice and why? Reasons for going to a specific person for advice were typically because of that specific person's training and knowledge, less so because of their years of experience. One healthcare worker noted that because of their training in integrated community case management, PHCU colleagues seek their guidance. Some respondents said that they relied on formal supervisory structures, however they appeared relatively infrequent with most people describing qualities of the individuals' knowledge and skills dictating their advice seeking behaviour rather than just formal structures. The examples of consulting supervisors related to situations where they "had some fear or discomfort with the situation and didn't want to take accountability for something going wrong." [Health officer, Tigray] Personality or level of comfort with the person was mentioned as a secondary factor that contributed to who was approached for advice. In general, current PHCU colleagues were sought for advice, however if they were not available, former classmates and colleagues were the most common individuals sought for advice. For example, one respondent said he'd first go to experienced people in his PHCU, but if they are unable to give advice he would: "call some peers, people who [I] went to school with and grew up with, working in other HCs or hospitals who have several years of experience, or even professors to ask for advice in complicated cases." [Health Officer, Amhara] This was consistent across cadres: health officers, nurses and HEWs for both routine and urgent questions. A nurse in Tigray described a case of postpartum haemorrhage when his supervisor was away at a training, so he had called a midwife he had previously worked with who was now at a different health centre. Several people within this PHCU mentioned seeking advice from this same midwife who had been transferred. The furthest afield anyone mentioned seeking advice from was from a friend in Addis Ababa because colleagues within the PHCU did not know how to handle the situation. Only one person mentioned seeking advice from someone outside of the PHCU because of not being comfortable asking for a colleague's advice. This does not appear to be a widespread concern for most healthcare workers. --- Reasons for advice exchange According to our respondents, the range of advice given on providing ANC care included many topics already Fig. 2 Primary Health Care Unit "A" Valued Postnatal Care Advice Seeking and Giving covered in their training. One nurse explained that they need repetition because learning the content theoretically is so different from doing it practically. This was also the case for advice exchange around other service delivery areas. "While the integrated community case management manual is very clear it seems that [the HEW I was advising] lacked confidence and contacted me at the health centre for reassurance." [Nurse, Tigray] While there are examples of advice being sought for providing antenatal care, it is noteworthy that many respondents said they felt comfortable providing ANC and believed they did not need advice. Several healthcare workers mentioned fearing deliveries and that those with less experience sought those with more for reassurance and guidance. A HEW described seeking advice from another HEW because she had referred more women to the health centre for delivery and she wondered what methods she was using that might be helpful in her own work. --- Are advice needs being met? All interviewees indicated that they had always been able to have their specific questions answered when seeking advice. Interviewers probed further, asking if there were No interviewees said they were unable to get the advice they needed, although some described asking more than one person or consulting other resources. This could reflect a response bias, an unwillingness to admit to providing care while having questions about providing that care. Or it could be that when they had a clear question they could generally find an answer. However, they only asked when they were aware that they did not know something. While their specific questions were addressed, many described a desire for additional training, because as one respondent said, "I will benefit from additional information I am not aware of" [Midwife, Oromia] --- Barriers to advice exchange Many of the HEWs said that they do not deliver babies, although as the closest HCWs to the community they are often involved in referral to the health centre. Some HEWs said they are not involved in postnatal care, although others said they are involved just for identifying dangers signs and referral. 1 When asked why people do not seek advice from her, a HEW said "the HC staff don't ask because they are at a higher level of knowledge, education and training than me and that they would ask each other. They wouldn't think to ask a HEW." [HEW, Amhara]. Respondents commented on the logistical constraints in seeking advice. The poor mobile phone network in some rural areas was mentioned as a barrier, particularly in one PHCU's catchment area, which had only gained network access within the last 8 months. Respondents commented that workload sometimes interfered with seeking advice for non-urgent cases. Advice exchange for these non-urgent situations typically happened in person in either at ad-hoc or routine meetings. If urgent, [from respondent's perspective] mobile phones were used, particularly for HCWs seeking advice from more skilled providers at the health centre or woreda. "[I] usually ask for advice by phone, especially in the case of emergencies and this advice seeking comes whenever a difficult case arises, once or twice a month. There is nothing stopping [me] from asking for advice as long as the phone networks are working. The network rarely fails around the HC so this is not a big hurdle "[Health Officer, Tigray]. One HEW described her fellow HEW as being "intimidated easily to ask questions, so [I] served as a conduit." [HEW, Amhara] Another said that "If I do not know the answer, I would call someone who is not here, does not work in this place." [Midwife, Oromia] These were the only examples given even with probing with hypothetical reasons for why someone might not feel comfortable seeking advice from their colleagues within the PHCU. Another mentioned language as a possible barrier and that they "could access advice more easily" if rather than speaking in Amharic, they spoke in a local language as "this would avoid missing out on any information." --- Discussion Professional advice networks present an opportunity to more effectively change health provider practice than mere training as healthcare workers are more likely to be convinced by trusted colleagues. This is particularly relevant in resource constrained settings where funding for and feasibility of training everyone does not exist. This study explores existing advice networks in the context of a new program introduction to learn about these networks and see if there may be potential for harnessing them for future program needs. There are no standards for appropriate frequency of advice exchange among healthcare workers in Ethiopia or elsewhere. Presumably most healthcare workers should have some advice needs or if not, should be the source of advice for their colleagues, particularly in the context of a new programme being introduced for which not all have been trained. Further complicating interpreting these data is that network scientists debate what network properties constitute a "healthy network" [5]. There were no previous studies of PHCU healthcare worker professional advice networks documenting network properties. Neither were there previous studies of the same healthcare professionals comparing their professional advice exchange networks for antenatal, maternity, postnatal or newborn care provision. Therefore, this study contributes a foundation which future studies can use to compare their findings. The advice networks observed had few isolates with limited distance between actors. Five PHCUs had low density and high centrality. Taken together these network properties can be interpreted that most providers are participating in advice exchange and that typically there are actors that serve as hubs. There is diversity across the PHCUs in terms of network properties, with variability across all metrics. While some PHCUs fit a pattern whereby there are more ties for antenatal and maternity advice networks, other network properties and other PHCUs were more nuanced. Of note is the willingness of healthcare workers to seek advice and the lack of rigid adherence to the supervisory structure. Given the Ethiopian healthcare delivery context is very hierarchical the existence of informal advice networks is a noteworthy finding. Similarly, a study of Dutch nurses in a long-term care facility found advice networks to be non-hierarchical, although they only looked within one cadre [53]. In studies of interprofessional communication hierarchy is common [7,35,[54][55][56][57]. That said, the reliance on informal advice networks may be a product of infrequent supportive supervision visits and review meetings. Advice exchange varied by cadre, but universally there was more advice exchanging between cadres than within cadres. This runs counter to what has been found in some western contexts, where professional homophily runs high and advice is primarily sought within a cadre [7,35,58]. The direction was typically from the cadre with less training to that with more, with midwives being more engaged for maternity care advice and health officers more for newborn care advice needs. It is important to note that there is specialisation and division of labour within the health centre and between different cadres which could explain some of the patterns observed in the advice seeking. While HEWs say they are not sought by other cadres for advice because of knowledge differentials, the subtext could very well be that a HEW's advice is not sought by other cadres because of power and hierarchy. Advice exchanged between cadres often took place over mobile phones if it was urgent. If it was not urgent, they would wait for supervisory visits or routine meetings. This preference for informal, in person communication is consistent with findings among other, albeit western healthcare contexts [58,59]. It is a significant finding that the advice networks are largely meeting the advice needs of healthcare workers. Barriers to advice exchange included poor phone networks and knowing when to ask for advice. Advice needs are being met only in those situations where healthcare workers know what they do not know and seek advice. While a preference was expressed for in person communication, evaluations of the HEP note that routine PHCU meetings and supportive supervision fall short of standards [24]. The lack of regular supervision may also account for why these supervisors were not featured more prominently in these advice networks. Professional advice networks of PHCU staff prioritise proximity, but are not restricted to it. This is relevant both within the PHCU and to those working outside the PHCU. Another study in Italy found geographic distance a factor in physician advice networks [60]. The individuals who were not PHCU staff sought for advice ranged from peers from training programs, former colleagues who have since been transferred to those working in nearby PHCUs or hospitals. Typically, they were engaged only if colleagues within the PHCU were unavailable or in one case could not answer the question. This suggests that the personal connection may matter less as the established relationships are not prioritized as a first stop for advice. This study shows the value of combining quantitative network methods with qualitative inquiry. This is an approach that more network studies may consider applying should their research objectives include understanding the context surrounding network ties. Such complementary methods could strengthen proposed role of SNA in program implementation [5]. There is a need for further research to understand why there is more advice exchanging around providing antenatal and childbirth care than providing postnatal and newborn care. The aim is to be able to answer the following questions: is this pattern a reflection of true information needs, a product of more patients engaging with the health system for those services, or a reflection of knowledge gaps in providing postnatal and newborn care sufficient enough for providers not to be able to identify their knowledge limitations. Answering these questions will help identify the appropriate policy response. This could be additional pre-service training on postnatal and newborn care such that once in the field healthcare workers are primed to "know what they don't know." This study's findings, while foundational, could have relevant policy implications for the Ethiopian Federal Ministry of Health and other LMIC contexts. Language was mentioned as a barrier in some advice exchange and some providers described relying on guidelines and reference materials, if these materials could be available in local languages this would be an easy-to-implement "quick win". These findings suggest that healthcare workers value training and knowledge over years of experience when seeking advice. Therefore it may be possible to achieve better outcomes through focusing training on specific individuals within a PHCU whose primary role could be sharing knowledge. Additional research would be needed to test such a concept and see if this targeted training approach yields at least equivalent learning across PHCU staff and patient outcomes. Additionally, these findings point towards the potential of cadre-based targeted in-service trainings with more central figures within informal advice networks. Further studies, building on what has been done elsewhere exploring the value of network-based training models [61] would be needed to pilot both the feasibility and measure the impact of such an approach to ensure the intervention is scalable and that equivalent outcomes are achieved. Strengthening supervisory structures may enable them to be more commonly used fora for advice exchange. Health centre staff were noticeably transient relative to HEWs, this may affect advice networks. Further research is needed to understand the directionality of this affect and the policy implications. --- Limitations Approximately 8 months passed between the collection of the quantitative data and the collection of the qualitative data. Recall bias may have compromised data quality, but this should be minimal given the study asked about hypothetical situations and the most recent example of advice exchange. The qualitative data are not explicitly linked to the quantitative network data as the time periods of reflection are inherently different and staff changes were noted. While there were anticipated advantages to analysing the quantitative network data and using that data to select participants for the qualitative study, future mixed methods network studies may consider conducting the qualitative interviews in parallel with the network surveys. While effort was made to limit misunderstandings around the type of communication of interest through use of a scripted description to clarify and standardise meaning of "advice seeking and advice giving", it remains possible that study respondents did not have a common understanding. --- Conclusions This exploratory study provides foundational information regarding professional advice networks of PHCU healthcare workers in Ethiopia. This study establishes that PHCU staff involved in delivery of maternal and newborn health services have informal advice networks outside of supervisory structures. Advice exchanges occurred between cadres and used both face to face meetings and mobile phones to exchange advice. More research is needed to understand if the patterns in advice exchange across antenatal, maternity, postnatal and newborn care accurately reflected advice needs or if they reflected a bias towards antenatal and maternity care knowledge and thus individuals being better placed to self-identify knowledge gaps in those areas. Fellow PHCU staff were prioritised but networks were not limited to those within their geographic area. Policy implications include focusing future training on cadres more central in advice networks, such as midwives for antenatal, maternity and postnatal care and nurses or health officers for newborn care. One possibility could be training an individual or two per PHCU to be the "knowledge sharing focal persons", who attend trainings and are responsible for sharing learnings. Another could be cadrebased in service trainings with the same mandate for sharing learnings. Further studies would be needed to pilot such approaches to ensure achievement of equivalent learning and patient outcomes. A simple policy implication of this work could be providing guidelines and reference material in local languages. Additional research is needed to more accurately measure performance to link network properties to patient outcomes as well as investigate the impact of turnover and absence on advice networks, ideally through a longitudinal network study. This study demonstrates the feasibility of using social network analysis methods in rural Ethiopia, which has implications for other African and low or middle-income countries. This study also shows the value of combining quantitative network methods with qualitative research to lend a greater understanding of network properties. Mixed SNA method studies should be used more widely in these contexts as they provide a different lens and understanding of professional advice networks in settings where resources for health are increasingly constrained and as such networks may be an efficient and effective way to change practice. manuscript. NS Guided conceptualisation of research questions, reviewed protocol, study tools, advised on qualitative data methods and analysis, and reviewed drafts of the manuscript. All authors reviewed the final version of the review. --- --- Supplementary information Supplementary information accompanies this paper at https://doi.org/10. 1186/s12913-020-05367-3. Authors' contributions KS Wrote the study protocol, managed ethical clearance, engaged the ministry of health for input, conducted pre-testing, trained enumerators, entered and analysed data and wrote all drafts of the manuscript. JS Guided conceptualisation of research questions, reviewed protocol, study tools, advised on data collection procedures and drafts of the manuscript. KB Guided conceptualisation of research questions, reviewed protocol, study tools, advised social network analysis methods and analysis, and drafts of the manuscript. DB Guided conceptualisation of research questions, reviewed protocol, along with KS engaged ministry of health for input, reviewed study tools, advised on data collection processes and procedures, provided input into selection criteria and use of coverage data, and reviewed drafts of the Consent for publication Not applicable. --- Competing interests None. ---

Background: In an era of increasingly competitive funding, governments and donors will be looking for creative ways to extend and maximise resources. One such means can include building upon professional advice networks to more efficiently introduce, scale up, or change programmes and healthcare provider practices. This crosssectional, mixed-methods, observational study compared professional advice networks of healthcare workers in eight primary healthcare units across four regions of Ethiopia. Primary healthcare units include a health centre and typically five satellite health posts. Methods: One hundred sixty staff at eight primary healthcare units were interviewed using a structured tool. Quantitative data captured the frequency of healthcare worker advice seeking and giving on providing antenatal, childbirth, postnatal and newborn care. Network and actor-level metrics were calculated including density (ratio of ties between actors to all possible ties), centrality (number of ties incident to an actor), distance (average number of steps between actors) and size (number of actors within the network). Following quantitative network analyses, 20 qualitative interviews were conducted with network study participants from four primary healthcare units. Qualitative interviews aimed to interpret and explain network properties observed. Data were entered, analysed or visualised using Excel 6.0, UCINET 6.0, Netdraw, Adobe InDesign and MaxQDA10 software packages.The following average network level metrics were observed: density .26 (SD.11), degree centrality .45 (SD.08), distance 1.94 (SD.26), number of ties 95.63 (SD 35.46), size of network 20.25 (SD 3.65). Advice networks for antenatal or maternity care were more utilised than advice networks for post-natal or newborn care. Advice networks were typically limited to primary healthcare unit staff, but not necessarily to supervisors. In seeking advice, a colleague's level of training and knowledge were valued over experience. Advice exchange primarily took place in person or over the phone rather than over email or online fora. There were few barriers to seeking advice.

Introduction Islam has teachings about human equality, so the basic concept of Islam that must be interpreted together is that God created men and women to be leaders. positive law that men are responsible for their households. And related to the leadership of women in the household where on the other hand there is a husband who should carry out his role, it needs to be studied further from various aspects of the law. The legal aspect cannot be separated from the life of a Muslim is Islamic family law, this is because Islamic family law is the main gate in entering other provisions of Islamic shari'a law. 2 One of the concepts often discussed in Islamic family law is Qiwamah, which comes after marriage between a man and a woman, in which the man is meant to be the head of the family who is most responsible for family affairs. 3 After the marriage takes place, the smallest group in society is formed, namely the family, consisting of husband, wife and children or at least husband and wife only. 4 Serves as a place to realize a safe, peaceful and prosperous life in the nuances of love and affection between family members. 5 In the 2 Lilis Hidayati Yuli Astutik, Muhammad Ngizzul Muttaqin, "Positivity of Family Law in the Muslim World Through Family Law Update", Islamika: Journal of Sciences Islam, Vol. 20, no. 01, July 2020, hlm. 56. 3 Dedi Masri. "Priority Man as Leader." ANSIRU PAI: Development Islamic Religious Education Teacher Profession Vol. 5, no. 2 of 2021, hlm. 156-167. 4 Sofyan Basir, "Building Sakinah Family", Guidance Journal Islamic Counseling, Vol. 6, no. 2, December 2019, hlm. 100. 5 Mufidah, Psychology Islamic Family with a Gender Insight, Cet. Ke-3, hlm. 33. family, a qualified leader is indispensable, because the well-being of the family depends on how a family leader performs his functions and duties well. 6 In a society that still adheres to the patriarchal system as if it has become a general agreement that places men as a leader or head of the family, therefore it will be considered strange and unfamiliar if the roles and functions of both are reversed. However, in practice, arguments and thoughts about male superiority are faced with shifting values, conditions and times. Because there are many cases that the dominant and superior in the family and place it as the party who holds a key role and control in the household is the woman. 7 Whereas as a leader in the family it has a very important role in directing, guiding and educating family members. Therefore, the application of the concept of absolute Qiwamah to men in this case as husbands raises controversy and debate about its implications for gender 6 Siti Halimah Putung, Raihanah Azahari, "Husband's Leadership in an Islamic Perspective: Function in Strengthening Institution Family", Sharia Journal, Vol. 28, no. 2 Years 2020, hlm. 128-129. 7 Nurliya Ni'matul The third, Wardah Nuroniyah in her journal entitled "The Concept of Qiwamah and the Phenomenon of Women Heads of Families", stated that the understanding of the concept of Qiwamah was quite good, but some of her informants required only the male gender as the head of the family. The concept of Qiwamah is actually contextual, because the Qur'an adapted when it was revealed to the situation of women in the Arabian Peninsula 14 centuries ago, then over time the adaptation process will continue if the Qur'an is seen in terms of its universality. 11 10 Tatik Hidayati, "Representation Social and Autonomy of Women Salt Workers ," Anil Islam: Journal of Culture and Science Islam. Vol. 12, no. 2 December 2019, hlm. 161-193. 11 Wardah Nuroniyah, "Concept Qiwamah and the Female Head Phenomenon Family", Equalita, Vol. 4, no. 1, June 2022, hlm. 114-135. So, it is clearly very necessary to conduct further research related to the concept of Qiwamah. Therefore, the focus of this research is the concept of Qiwamah and its implications for gender justice in Islamic family law in Indonesia. This study aims to examine more deeply related to women's leadership in the household so that we can find out how the concept of Qiwamah and gender justice in Islamic family law against the concept of Qiwamah in its implications for gender justice. The previous research has only used the concept of leadership in the household but this study will examine more deeply about the concept of Qiwamah in the household and its implications for family law in Indonesia. --- Method This research uses a type of library research. The meaning of library research is that research activities are carried out by collecting data and information with the help of various sources in libraries or other places such as reference books, previous research results that have similarities, notes, articles and various journals that have links to the problem that the researcher wants to solve. Research activities are carried out systematically to collect, process, and conclude the results of the data and information using certain methods / ways to find answers to the problems studied. 12 The approach method that the author uses is normative approach ; and sociological juridical approach (a legal perspective based on reality and social reality that exists in the community, namely regarding The Concept of Qiwamah and Its Implications for Gender Justice In Islamic Family Law In Indonesia. 13 In data processing, the author uses several methods, namely: 14 Data Reduction, in this case the author selects and concentrates attention on simplification so that it is easy to classify similar data by carrying out several activity steps such as selecting data based on the level of relevance and relation to data groups, grouping similar data and coding data in accordance with the research work grid. Data display, in this case the author develops and describes data that has been classified according to the subject matter and compiles relevant data so that it becomes information that can be concluded and has a certain meaning. And in data verification, after the data processing is In relation to the family, the family itself is the smallest unit or institution in society. A family will be able to run well if each performs its functions well, so in the process a leader or head of the family is needed. Thus, Qiwamah can be interpreted as leadership in a family. 17 In the Big Indonesian Dictionary it states leader is a person who shows the way, lead as well as guiding. According to The word qawwamun is found in sura an-nisa verse 34, etymologically qawwamun is the plural qawwam which comes from the word qama yaqumu qauman wa qiyamun which means to guard, maintain and do good. In Arabic words, the word qawwamun can be interpreted as a person who is responsible for matters related to his wife and cares about his wife's condition. 19 Some scholars agree that the concept of Qiwamah contained in Sura an-Nisa verse 34 is an elaboration of male leadership in the domestic area or household because men have several advantages over women. According to at-Thabari, among the advantages of men over women is to give dowry and provide for his wife. Likewise, according to al-Zuhayli, where men have advantages because they give dowry, besides having excess intelligence, better physical strength, and stable emotions. Although the above interpretation is biased, it shows that the leadership of the man is the domestic or household territory. 20 When the Qur'an has laid down the duties of leadership to men, it is because of two main things. First, this is because between men and women there are different privileges, but if it is related to the context of qawwam, male privileges are considered appropriate and worthy to carry out the task. Secondly, the Qur'an speaks of leadership to men because they are the ones who will provide livelihood for their family members. 21 Traditionally the head of the family is someone who is responsible for meeting the material needs, education of children and a good environment for the family. Viewed from the physiological side, a man has advantages over women such as being strong and having greater energy, so that men must be the head of the family who plays a role in earning a living to support his family or household. A leader or head of the family is required to meet the material, emotional, intellectual and socio-moral needs of each family member. So, the man with his status as a head of the family must have a responsible nature, able to control the situation in the family at home, as a founder and originator of policies, guides and decision makers for his wife and children. 22 Islam recognizes the importance of the existence of leaders in a group, how and whatever that group is. Muhammad Ali Albar in his book entitled 'Amal al-Mar'ah fi al-Mizan states that there is an Islamic rule that dictates that if there are three people who travel then the leader must be appointed from one of them. The leader is someone who must be the most responsible party and to his leadership has the least interest. Based on this rule, he held the view that in the home it is not worthy of leadership to be in the hands of women, because men need less protection and more steadfast, so on his shoulders God places responsibility in leading the household. 23 However, if the man does not have the ability to lead and provide for his family or the situation reverses where the wife --- Gender Justice Gender is about ideals between men and women at a certain time and place, which can change according to changing times, times or certain community conditions. 25 In general, gender is used to identify or show differences between men and women that can be seen from behavioral, social or cultural. 26 The word gender can also be defined as differences in functions, status, roles, and responsibilities between men and women as a result of socio-cultural constructions that exist in society. This difference is then rooted in society through the process of socialization Therefore, men and women can participate according to their respective natures, which should not be limited by man-made legal regulations. 28 If there are those who hold or think that the verses of the Qur'an and Sunnah that are shahih contain gender bias, the subject matter is not from the Qur'an and Sunnah but from the views arising through that party that is the subject matter. So the emergence of man-made legal rules related to gender is very possible as long as it does not violate the provisions of Islamic Shari'a. 29 Justice is a very important and essential thing in the legal world until it has become a Some people think that justice must be equal to equality, because it is not absolute that equality between men and women is fair. 31 Gender justice is a process and fair treatment between men and women, so to be able to ensure that the process is fair to both requires actions that stop things that historically and socially hinder and hinder men and women from participating in their respective duties and functions. Fair treatment between men and women in all respects makes for a good relationship to lead to more positive goals. 32 --- The Role and Responsibilities of --- Husband and Wife in the Family The success of a man is supported by love, motivation and prayer from a woman . Similarly, the success of a wife is also supported by the provision of 30 motivation, access, and sincerity of a husband 33 Therefore, in their roles as husband and wife, both can perform balanced roles, including: 1. Sharing joys and sorrows and being able to understand the function, role and position of husband or wife in their social life and profession, provide access, support, be able to share roles in certain contexts and can play certain roles together. 34 Since families that allow for a division of roles in domestic affairs will eliminate gender discrimination, sharing these roles is also very good to avoid multiple roles for one party ; 2. and can position oneself as a wife as well as a friend and lover for the husband. Similarly, placing the husband as a lover and friend who both need affection, attention, advice and motivation also have the same responsible nature to empower each other in spiritual, intellectual and social life. 35 The role of husband and wife in this case aims to cultivate a sense of rahmah, mawaddah and sakinah, because there are many attempts to position both of them in obtaining their basic rights well; People who still adhere to patriarchal culture determine that the responsibility of earning and providing for family members is a husband. As for a wife, it is more focused on roles in the realm of the household. 37 The standardization of roles between husband and wife will not be a problem if the wife wants it, or the wife has decided to become a housewife without pressure from any party, and based on reasons and considerations that actually provide comfort for the wife, then the choice of role will not be a problem. 38 The role of the husband as the head of the family and the wife as a housewife, is directly proportional to the noble duty attached that the husband is obliged to give everything for the needs of married life in accordance with his ability and is also obliged to protect his wife. In the Compilation of Islamic Law it is detailed and clearly mentioned about the duties of the husband as the head of the family, namely. 39 1) The husband is a guide to the wife and her household, but in matters of household matters the importance is decided jointly by the husband and wife; 2) the husband is obliged to protect his wife and give everything for the needs of married life according to his ability; 3) husbands are obliged to provide and teach religious education to wives and provide opportunities to learn about knowledge that is beneficial and useful for religion, the archipelago and the nation; 4) According to his ability, a husband must bear: a. bread, clothing, and shelter for the wife; b. expenses for household needs, medical expenses and care for the wife and children; and c. Tuition fees for children. --- The Concept of Qiwamah and Its --- Implications for Gender Justice Regarding leadership in the family is found in verse 34 of Surah An-Nisa, which means, "Men are leaders of women, because Allah has favored some of them over others , and because they have spent part of their property..." The placement of the man as the head of the household was actually a reaction to the conditions and social situation of Arab society before and when the Qur'an was revealed. So, in this case, the role of men is very dominating in various areas of life even including the area of the family system. In addition, the above verse indirectly shows that by nature, men tend to want to protect women. 41 Men in Arab society, having the duty of defending and defending all members of their family, also have the responsibility to meet all the material needs of family members. As a result, men master leadership at all levels. 42 The reason why men dominated to lead their households was, "First, the superiority of human resources , which at that time 41 Abdul Rahim, "The Role of Women's Leadership in a Gender Perspective, Al-Maiyyah Journal, Vol. 9, no. 2, December 2016. hlm. 288. 42 Ahdiyatul Hidayah, "Scholar's Perspective on Wives as Family Breadwinners During the Covid-19 Pandemic", An-Nisa': Journal of Gender Studies, Vol. 16, no. 1 of 2023, hlm. 43-58. 43 Surya Sukti, et al. "Gender justice: education, leadership and rights Islamic inheritance." Journal of Religion and Society Studies Vol. 18, No. 2 of 2022, hlm. 130-137. generally still belonged to men. Second, at that time economic power was also mostly still in men, so they were the source of income for family members. 43 Regarding male leadership in his household, the Prophet Muhammad said, "Each of you is a leader and responsible for what he leads. A priest who leads people is a leader and responsible to his people. A husband is a leader in terms of taking care of and leading his family members, so he is responsible for those he The institutionalization of the distribution of roles that states that men are the head of the family and women as housewives, by women activists are feared to reduce equality between men and women. They also criticize it as discriminatory and legitimize the power of the husband to be able to regulate all matters related to his domestic affairs. Along with the emergence and strengthening of awareness about gender equality, the division of roles that the husband is the head of the family and the wife as a housewife began to be questioned. In the context of conception, the division of roles seems to give birth to an implication that women are subordinates of men, which is clearly contrary to the spirit of gender equality. In reality in the field, there are many households where his wife performs a very strategic role, such as being the breadwinner as well as managing the affairs in the household. Therefore, regarding the absolute husband as the head of the family, according to them, it is not in line with existing practices in the field. 48 With all the progress of the times that have changed a lot. The impact of modern industry, as well as the increasing need for health and education, this opens up new opportunities for women to take part outside the home, whether it is working or participating in institutions in the community. 49 In the reality of social life that has undergone many changes, the sole and Gender Studies, Vol. 4, no. 1, March 2018, hlm. 174. breadwinner in the family is not a big problem if it can meet all the needs of the family, so as to form a sakinah and prosperous family. But if the sole breadwinner is unable to provide for his family, then the reality is that society has undergone a shift where whether or not able, ready or unprepared wives can take on roles by working outside the home. 50 In addition, the above verse shows that these two things strengthen the reason for providing opportunities for women to become leaders. The Qur'an mentions men, because at the time of the Qur'an's descent the leadership potentials according to Q.S An-Nisa: 34 above were still the domain of male leadership. But looking at the reality now, the potential for excellence could come from the wife's side as well. 51 Aminah Wadud Muhsin stated that the man as the leader is not intended to show superiority to the man Qur'an, namely being able to prove his superiority and provide for his family members. This means that two criteria must be met by men to become a leader in their family, namely managerial ability and the ability to provide a living. 52 According to Fatimah Mernissi, leadership in the family will not always be in the hands of a man. A man becomes the leader of his family if he can provide a living, that is, the man is functionally able to work and able to provide for his wife and children. However, if the man is unable to provide for his wife and children and is unable to provide a living, then he does not become a leader in his family, because the man has no advantages. Thus, his leadership in the family is lost if he cannot provide for his wife, while one of the factors of superiority of men over women is in terms of providing for them. 53 Because the family is the backbone and soul of society that determines the inner welfare and birth of a country. So, in Islam it is highly recommended to build and form an ideal family which is often referred to by the Muslim community in Indonesia as sakinah. 52 Sulaiman Ibrahim, "Domestic Law and Women's Leadership in the Family", Al-Ulum Journal, Vol. 13, no. 2, December 2013, hlm. 230. 53 Achmad Ghufron, "Leadership Family Perspective Islamic Femenism ", al-Thiqah Vol. 3, no. 2, October 2020, hlm. 135. must be carried out and must even be able to be accounted for by the man if the other party feels disadvantaged. However, the position of the husband as the leader or head of the family in his household does not necessarily make the position of the wife below the husband. The husband as the head of the family and the wife as a housewife are only related to function, not referring to the structure of the husband and wife relationship. 57 The head of the family is the husband and wife of the housewife, not intended as a division of roles that structurally place the wife under the husband, but rather a division of noble household duties only. So there is no need to worry too much that the position of the husband as a head of the family, there will be arbitrary actions of men against women. In --- Conclusion It is known that Qiwamah is someone who leads, someone who has responsibility for strength. In relation to the family, a family will be able to run well if each performs its functions well, then in the process a leader or head of the family is needed, so Qiwamah can be interpreted as leadership in a family. The concept of Qiwamah itself is not absolute, but can change and be exchanged. So that the role of leadership can be realized properly without any party who feels that he does not get justice in doing his role to participate in organizing, directing and guiding his family. This means that leadership in the family is held by performing balanced roles both male and female. --- superiority of men in organizing, directing and providing for their families is not natural, but social. 55 Therefore, the role of a wife in her participation in leading by participating in organizing, directing and guiding her family must also be recognized, because there should be no difference between husband and wife related to helping each other in meeting the needs of their family, educating and directing family members. So that the role of leadership can be realized properly without any party who feels that he does not get justice in doing his role to participate in organizing, directing and guiding his family. Also, a fair division of roles between husband and wife will encourage the realization of a sakinah, mawadah, and Rahmah family. 56 According to religious, positive or socio-cultural laws in Indonesia, all of them equally place the man as the head of the family, but it still contains balanced justice. That is, the position as the head of the family is given to the man , accompanied by duties and obligations that Musawa Journal of Gender and Islamic Studies Vol. 17, no. 2 Years 2018, hlm. 137-152. 56 Ramdanil Mubarok, "The Role of Leadership in the Family in Online Learning in North Sangatta Village." Literacy: Journal of Non-formal Education Sciences Vol. 7, no. 3 of 2021, hlm. 251-1262.

Family is the most important element in the formation of a society. Islam has a very fair concept in family leadership. In Islam, rights and duties are always regulated in balance. This balance results in the establishment of a harmonious relationship between husband and wife in the family. This paper uses an analytical descriptive method that examines Qiwamah issues related to women's leadership in the household. And the aim of this research is to find out how the concept of Qiwamah and gender justice in Islamic family law against the concept of Qiwamah in its implications for gender justice. This paper uses analytical descriptive method. The results of this study found that Islam gives leadership in the family to husband and wife, if both have the qualifications to be leaders. Leadership is not only given to men (husbands) only. Leadership is given to both parties if both have leadership qualifications, be it wives or husbands. In Islam, leadership/ Qiwamah in the family aims to create harmonious conditions so that the family can achieve happiness both in this world and in the Hereafter.

Introduction Globalization, internationalization, and human mobility have increased the number of persons considering education abroad [1]. Studies show that the number of international students living and learning abroad has risen from 800,000 in 1975 to 4.6 million in 2015, and it is predictable to reach 7.2 million by 2025 ). While the United States , the United Kingdom , Australia, France, Canada, Russia, and Germany have always been the favourite destination for students seeking higher education overseas, non-English-speaking countries, like, Japan, South Korea, India and China have also risen as desirable destinations for higher education [2]. Particularly, with about 500,000 overseas students enrolled in 2020, the rapidly expanding Chinese economy has developed its higher education into an "international education center" . China has moved from being the top sender of international students to one of the top ten host countries of international students. Since 2014, it has been the third-largest host country after the US and the UK [3]. Most international students studying abroad are engaged in various forms of social media platforms. Though the use of social media platform comes with positive effects such as creating and sharing information with others, examples include sites like Facebook, Instagram, Snapchat and Twitter, which allows us to connect and communicate with others, share a common challenge/interest, Access to Information, allows us to inspire others to do things, Furthermore, social media can increase voter participation as well. Besides, the ugly part of social media there are tons of unnecessary information shared by people and also bullying and harassment on social media has been increased. People can make brutal and negative comments about anything and anyone. Social media has good, bad and ugly impacts on our culture. As hypothesized, a separation in language preferences on social media has a significant and positive effect on enculturation and a negative effect on acculturation, suggesting that international students' immigrants' preferences for their heritage languages online lead to the maintenance of heritage cultures in offline contexts. There is a perceived discrimination and ethnic identity crisis including other risk factors in the process of acculturation in the use of social media platforms. Hence, social media use can also negatively affect students and teens, distracting them, disrupting their sleep, and exposing them to bullying, rumour spreading, unrealistic views of other people's lives and peer pressure. The risk factors are many, hence an investigation to ascertain the impact of social media. This topic is wealth research because it has a role to play in international students' life and acculturation in China. In line with the above, international students contribute significantly to the host country's economy and cultural diversity, they frequently struggle to adapt to a new culture, resulting in problems and stress during acculturation, which sometimes interferes with their academic performance [4]. Though social media comes along with many advantages, social media also has several disadvantages that in certain situations limit students' concentration and academic progress in school. Social media can be a distraction for some graduate students. Students may be distracted from their school work and the teachers will have no option of knowing which student pays attention during lecture sessions. Research on acculturation indicates that adjusting to a new culture can be challenging [5]. Generally, people experience adverse feelings such as anxiety, isolation, depression, and other negative emotions [6]. However, as internet service and modern telecommunication technology, especially the use of social media, has increased over the past few decades, international students' acculturation experiences and cognitive abilities to integrate into other communities have improved [4]. According to [5] social media contains technical and social features that allow international students to adjust quickly to the adaptation process while maintaining ties with their home country. Consequently, previous empirical studies in the area of acculturation have consistently found that international students are more likely to use social media to expand their social networks, learn about the host country's culture, form friendships with friends and family, and fulfil various needs in a non-native environment while quickly adapting to their new surroundings [5]. For instance, [7] surveyed 283 Korean and Chinese college students in the U.S. to assess their acculturative stress. They demonstrated that international students use social media to post writings and photos about their new lives in the host country, allowing them to maintain contact with important friends back home and thus improve their acculturation process. [8] use 146 international students in Argentina to show that social media and video chat could help international students find people with similar interests, make new friends in a new place, and feel less stressed about adapting to a new culture. [9] explore the role of major US social media in Chinese students' acculturation and show that social media contributes to higher acculturation and better adaptation. [10] explore the role of social media on international students enrolling in South Korean universities and find that the use of social media helps to reduce international students' acculturative stress and to enhance their satisfaction with Korean life. [6] survey international students at a large American university and discover that social media was associated with less stress from cultural adaptation because social support from the host country was greater. More recently, [11] examine how 27 transient international students in the UK use social media to negotiate and control their identities. The study shows that social media can help people maintain their home culture, adapt to the host culture, and integrate two cultures. Others have discovered that social media can help international students socially adapt to educational environments in other countries. By administering an online questionnaire to a group of 120 Chinese international students, [12] discover that those who utilize social media more frequently demonstrate greater social and academic adaptation to the foreign culture. Similarly, [4] investigate the psychological and behavioural effects of social media on the acculturation processes of Chinese students studying in the United Kingdom. Their findings indicate that the psychological acculturation of Chinese international students to the host culture is unrelated to their academic performance. We define social media as an internet-based application that allows users to create a public or semi-public profile to interact, share, and exchange information and ideas with others in online communities and networks. Researchers have increasingly examined the relationship between international students' media usage and their cross-cultural adaption in recent years, there have been relatively few studies on this topic in China. The majority of previous studies only concentrated on Chinese students' experiences in US and UK universities and how their expectations were met or not. Perhaps, a surge in the number of Chinese students in US and UK universities may have increased the scholarly study on Chinese students. According to the International Education Exchange Open Doors report published by the Institute of International Education , China accounts for roughly 31% of international students in the US. China also offers the UK, Australia, Canada, Japan, and New Zealand the largest number of international students. However, as indicated earlier, China, which once sent many of its students abroad, is progressively becoming a recipient of international students to boost inward student mobility [13]. International students in other nations may have similar experiences in host cultures, and certain cultural customs in China can make adaptation difficult for international students [14]. Despite historical and present influences from the West, China's socio-cultural structure is generally different from the Western world and even its nearest neighbours. Chinese educators are aware of these disparities and are attempting to assist international students to adapt to Chinese sociocultural and educational practices [15]. Even though various aspects of international student adaptation tactics, particularly social media usage, have been proposed to increase acculturation and student engagement in China, there is no study investigating this issue [16]. Moreover, China uses a different set of social networking sites and apps than the rest of the world, and more importantly western social media platforms like Facebook, Twitter, Instagram, YouTube, and WhatsApp are blocked in China. To maintain the connection to these western social networks, a Virtual Private Network is required to access the restricted Western social media platforms; however, VPN use is not always possible in China. Thus, the extent to which social media improves international students' adaptation to Chinese culture and, ultimately, their school activities is questionable and remains unclear in international academic circles. Hence, this necessitates scholarly research on the effects of social media on international students' acculturation and its outcomes on their university experience. Building upon the foregoing discussions, the current study uses a quantitative research design to increase our present understanding of international students' acculturation by examining the impact of social media use on international students in China acculturation processes in terms of psychological and behavioural dimensions, and b) exploring the link between acculturation and student engagement among overseas students in China. Ethnic self-identification is employed as a moderator to investigate the intermediation effect of social media use on acculturation among international students. Ethnic self-identification is crucial to the ideological and behavioural changes during an individual acculturation process. Ethnic self-identification is defined by how tenaciously and forcefully they keep their ethnicity and cultural characteristics when engaging with people from various ethnic and cultural backgrounds. The rest of the paper follows this particular order. Section 2 discusses the literature review and hypotheses. Section 3 highlights the methods and materials. Section 4 provides the results, and the section discusses the outcomes. Finally, section 5 closes the research with policy recommendations, limitations and further research. Berry refers to acculturation as "the dual process of cultural and psychological change that takes place as a result of contact between two or more cultural groups and their individual members" . In other words, acculturation happens as people from various cultures interact on a daily basis, leading to changes in their mental well-being and behaviour as a whole [17]. Mental acculturation is changed in the psycho-cultural orientations of individuals [18]. While behavioural acculturation consists of multi-faceted languages, such as media uses, religion and food consumption [19,20]. --- Literature review --- Acculturation and social media usage Like any other immigrant, previous studies have revealed that international students may find it hard to adjust to a new cultural setting and may suffer elevated stress levels due to the supposed cultural barrier. International students could face cultural "shock or stress" upon arrival in a new host environment, primarily impacting their mental well-being [21]. It is possible that they will experience significant cultural and ethnic differences between their own country and the host country [19]. They may struggle to communicate and study in different languages and cultures. However, given the rapid advancement in technology, there is wide use of various social media platforms such as Facebook, WhatsApp, WeChat, etc., mainly among young people [22]. Research indicates that these social media platforms enable international students to connect with family, friends and coworkers almost anytime and from any location expediting their acculturation processes. As a result, scholars have attempted to study the link between social media use and acculturation from multiple perspectives. Many of these investigate whether home or host social media affect the acculturation process [23]. According to these studies, using media from the host country can help acculturation by offering access to mainstream culture, whereas using media from the home country can obstruct the process by limiting opportunities for cross-cultural communication. Cemalcilar et al. , for instance, stated that using social media from the home country is linked to a greater degree of support for ethnic identity while using host countries' social media is linked to stronger social and cultural adaption to the host country. Similarly, in the United States, [24] show that media consumption in the host nation correlates positively with social interaction, English proficiency and awareness of the host culture. More recently, [23] found that among 121 Chinese students in New Zealand, those who used host social media more frequently were more likely to identify with New Zealand. In contrast to previous studies, this study focuses on how international students in China use social media content to enhance their acculturation process. According to [25], people use social media to seek information to fill knowledge gaps, fulfil entertainment needs, and connect with others. This research, therefore, aims to ascertain how international students use social media during their studies in Chinese universities, focusing on knowledge sharing, interaction, and entertainment. Bearing this in mind, we assume that when international students use social media to interact with their peers and family, enjoy various entertainment activities, and share knowledge across multiple channels, they are much more likely to understand the significance of what is happening in the host culture, maintain up-to-date knowledge, and reduce culture shock [20,26]. Therefore, hypothesize that: --- H1a. There is a positive relationship between the use of social media by international students in Chinese universities and their mental acculturation to the host culture . H1b. There is a positive relationship between the use of social media by international students in Chinese universities and their behavioural acculturation to the host culture . --- Acculturation and international students' engagement Students who move to a new country encounter various challenges, including adjusting to new cultural, ethnic, and academic surroundings [15]. Individuals attempting to re-establish themselves in a new setting are frequently driven to make adjustments that have mental and behavioural ramifications, making the transfer into the new environment extremely difficult [27]. International students come from different educational cultures and environments, and they may face a wide range of academic challenges and other forms of challenges such as the language barrier [28]. These challenges do not make it any easier for international students to adjust to a different educational setting [4]. Previous studies have examined the link between international students' acculturation to their host culture and their engagement in school activities [29]. For instance, [30] demonstrate that international students have less social support, adopt abnormal coping strategies, and have greater inconsistencies between their academic expectations and university life experiences than their domestic counterparts. [31] evaluate the academic achievements of both local and international students in five Netherlands schools and find that international students outperform local students. Using primary data gathered from 383 research students in various universities in eastern China, [32] demonstrate a positive correlation between social media usage and student engagement. [33] used 1050 international student nurses from three different countries to demonstrate that using social media is critical to improving the learning of international student nurses. As a result, they propose that social media be used as a learning tool to make nursing curricula more inclusive and equitable. Although these and other studies have demonstrated that the results are mixed and complex, their research focuses on either mental/psychological or behavioural changes during the acculturation process [31]. According to [17], acculturation occurs when people from different cultural backgrounds regularly interact, resulting in psychological and behavioural changes. These two acculturation processes are critical for international student engagement, but most previous research has focused on the psychological perspective [19]. This knowledge gap necessitates additional research into the impact of social media on the acculturation of international students from both a psychological/mental and behavioural standpoint. Therefore, following [4]., this paper contributes to the body of literature by investigating the impact of international students' acculturation on student engagement from both psychological/mental and behavioural standpoints. We suggest that international students in China with a better knack for adapting to the host country's culture may have higher levels of student engagement. Therefore, the study proposes that: H2a. There is a positive relationship between international students' psychological/mental acculturation to the host culture and student engagement . H2b. There is a positive relationship between international students' behavioural acculturation to the host culture and student engagement . --- Social media use and international students' engagement Using social media in the classroom opens up several possibilities, including increased student engagement, the development of students' social media skills, and the formation of students' professional and peer networks [34]. Given the various challenges that international students may face, such as adjusting to a new educational and cultural environment, social relationships can aid in this process [35]. To this end, research on international students' use of social media shows how emerging digital platforms can aid in their adaptation to new educational environments. According to [36], social media facilitates productive learning, networking, communication, and knowledge exchange. Similarly, [37],discover a positive relationship between social media and student engagement in the United States. [38], also show that most international students who spend more time on social media platforms feel more connected to their classmates than those who do not, thereby increasing academic collaboration. [39], noted that social networking sites significantly increased cooperation between domestic and international students, resulting in increased international student engagement. In a recent study, [34] find that incorporating social media into undergraduate and graduate courses helps increase student engagement and build peer and practitioner networks by showing students how course content relates to the real world and providing opportunities for them to connect with one another. The use of social media to foster collaboration may raise concerns about privacy and security [40]. Given that sometimes there is no face-to-face contact and the opportunity to approach other students physically, international students may be at risk when working with people they do not know [41]. Furthermore, social media exposes international students to a wealth of data and information that might be difficult to digest and judge for accuracy owing to cultural differences [42]. Exposure to inappropriate online content, online stalking, and abuse all contribute to the inherent difficulties of using the social media of the host country [43]. Other studies have also reported that social media may distract students from engaging in school activities. Therefore, these issues and concerns overshadow the debate about whether social media can increase student engagement. However, because our focus is on postgraduate students who are older and more mature and thus better able to manage the risks associated with social media, we hypothesize: --- H3. There is a positive relationship between the use of social media by international students in Chinese universities and their engagement in education in the host country . --- Moderating role of ethnic self-identification According to the identity theory, individuals are identified with a particular social group [44]. This social group is a group of people with a common identity. People find out which social groups they are a part of through the social group identification process, which helps them form their identities as members of an in-group or an out-group. Culture is one of the foundations for the identity of an individual [45]. Attempting to strike a balance between two different cultures or fitting in with the host culture will alter a person's cultural identity. To form a new cultural identity, international students must go through a constant and dynamic process of negotiating between host and home cultures, i.e., maintaining the culture of their native country while also adapting to the culture of the country in which they reside [30]. International students with strong cultural identities are expected to have more difficulty adapting to the host country's culture [46]. Ethnic self-identification highlights how migrants show and maintain their cultural characteristics when abroad [47]. Being highly motivated to make use of home media and communicate with people from the same country, especially speaking their local language, international students are considered unwilling to integrate into the host country which hinders their acculturation process [4]. But, according to [48] international students may exhibit dual cultural identities, which suggests that individuals can adopt the host country's culture while remaining loyal to their own traditional culture. [49] proposed the coexisting cultural self-identification model. In another study, [50] concluded that international students who love their cultural identities and who regularly associate with those from the same ethnic heritage are more self-esteem. Therefore, we hypothesize that: H4a. International students' self-identification has a moderating effect on the link between the use of social media and psychological/mental acculturation . H4b. International students' self-identification has a moderating effect on the link between the use of social media and behavioural acculturation to the host culture . --- Method --- Ethical approval Ethical approval was sought from the Jiangsu University Institutional Review Board and the International Education Unit in China to conduct the study. A recording device was used to take oral consent from participants as indicated by the ethics committee. The purpose of the study and other details were disclosed to the authorities and participants. Participants were not financially induced or coerced to take part in the study. It was explained to them that their participation was voluntary. --- Participants and materials To test the hypotheses stated above, a purposive sampling approach was used to select postgraduate international students in China who were 18 years of age or older and had studied in a Chinese university for more than six months. We conducted online surveys by administering a questionnaire to gather responses from international students in several universities across China. The research is quite interesting because the survey was administered according to the research plan and methodology. The investigation proceeded without deviating from the chosen methods of analysis towards achieving the research objectives. For instance, to get adequate and first-hand information to support the research findings and make a generalization on social media usage, survey research was chosen over other research designs. This gives room for more generalized results and outcomes in situational analysis. We also advised everyone qualified to partake in the research to spread the word by inviting their friends and colleagues to join them. Participants were asked to provide accurate information and were guaranteed that their data would be collected without revealing any personal information and would only be used for this study. Meanwhile, participants were asked to stop participating in the online questionnaire if they had any hesitations about it. The SEM method is appropriate for establishing a correlation among variables, assessing the theoretical model and getting the research hypotheses tested with firm results. More so, the chosen method offers immediate and long-term results. The research model was tested in a pilot study on a convenience sample of forty social media users drawn from the studied population to ensure its suitability, reliability, validity and freedom from error. Following the validation procedure, a total of 395 responses were collected using online questionnaires in the English language over three months from 32 different universities located across mainland China. These universities were selected because they have sufficient numbers of international students, who enroll in English language programmes. However, 41 responses were removed from the dataset due to uncompleted data. We did this to get a complete dataset for the analysis. Besides, in dealing with missing data at random, related data can be deleted to reduce bias. Especially in cases where the data gathered are more than enough for generalizing the subject under study. --- The measures considered in the survey The questionnaire used for the study is made up of two parts. The first portion included questions about the participants' demographics . The second part contains measures, which were built on earlier studies. i. Demographic variables. Participants were instructed to state their gender , age , ranging from 1 to 5 and the number of residences in China on average-point Likert scale, ranging from 1 to 4 . ii. Social media usage. In the same way as [4] measured social media usage, we measured it using three constructs: information sharing, interpersonal interaction, and entertainment. The participants were asked to indicate whether or not they believe that using social media platforms improves their ability to share knowledge, allows them to interact with their peers, and enables them to entertain themselves. Each statement was gauged according to the average-point Likert scale, ranging from strongly disagree to strongly agree. Sample statements related to knowledge sharing, entertainment and social interaction are provided in Table 3. Though the result regarding the hypothesis of 'usage' seems quite broad, the usage component of social media analyzed in the study can further be explored in other studies to complement the literature on research and development under the subject under discussion. iii. Acculturation. Acculturation was also assessed using two constructs: Psychological/ mental and behavioural acculturation, both of which were used to measure it as in [4]. Sample questions for psychological variables include: "With which group of people do you feel you share most of your beliefs and values?", "With which group of people do you feel you have the most in common?" etc. The variable 'behavioural' was also developed based on earlier conceptual frameworks, mainly those outlined by [19]. Like in [19] and [4], we consider behavioural variables under two frameworks, i.e. language and media use. Sample terms include: "I feel very comfortable speaking my local dialect with my friends"-for language and "The Internet websites that I browse are mostly in Chinese"-for media . Each statement was scored on a 5-point Likert scale between 1 and 5 iv. Self-identification. Self-identification is a feeling of belonging and devotion to a particular ethnic group. In other words, it involves being associated with a specific ethnic group that shares common values and attitudes and sentiments of belonging and commitment. This variable was developed based on earlier conceptual frameworks, mainly those outlined by [51]. On this note, we measure this variable using four constructs. See details in Table 3. iv. Student engagement. [52] describe student engagement as "participation in educationally effective practices, both inside and outside the classroom, which leads to a range of measurable outcomes". Based on this definition, like in [32], we adopt the student engagement measure from [53]. Sample terms include: "I often miss my school/research lab", "I often cut/ skip my classes/research workshops", etc. See Table 3 for details. Each statement was scored on a 5-point Likert scale between 1 and 5 . v. Language. Research by [54] confirmed that the utilization of social media has been significantly perceived to have positively impacted learning the English language in terms of writing style, reading skills, listening and lexical variation, communication skills and grammar usage. In the future other potential underlying variables other than what we have tested could be factored into the analysis of student usage of WeChat and other social media platforms that is tied to understanding the language before its effective usage. Because having a prior understanding of language makes it easier to use other social media platforms for timely purposes with acculturation and easier time engaging in the coursework of study among students. --- Data analysis In the first section, we present the descriptive statistics which offer insight into the general characteristics of the respondents . Subsequently, a correlation matrix was executed to investigate relations amongst the main measures . In the second section, we further examine the correlation outcome using structural equation modelling to assess the above theoretical framework of the study and the proposed hypotheses. As a result, in this second stage, we first check the measurement model using the Amos 24, which was then used to find the causal relations between the observed and unobserved construct . The model was evaluated by looking at its reliability and validity testing results, among other things. In the second step, a SEM is evaluated using a multiple regress technique like hypothetical relationships based on the sign, magnitude, and significance level . --- Stage one: Descriptive statistics Of the final 354 international students, 82 were female students and 272 were male students . Concerning age, 3.39% of the participants were of the age bracket 18-25 years, 35.03% were of the age bracket 26-30 years old, 39.54% were between 31 and 35 years old, 20.34% were between 36 and 40 years old and 1.70% between 41 and 45 years old. Also, of the total number of international students, the majority of them had been living in China for between 3 to 4 years , 11.86% between 1 and 2 years, and 8.76% have been in China for 5 years and above. 70.06% were pursuing their master's degree and 29.94% their doctorate. According to the survey, the most highly used social media is WeChat , followed by WhatsApp , Facebook and Instagram . These social media is used between 2 and 5h per day and more than 8 hours . --- Correlations among variables Table 2. --- Stage two : Measurement model results As previously stated, the current study used SEM and AMOS 24 software to assess the theoretical model and research hypotheses. This process is carried out in two phases: In the first phase, a confirmatory factor analysis was performed to test the measurement model. In the second phase, we assessed the hypotheses of the theoretical model employing multiple regression techniques. SEM was selected because it makes it easier to assess hypotheses based on empirical measures that propose complex links outlined in theoretical models. In the first phase, we carried out a confirmatory factor analysis . CFA is a statistical method for confirming the factor structure of a group of observed variables. It aids in the identification and determination of construct validity and reliability . The total model was subjected to CFA using AMOS version 24. A reliability measure is applied in examining internal consistency by computing observed items and avoiding superfluous dimensions generated by factor analysis owing to garbage items [15]. The coefficient alpha was utilized to test the internal consistency reliability because it is the most generally used internal consistency technique that reveals how distinct items can measure diverse aspects of a construct [26]. The CA scale runs from zero to one, with values below 0.6 indicating low reliability [55]. [56] stated that when internal consistency is low, the content of the items will be too heterogeneous. CA should have a minimum threshold of 0.7 [57]. The CA in this research was greater than 0.70 , indicating that the findings are acceptable because of past studies see [57].. The construct-level reliability, also known as composite reliability demonstrated that items belonging to the same constructions had a strong association. It was suggested that the composite reliability be larger than 0.7 [57]. Similarly, in the present study, the CR were above 0.70 indicating that the results are acceptable in light of the previous research . In this study, the discriminant and convergent validity tests were carried out to validate the measure. Discriminant validity relates to how much measures differ from other operationalizations, demonstrating that the construct is genuinely different from other constructs [58]. Another technique for checking discriminant validity is by computing the extracted average variance for the constructs and comparing it to the square correlation between them [51]. [56] recommended that values for each construct be larger than 0.50 when calculating the average variance extracted . The results for the AVE in this study were above 0.50 ranging 0.678 from to 0.838, supporting the discriminant validity. An alternative measure to confirm the evidence for discriminant validity is when estimated correlations among factors were lower than the required value of 0.92 [4]. Convergent validity was investigated based on construct reliabilities and was used to test the construct's homogeneity. The correlation matrix for constructs with a cut-off value of 0.92 is shown in Table 2. The Mann-Whitney U test was used to calculate possible non-response bias by comparing the differences between early and late responders about the average of all variables [55,59]. Here, the first fifty observations were considered early responders, while the final fifty were considered late responders, based on the proportions of the times at which the online questionnaires were answered and returned. The results revealed there is no statistically significant difference between early and late responders; thus, non-response bias is not an issue. We used the Harman single-factor test to assess the common latent factor and common method bias suggested by earlier research, using differences in chi-square values among the original and fully constrained model [4,46,55]. The findings suggest that the two models are statistically different and share a variance. The model's initial results were assessed without considering the method biases. Furthermore, we adopt the [60] four common method variance categorization sources. As a result, measuring context effects played a large role in the magnitude of CMV in our investigation. The model's original results were then examined without considering the method biases, and CFA was suggested. The model fit was appraised using the Comparative Fit Index , with a value of 0.932 greater than 0.90 indicating an excellent fit and for the Root Mean Square Error of Approximation , a value of 0.075 less than 0.08 indicates an acceptable fit. The Tucker-Lewis index , incremental fit index , and normed fit index of the suggested operational model were 0.915, 0.912 and 0.882, respectively, with a chi-square of 2442.285 . Furthermore, the findings reveal that CMV was not the primary cause of the changes in the observed items [61]. Having done all these, in the second phase, we carried out the hypotheses testing of our model. --- Stage two : Hypotheses testing In this section, we tested our hypothesized model. To test the model fit, a variety of goodnessof-fit measures were used and this includes, including the comparative fit index , the Chi-square statistic, the root mean square error of approximation , and the standardized root mean square residual . An insignificant Chi-square, an RMSEA value of less than 0.06, a CFI value above 0.95, and an SRMR below 0.05 are all characteristics of a good research model. The model fits the data based on a variety of goodness-of-fit measures with a chi-square of 2324.342 , RMSEA = 0.032, SRMR = 0.036 and CFI = 0.953. Table 4 shows the results from the model. The H1a and H1b hypotheses, which propose a direct connection between social media use and mental and behavioural outcomes, respectively, suggest that students' use of social media is positively related to students' psychological/mental acculturation and behavioural acculturation . For research hypothesis H2a, the findings indicate that there is a significant linkage between psychological/mental acculturation and students' engagement . Similarly, H2b shows a statistically significant link between behavioural acculturation and students' engagement . The results for hypothesis H3 show that social media usage is positively linked with students' engagement . Finally, the outcomes reveal that self-identification is a mediator in the relationship between social media use and mental acculturation and behavioural acculturation . In other words, the outcomes suggest that the effect of social media on behavioural and psychological/mental acculturation could increase due to self-identification. The results of the regression coefficient are also illustrated in Table 4. --- Discussion and implications --- Discussions This research paper is one of the first to investigate the linkages between international student's use of social media and the impact it has on their acculturation and participation in school activities while in China. Specifically, we discovered that international students use of social media is positively linked with their acculturation which supports H1a and H1b. In other words, the study's findings indicate that when international students use social media to share or exchange information, create contact, or entertainment, they are more likely to become mentally and behaviorally acclimatized to the host culture throughout their time studying in China. Researchers have previously discovered that this particular group of students may experience greater challenges with language and communication as well as sentiments of homesickness and loneliness than the general student body [24]. However, given the widespread use of social media, this research reveals that international students may obtain the majority of their knowledge from online media platforms. According to our findings, international students in China spend anywhere between one and six hours every day on social media. They primarily use WeChat , followed by WhatsApp . This is consistent with previous findings that the amount to which international students rely on WeChat appears to contribute to their overall contentment with Chinese living [14]. A large number of previous research have demonstrated that the more an international student interacts with the host country's social media, the more likely he or she is to acclimatize to the host culture [6]. As a result of this finding, international students may use Chinese social media such as WeChat to communicate with local communities and students in their host countries. The research findings also reveal that there is a connection between international students' psychological/mental and behavioural acculturation to Chinese culture and their participation in school activities . This suggests that when overseas students are acculturated to the host culture from a psychological standpoint, their self-esteem is boosted, which leads to an increase in their participation in school activities, hence boosting their student engagement and overall academic success. Similarly, from a behavioural standpoint, international students who utilize social media are more comfortable connecting with people from the host country, which may make it simpler for them to participate in school activities and, as a result, increase their level of student engagement. The use of social networking sites such as WeChat from the host nation may also be connected with an improved comprehension of lecture content and involvement in school activities thereby improving student engagement. It is in line with the findings of other experiential research, such as [32] which observed a positive connotation between acculturation and student engagement in a university setting. Regarding hypothesis , the study also discovers that there is a direct positive link between students' use of social media and their engagement in school activities. Student engagement has been recognized as one of the most essential markers for assessing the quality of an international educational experience. According to our findings, the use of social media can encourage international students to become more involved in their academics and extracurricular activities. These findings are congruent with the findings of previous studies conducted in different circumstances. Example: According to [32], social media improves students' communication, cooperation, and relationship-building skills when they interact with others. [4] asserted that the online setting offers access to a variety of resources and educational tools that can help international students learn more effectively and efficiently. We also found that self-identification plays a moderating role in the connection between international students' use of social media and their acculturation. This is notable since some earlier study has seen acculturation as a one-dimensional process. This emphasizes the notion that overseas students are less likely to adopt another country's media or connect with those from another culture despite their deep connection to their culture [35]. This suggests that having a strong ethnic identification does not inevitably result in less acclimation to another culture and therefore, people can be caught between two different cultures without sacrificing their cultural identities supporting the bidimensional acculturation [62]. These findings support [63]. conclusion that self-identification with the host country has little bearing on psychological well-being. Instead, sojourners' selfesteem can be boosted by maintaining and practising their cultural identity. This implies that when international students are comfortable with their own identity and culture, they are less concerned about cultural differences and tend to respect other cultures. Similar results are presented by [4] who observed that Chinese students in the United States with strong ethnic identities and who frequently communicate with family or friends from their ethnic background appear to combine excellent academic accomplishment with low stress. We believe that the findings of this research add to the four-acculturation framework proposed by [17]. , which emphasizes the importance of cultural balancing and the coexistence of two or more cultures in society [64]. --- Theoretical and practical implications Theoretically, this study broadens the scope of literature on acculturation and social media by providing a more in-depth picture of the relationship between various types of acculturation processes , social media, and student engagement. Most prior research has examined the relationship between social media use and the acculturation stress faced by international students during their transition to a new culture from a psychological standpoint. The study could contribute to acculturation theory because it examines the possible effects of social media use on the process of acculturation for international students in China, both psychologically and behaviorally. The findings of this study highlight the significant role that using social media plays among international students, as well as its ultimate impact on the students' acculturation process and academic performance while abroad. Practically, our results are quite relevant in the field of social media and higher education. This research sheds light on the experiences of international students and will assist school administrators in developing a solid universitystudent connection that will benefit both parties in the long run. Even while some studies have suggested that using social media to learn about a new culture may hinder immigrants' ability to acculturate to the new culture, our data reveal that the opposite is true among international students in China who use social media to learn about their new surroundings. University officials should not be concerned that international students with strong ethnic identities will struggle to integrate into the host culture. Instead, they should increase their investments and support these students. These investments and forms of support can include: giving international students a welcome package with stationery, gift cards to a local store, and a sim card so they can use their phone; using visual aids when teaching; encouraging them to use examples from their own cultures in classroom discussions; and provide them full support in language learning and event planning to foster interactions between international and indigenous students. Aside from that, there is a positive association between the use of social media by overseas students and their ability to participate in school-related activities. It is hoped that these findings would assist educational institutions, lecturers, and research supervisors to better recognize the benefits related to using social media to increase student participation in school activities. To put it another way, our study reveals that social media can be quite beneficial for international students in several ways. For example, social media can enable overseas students to collaborate more effectively with persons in the host country who share their research interests, thereby increasing their chances of participation in academia. In addition, the more connections international students made with people in the host country may result in greater information sharing. This information sharing may then result in increased participation in school activities. On a more general note, international students require programs that enable them to better integrate with students from the host culture, reducing the negative impacts of acculturative stress. Therefore, the following strategies and actions can be implemented. First, it is recommended that school educators ask their school administration for specialized materials and training to aid in the acculturation of international students. Second, some ethno-specific leisure activities are needed to better cater to people who are still in the acculturation process and those who do not wish to become acculturated. Faculty can help by clarifying classroom and academic standards through class discussions and course information and encouraging an open and tolerant classroom environment to embrace international student diversity. Third, universities should provide cultural programs like cultural tours, taichi , and a learning atmosphere that allows students to adapt to the university and Chinese culture through various experiences. International students should also have the opportunity to participate in voluntary social programmes and activities with the local Chinese community, like cultural sharing, counselling, and recreational activities where they can study Chinese culture with the Chinese, as it is an effective way to get social support from Chinese people. Fourth, by actively fostering the creation of broader social networks involving a wider range of nationalities, school counsellors or authorities can aid in the acculturation of international students. This could include assisting students in developing extra social skills and providing information on how to meet new individuals on campus, not just in the classroom but also outside the campus. Finally, employing teachers and school administrators from different countries may help promote international students' acculturation. These teachers may provide inspiration and support to new students who are still in the acculturation process. --- Conclusion, limitations and future studies There are several studies investigating the connection between social media, acculturation process and student engagement. However, most of these studies focus on Chinese students' experiences in UK and American universities and how their expectations were met or not. Given that recently China is increasingly becoming a recipient of international students to boost inward student mobility, in this research, we focus on international students in China by examining the influence of the use of modern social media on international students' acculturation and their engagement in school activities. But more importantly, the findings of the paper established the tentative arguments before the study revealed the following outcomes; direct connection between social media use and mental and behavioural outcomes, respectively, suggests that students' use of social media is positively related to students' psychological/mental acculturation. There is a significant linkage between psychological/mental acculturation and students' engagement. There is a significant link between behavioural acculturation and student engagement. The results for the hypothesis show that social media usage is positively linked with students' engagement. In our regression analysis, self-identification mediates the relationship between social media use and mental acculturation and behavioural acculturation. In other words, the outcomes suggest that the effect of social media on behavioural and psychological/mental acculturation could increase due to self-identification. The outcomes of this study are particularly useful for schools that host a large number of international students and that have a multi-national recruitment strategy. The findings of the study demonstrate the critical role played by social media among international students, as well as the ultimate impact it has on their studies. The study has several flaws which can be explored in future studies. First, though social media use is an important predictor of acculturation, this study did not investigate the influence of demographic variables like age, sex, personal traits, university program, and years of sojourning on acculturation. Another problem is that the participants' gender ratios were not evenly distributed. There were more males than females which means future research is should try to even the balance between gender ratios. Furthermore, because the survey only included a small number of international students in China, the study's sample may not be a true representation of the sample of the overall population of international students in China. Another issue is that the study treated all of the international students as if they were from the same country and culture, even though they came from diverse countries and cultures. In other words, the current study was unable to distinguish between participants' cultural backgrounds as a factor that influences the study's outcomes. It would have been preferable to divide the participants into groups based on their country of origin before conducting the analysis. The findings inspire future studies to recruit a larger number of participants while also taking into consideration the aforementioned issue to re-examine the effect of social media on acculturation and student engagement in the analyses. --- This is the published link and DOI to the data repository site. https://data. mendeley.com/datasets/c27jz8y4wx/1 DOI:10. 17632/c27jz8y4wx.1.

Despite the widespread use of modern social media, relatively less is known about the impact of social media on the acculturation processes of international students in China and their engagement in school activities. Accordingly, this research intends to assess the influence of social media usage while answering questions such as how using social media can improve international students' acculturation process from both psychological/mental and behavioural standpoints, as well as whether international students' acculturation promotes students' engagement in school activities, among other questions. The role of self-identification in mediating the connection between social media usage and international students' acculturation is also investigated. Primary data were gathered from 354 international students studying at various universities around China. The results show that international students use of social media improves their acculturation process and engagement in school activities through information sharing, establishing contacts, and entertainment. The study's limitations and future directions are also highlighted.

COVID-19 and social psychological contributions During the COVID-19 pandemic, social psychology entered the popular imagination in new ways. In many countries, the media asked social psychologists to explain how people's behaviour can be influenced, and the impact of policy decisions and leadership on adherence to COVID-19 rules . Various social psychologists worked to inform public debate and show the utility of behavioural science for ameliorating the pandemic. This increased visibility of social psychology is facilitated by three main issues. First, long-established bodies of work on social identities and group processes , communication and language and health and community were available to be mobilised and made accessible to the media. Second, many social psychologists are committed to making a positive difference to their societies . Third, the conditions created by, and reactions to, COVID-19 afforded possibilities of collaboration, including across areas of social and health psychology that are sometimes critical of each other . Kazak suggests that "psychology as a hub science can partner with others to provide scientifically based information about human behavior and the biopsychosocial concomitants of COVID-19." COVID-19 also facilitated media interviews and speedy publishing of preprints and texts in recognition that COVID-19 is a turning point for social psychology as well as for global populations. It quickly became apparent that the pandemic exposed and exacerbated global and national inequalities linked with racialisation, ethnicization, social class, gender, nation, whether or not occupations involve dealing with the public and housing . Dramatic and traumatic as it is, COVID-19 does not stand alone as a globally transformational event. Smith et al., document several events that have produced rapid societal change. They argue that these require the conceptualisation of how societal-level, macro events are related to more proximal psychological processes. As they show, social psychology research frequently focuses on the micro or meso level, without considering the macro level. The insight that all levels operate simultaneously and that there are innovative methods for studying them, is central to the discussion below. --- Theoretical underpinning: Racialisation, intersectionality and hauntology Racialisation is a term coined by Frantz Fanon but made popular by Omi & Winant and Miles . It illuminates the ways in which 'race', which many people think of as simply black, white, and Asian, for example, is not fixed, but changes over time and is different in different situations. It is made socially significant and meaningful rather than being fixed or natural. It is a socially constructed, relational process in which everybody has a racialized, and an ethnicised position. It involves exclusions and inclusions, in that people who are less powerful are often excluded from participation in meaningful processes in society. It is also a socially constructed process of subjectification in which people come to understand themselves as positioned in society and having particular identities . Intersectionality, a term coined by Kimberlé Crenshaw has gained popularity partly because it allows an engagement with the complexity and multi-layered nature of everyday life and social categories . Its central point is that people are always simultaneously positioned in many social categories so that there is no essence to any category. A focus on any one social category can, therefore, only be understood in the context of other categories and of differences, as well as commonalities, within groups and between them. Gender and sexuality, for example, are also class-based and racialised social relations . Any social category is, therefore, always decentred by other social categories and stands in power relations to other categories. Collins & Bilge and Collins suggest that it is a critical social theory with six core ideas of intersectionality on which methodological strategies for decolonizing knowledge has been built: social inequality, epistemic power relations, relationality, the significance of historical and social contexts, complexity, and social justice. From an intersectional perspective, the understanding of which categories are being evoked in any social situation and how social locations, emotional attachments, positioning in relation to nations and power relations make a difference is not necessarily self-evident. Nor can it be assumed that racialised relations are only playing a part in interactions when identified as such. Intersectionality is thus a heuristic reminder that all categories are associated with power relations and cannot be neutral . It is an onto-epistemological theory in that it is predicated on the notion of knowledge as situated and so differentiated since people occupy different social positions that change over time and across circumstances. At the same time, the nature of being is that people are multiple, constructed through difference, complex, nonessentialist and subject to change. Notions, of knowledge as partial, dynamic, and situated and people as multiple, complex and in process are central to intersectionality and fit with many current ways of theorising identities and social positions. It is particularly valuable to thinking about the humanising of racialisation in that it allows a focus on hierarchical and horizontal power relations and contradictions within social relations as well as considerations of embodied materiality. As intersectionality has burgeoned, it has been drawn on in many disciplines, including psychology . However, Settles et al., suggest that it has systematically been 'epistemically excluded' in psychology. The third theoretical concept employed in this paper is Derrida's concept of 'hauntology'. This explains that history remains part of the present. Colin Davis suggests that "undisclosed traumas of previous generations …disturb the lives of their descendants even and especially if they know nothing about their distant causes." A contemporary indication of how this may happen relates to a recent BBC documentary on Caribbean origin children, many of whom were, in the 1960s and 70s, misclassified as educationally subnormal by British educational psychologists and put into schools for the 'educationally subnormal'. The day after that program aired, the Division of Educational and Child Psychology in the UK apologised for this epistemically violent history , which has been marked by repeated attempts to prove that black people are less intelligent than white people . In terms of the haunting of undisclosed histories, some of those children, now adult, spoke for the first time about the trauma that had led them to keep that aspect of their histories secret and how they had compensated for it by, for example, spending long periods as students and, in consequence, having less time to spend with their children than they otherwise would. Their history, therefore, affected their children in this, and probably many other implicit ways. This exemplifies the ways in which "collective histories inform our current individual modes of being and the traces of the past create the contours of present life, and constrain our imaginative future projections" . Although such hauntings are frequently disquieting, they can be spurs to consciousness and calls for future political action. . The UK sociologist Paul Gilroy suggests that 'postcolonial melancholia' characterises postslavery nations and results from the denial of the impact of slavery and the losses that would be entailed by acknowledging them. Similarly, the US political scientist and psychoanalyst, Jane Flax , similarly suggests that denial of the impact of slavery leads to 'sometimes paralyzing wishes to magically erase the living past rather than engage in the arduous processes of realistically facing its effects and constructing practices to ameliorate them'. Once hauntings, absences and erasures make their ghostly appearance, repression no longer works, and something must be done The relatively new social psychological notion of collective victimhood chimes to some extent with the conceptualisation of hauntology. It refers to the psychological experience and consequences of the harm that arises from victimisation and includes affect, behaviours and cognitions that produce collective identities and intergroup relations . The theorisation of collective victimhood constructs it as complex, multi-layered, and impactful experiences. However, hauntology does not make victimhood central, focusing instead on the continued presence of history in the present. The section below considers current social psychological approaches to racialisation. --- Current social psychological approaches to racialisation Social Identity Theory, which underpinned much behavioural science response to COVID-19, was developed from a desire to understand how the Holocaust could happen. Yet, Tajfel, who was central to its inception preferred to approach racism and its relationship with identities only indirectly , an approach that continues to be common . Partly as a result, much social psychological work has not kept pace with current understandings of social categories and processes of racialisation. Given social psychology's tradition of considering one social category at a time and doing so from outsider, researcher perspectives, it is not well placed for understanding the complexity of social inequalities. Some psychologists are, however, now arguing for engaging with complex social inequalities by taking intersectional approaches to psychological science . Social psychology itself is a multifaceted enterprise that is neither neutral nor necessarily committed to social justice. As members of society, social psychologists are differentiated by whether they take for granted implicit social norms that perpetuate, rather than challenge, social inequities. Gergen , challenged the presumptions that social psychology is a cumulative or progressive science; that it is value neutral and that social psychologists are objectively distanced from the research participants they observe. Equally, in the landmark text, Changing the Subject , showed how social psychological concepts and research frequently served to maintain the political status quo and already-existing power relations. Reicher , however, argues that criticisms of Social Identity Theory frequently traduce it by failing to recognise its focus on the importance of context or on the multiplicity of the social categories to which everyone belongs. Critiques such as these inspired generations of social psychologists to do social psychology differently and to put time and effort into demonstrating personal commitments to social justice. More recently, two areas of work illustrate ways in which interpretations of social psychological findings on difference, including racialisation, continue to be contested. First is the long-established area examining whether, and if so how, contact between groups reduces prejudice . While work in this area has produced sophisticated methodology and many insights, as Dixon et al., point out, it repeatedly binarises the social psychology of intergroup relations as about 'us' and 'them', where 'us' in studies of racialisation are white people and the focus is on how white people think and feel about black people. As they argue, it is important to move beyond such binary thinking. In addition, the taken-for-grantedness of insiderness as white, perpetuates the outsiderness and exclusion of black people. In itself, therefore, contact cannot be the panacea for racism and leaves power relations and racialised hierarchies unaddressed in, for example, contact between Protestants and Catholics in Northern Ireland . The second example concerns the drawing into psychology of the concept of 'microaggressions' to study the everyday, apparently minor ways in which racism is perpetuated in interactions. This area is one that focuses on the microinsults, microinvalidations, and micro assaults to which people from minoritised groups are routinely subjected . This area has the potential to 'humanise' people from minoritised groups, taking seriously their experiences, viewpoints and feelings and refusing the routine normalising of majority group practices. Freeman and Stewart , however, suggest that many accounts of microaggressions are 'act-based' rather than 'harms-based' and that the focus needs to be shifted from those who are the micro aggressors, to the harms done to those who are aggressed. This is a hotly resisted area both outside, and within, psychology with some commentators eschewing the existence of microaggressions. Syed suggests that disagreements arise because diametrically opposed conceptualisations of society are at stake, one that views society as racist and the other that assumes it is equitable. Syed suggests that, until this is openly discussed, these positions cannot be reconciled. It is, however, questionable whether debate would lead to reconciliation, since debates are not neutral, but motivated by epistemological, ontological, and axiological commitments. Social psychologists have been at the forefront of opposition to publications that explicitly reproduce racism by suggesting that there are inherent black-white differences that prove black people's inferiority . That personal commitment characterises many psychologists, although others have complained that psychology is becoming too political in being avowedly anti-racist, . Personal commitment does not, however, necessarily translate into theory and professional and research practices that promote social justice. This is evidenced by two apologies in 2021 from psychological bodies: the UK Division of Educational and Child Psychology and the American Psychological Association resolution adopted by their Council of Representatives, evocatively titled 'Apology to People of Color for APA's Role in Promoting, Perpetuating, and Failing to Challenge Racism, Racial Discrimination, and Human Hierarchy in U.S'. The global eruptions that followed the murder of George Floyd and the surprise this engendered in many quarters, illuminated the fact that psychology has had little to say about what it means to be black in many societies. A major reason for this is the lack of explicit recognition that black people's positioning in societies is inextricably linked with histories of racism and colonialism. Thus, while since George Floyd's murder, there has been increasing acknowledgement that many Minority World societies are affluent because of their histories of enslavement and colonialism, blackness remains an outsider position in a discipline disproportionally populated by white psychologists. Since many white social psychologists , see themselves as implacably opposed to racism, it is difficult for them to conceive that they may fail to see how social psychology perpetuates the 'normalised absence' and 'pathologised presence' of black people in social psychology, as in society . The African American sociologist W.E.B. Dubois's asked, 'How does it feel to be a problem'? which is consonant with a 'pathologised presence' approach. The fact that there are consequences to leaving blackness and black people invisible was illuminated after the resurgence of Black Lives Matter, when various UK institutions, including universities, "sympathised with their BAME staff" and were surprised by angry reactions from black staff who pointed out that this was either a gross failure to recognise, or denial of, the specificities of antiblack racism and the experiences and feelings of black people about these. The conjunction of the visibility of the murder of George Floyd occurring and statistical evidence that black people were more likely than white people to die of COVID-19 led to an outpouring of testimonies about the pervasiveness and painfulness of experiences of racism. These gained traction as white young people joined demonstrations and various media published first person accounts. As a result, some people learned for the first time that multiple forms of racism are routine in the everyday lives of many black people but are generally left silent in the expectation of either not being believed or of having to deal with rebuttals or the explaining away of their experiences. The ubiquity of such testimonies led many people to argue that it is now no longer possible to deny the pervasiveness of racism. This, however, is questionable, since many studies find that racism is routinely denied and race and racism are routinely avoided in interactions. As Sambaraju & McVittie suggest, "acknowledging that racism organizes much of our social lives… may raise various levels of discomfort" so there continues to be a dilemma for social psychologists between the visibility of race and the studied invisibility of racism. For social psychologists to address that dilemma requires a rethinking of ontology and epistemology, the nature of society, racialised social relations and intersectional social differences. It has long been recognised in social constructionist and poststructuralist work as well as positioning theory, that we are all socially located and that this makes a difference to worldviews and experiences. Some social psychologists, particularly taking feminist perspectives also reflexively situate themselves in their research. This, however, is frequently superficial, with simply naming select social categories being treated as equivalent to analysing positioning and often being essentialised. To some extent this results from the space that would be taken up with discussion of the researcher in relatively short research papers in what sometimes seems to be self-indulgent ways. Yet, just as we recognise that our participants' positioning impacts on their lives and the experiences they report, so it is equally important to consider how researchers' histories impact on the analytic insights they produce. Methodologically, there is a need for methods that recognize that psychologists are temporally, geographically, and institutionally located in institutions and nations that have particular ways of constructing people. --- Humanising blackness Following the Los Angeles riots in 1992, the African Caribbean novelist, dramatist and philosopher Sylvia Wynter wrote an open letter to her colleagues at Stanford University . In "No Humans Involved: An Open Letter to My Colleagues", she argued that academia discursively ignores, implicitly legitimates, and sustains social hierarchies where humanness and north Americanness are white and middle class. Yet where did this system of classification come from? One that was held both by the officers involved in this specific case of the routine "nigger breaking" of Black males, as well as by the mainly white, middle-class suburban Simi Valley jurors? Most of all, and this is the point of my letter to you, why should the classifying acronym NHI [No Humans Involved], with its reflex anti-Black male behaviour-prescriptions, have been so actively held and deployed by the judicial officers of Los Angeles, and therefore by "the brightest and the best" graduates of both the professional and non-professional schools of the university system of the United States? By those whom we ourselves would have educated? How did they come to conceive of what it means to be both human and North American in the kinds of terms within whose logic the jobless and usually school dropout/push-out category of young Black males can be perceived, and therefore behaved towards, only as the Lack of the human, the Conceptual Other to being North American? The same way, as Zygmunt Bauman has pointed out, that all Germans of Jewish descent were made into and behaved towards as the Conceptual Other to German identity in its then PanAryan and Nazi form. If, as Ralph Ellison alerted us to in his The Invisible Man, we see each other only through the "inner eyes" with which we look with our physical eyes upon reality, the question we must confront in the wake of the Rodney King Event becomes: What is our responsibility for the making of those "inner eyes?" . --- Thirty years after Wynter's open letter to her colleagues, the reactions to the murder of George Floyd indicate that academia has not systematically risen to Wynter's challenge. There is a failure to recognise the implications of the ways in which racialisation and racisms continue to permeate everyday life in ways that construct black people as 'less than human'. Wynter, draws on Frantz Fanon's notion of sociogenesis to recognise that it is not produced in the body, but in the 'social imaginary grounded on colonial differences... I am who I am in relation to the other who sees me as such; and, in a society structured upon racial hierarchies, becoming black is bound up with being perceived as black by a white person' . According to this formulation, selfunderstanding is always permeated and fractured by colonial histories of racialisation and the denial that conceptualisations of humanness is inextricably linked with such racial logics. The title of this paper, 'Humanising racialisation' draws attention to the ways in which social psychological attention to racialisation continues the twin processes of ignoring and/or including black people in ways that implicitly construct them as outside humanity. The point that racialisation is both about the nature of being and about unequal, racialised, but unacknowledged social constructions of people finds some support in different areas of social psychological literature. The literature on unconscious, or implicit bias, built on research in Social Identity Theory, Self-Categorization Approach, entitativity and other social psychological work focuses on those who perpetuate discrimination and is designed to give insights into how to eradicate it. .This is not easy to do in that, as West and Eaton found in two studies on racism, participants can overestimate their levels of racial and gender-based egalitarianism, particularly those rated as most prejudiced These processes fit with Wetherell and Potter's findings from discursive psychology research, that everyday racialised categories are complex and contradictory and are constructed in ways that routinely mask the hierarchies they produce. Wynter's sociogenic formulation raises three further issues relevant to the future of social psychology. First, Wynter addresses the question what it means to be human by recognising the simultaneity of race, class, gender, amongst other categories. This underlines the issue of intersectionality and its place in researching racism and racialised identities in social psychology. Second, is the question of how historical legacies are represented and recognised in thinking about group processes and social identities. Michael Billig reminds us that although Henri Tajfel, whose collaborative work produced Social Identity Theory was an experimental social psychologist, he "insisted upon the historical nature of social psychological phenomena. In general terms, rather than in specific detail, his arguments about the importance of historical understanding resemble those of Kenneth Gergen . Also, as will be seen, Tajfel insisted on the limitation of social psychology. In his view, social psychology on its own was insufficient for understanding important social phenomena; nor could social psychology ever hope to produce complete explanations." This is a salutary reminder that while the experiments that produced general understandings of intergroup relations can be criticised for reducing the content of intergroup conflict to relatively trivial differences and implying that intergroup conflict is inevitable , Tajfel having himself survived the Holocaust, did not consider that Social identity Theory could provide a full explanation of the racist mass slaughter of the Holocaust. Equally, while as we have seen above, current work on group processes grounded in Social Identity Theory has been invaluable to addressing the issues raised by the COVID-19 pandemic, they are not designed to address the sociogenic issues Wynter and Fanon raised . The importance of an intersectional approach is illustrated by the recent testimonies and analyses of Tajfel's repeated sexual harassment of women over whom he had power, which is only being made public almost forty years after his death . This demonstrates how people who make valuable contributions to social justice in one way can use their power to oppress other groups in unacknowledged ways. While Tajfel was solely responsible for his harassing behaviour, the general point for this paper is that it is frequently difficult to see the ways in which we ourselves are privileged and perpetuate the oppression of others . In relation to social class, Valerie Walkerdine brings together intersectional sensibilities and histories to suggest that "We must understand the complex affective histories and entanglements as they relate to lives across the class divide-urban, rural, middle and working class, black and white, gay and straight, other possibilities and sexualities." The third issue relevant to the future of social psychology concerns the question of how to analyse lived experience. The plethora of testimonies on experiences of racisms of different kinds are of course important. However, the telling of stories of experience is historically contingent and sociogenically produced. Yet, current testimonies are frequently individualised and transformed from being simultaneously about psychological and social issues to decontextualised stories to be consumed for edification, marvelled at, and pitied without any commitment to political change or to shifting disciplinary understandings. Paradoxically, this reinforces the disempowerment of those telling the stories since their understanding of their experiences ceases to be within their control and can be treated as tokenist. Perhaps even worse, it may be that such testimonies first elicit responses of care and then 'moves us into indifference' . This is far removed from Fanon's and Wynter's focus on lived experience in order to theorise the imbrication of the personal and sociogenic . This problem was much written about during the period when Racism Awareness Training was common in the UK and USA . Since it is in the everyday that understandings of lived experience are produced, we need to analyse everyday complexities. In a now classic, and highly influential publication, Michael Billig coined the phrase, 'banal nationalism' to show the power of everyday, endemic nationalism that is largely unexamined and unremarked, but buttresses more extreme, dangerous forms of nationalisms. In a parallel way, we need to think about the banality of racism and its everyday manifestations and invisibilities. Cheryan & Monin draw on self-categorization theory in bringing together nation and racism by considering the mundane question to visible minorities in many countries , 'Where are you really from?' This denies them belonging to the nation, constitutes identity denial and impels Asian Americans to react by showing that they have national cultural knowledge and participate in practices generally considered American. From the viewpoint of universities, Osbourne, Barnett & Blackwood researched the experiences of seventeen black students . They point out that the hypervisibility of black students in those contexts enables their white colleagues to produce cultures of exclusion that are cognitively and emotionally salient for black students. They conclude that, in seeking to increase diversity, universities need to think about the attitudes, knowledge and experience that white students bring to university with them since these impact on black students. As Kelly et al., point out in relation to professional psychologists, white people are significantly less likely than black people to perceive individual and systemic racism, partly because of their relative ignorance of historical racism. Equally, the ways in which teaching is done in universities also matters for the disruption or reproduction of everyday racism. Adams et al., showed in an experimental study, that 'standard pedagogy' may reproduce a modern form of scientific racism by understating the significance of racist oppression. Just as the conjunctions of COVID-19, Black Lives Matter and climate change have exacerbated already-existing inequalities and fuelled action for social justice, so they have intensified alreadyexisting attempts to rethink epistemology, ontology, and methodology. This unforeseen conjunction of events and protests therefore has the potential to produce transformative reimaginings of society and of political possibilities. In this process, there is clearly a new appetite, across racialised and ethnicised divides, to listen and to try to understand the nature of everyday racism and its insidious, damaging effects, recognising that this is not comprehensible without historicising it. Writing from South Africa, and a feminist psychological perspective, Boonzaier & van Niekerk argued that psychology has to be decolonised. In South Africa in particular… there has been longstanding recognition of the role of psychology in enabling and justifying institutionalised racism through apartheid... Given this history, as well as the contemporary ways in which knowledge production within psychology continues to pathologise those with long histories of suffering…, there is little argument against the idea that psychology must indeed be decolonized. Many social psychologists acknowledge that social psychology requires an infusion of new thinking that includes engagement with histories, interdisciplinary insights and disciplinary commitment to a social justice axiology . As is clear from social psychological work on group processes, however, groups are socially produced in different ways, and everybody is intersectionally positioned in several groups at once, so that in any context, some group memberships are given recognition and/or make a difference and others do not. The ways in which groups are brought into being in local contexts are themselves materially produced and sociohistorically constructed. The final main section below discusses concepts that facilitate the taking serious of the historical in the everyday and the ways in which people are always simultaneously positioned in multiple groups. It suggests the sorts of questions these theoretical perspectives might enable social psychology to answer in humanising racialisation within social psychology. "This stuff came up that I'd suppressed over all these years" The example below comes not from a research interview, but from a journalistic interview that elicits testimony from a well-known mixed-parentage racing driver, Lewis Hamilton, interviewed by the journalist and professor of Sociology, Gary Younge. His narrative illuminates the everyday relevance of the humanising of racialisation for social psychology. It is intended to provide an indication of how narratives may illuminate the issues central to the humanising of racialisation, rather than as a template or ideal type. "But in the past year, off the track Hamilton has started to find a voice about his racial identity. He has been taking a knee; raising a clenched fist. Long dormant concerns about racism and discrimination have been rudely awakened following the Black Lives Matter uprisings. In the process, Hamilton has transformed the way he sees himself: from a compliant go-with-the-flow character to a change agent who is determined to make waves. He has shaped the way others see him too, going from an inoffensive, if gaffe-prone, socialite focused only on his sport, to a politically aware role model conscious of his wider cultural significance. Now, he is about to take on the sport that brought him fortune and fame, ... "It's been a long journey for Hamilton, and there have been bumps along the road. "I'd be in Newcastle and people would shout, 'Go back to your country,'" he says. "Or in Spain, in 2008, when people painted themselves black and put on wigs, and were really mocking my family. And I remember the sport not saying anything about it."/…/ "Last year, that attitude changed. Before the Austrian Grand Prix, just a month after George Floyd's murder, Formula One's only ever Black driver donned a Black Lives Matter Tshirt and took the knee. When some drivers refused to follow his lead, he warned them that "silence is complicit". In the end, they all wore End Racism T-shirts and 14 drivers joined him in the gesture, while six stood behind. /…/ At the same race, Formula One, which controls the cameras broadcasting the event, cut away from the moment some of the drivers took the knee, to instead show Red Bull skydivers dropping from the sky. It was as though a dam broke. "This wrath of emotions came up and I couldn't contain myself," he says, recalling this profoundly emotional moment in a matter-of-fact way. "I was in tears. And this stuff came up that I'd suppressed over all these years. And it was so powerful and sad and also releasing. And I thought, 'I can't stay quiet. I need to speak out because there are people experiencing what I'm experiencing, or 10 times worse. Or 100 times worse. And they need me right now.' And so when I did speak out, that was me letting the Black community know: 'I hear you and I stand with you". There are numerous points of analytic interest in Hamilton's account in terms of humanizing racialisation. Five of these are immediately apparent. First, his narrative identifies a key trigger that acts as a turning point, Black Lives Matter. He claims agency in transforming himself from someone who recognised that he was subject to racism but made no public response to actively symbolising Black Lives Matter by 'taking the knee'. Many testimonies following George Floyd's murder discuss it as a turning point . Winters coined the term 'black fatigue' to explain George Floyd's murder as a 'tipping point' because 'it is physically, mentally, and emotionally draining to continue to experience inequities and even atrocities day after day when justice, equity, and fairness are purportedly legislated rights of all citizens of these United States of America.' This shift, reported by Hamilton, alerts us to differences between people who are apparently in the same racialised categories and disrupts essentialist notions that, for example, black people all have the same perspective on racism or are the group most subjected to racism. As Brah suggests, there are multiple formations of power embedded in the intersections between gender, race, class, ethnicity, and sexuality. As in demonstrations for Black Lives Matter, politics of connectivity are, according to Brah, produced in and through struggles over 'difference'. These can impel change. Second, the racism Hamilton mentions is both geographically specific , and evokes histories of exclusion. It is noteworthy that, while Hamilton is of mixed parentage, having a white mother, he identifies as black in this account. His account is recognizable because it is intertextual with common forms of racism. Third, Hamilton's racialised positioning intersects with his socioeconomic status, success, fame, power relations, and , gender. This enables him to demand concerted action from the other Formula One drivers, persuading them to take a stand against racism and is able to set up his own Commission because he has the wealth and inclination and is buttressed by Black Lives Matter. The other Formula One drivers constitute one of the many groups to which he belongs, despite their differences of racialisation. Fourth Hamilton evocatively invokes the political power of the affective: 'a wrath of emotions… couldn't contain myself… all this stuff came up that I had suppressed…' While he does not say so, those emotions are made legible because of the politically mobilising anger and upset that became manifest following George Floyd's murder. Fifth, he makes common cause with 'the Black community' and wants them to know that he does , despite being more protected by his positioning than many. His relationality here is also narrated as intergenerational and personal in that he goes on to speak about his hopes for his nephew and niece. Overall, the context Hamilton evokes is one in which personal and historical relations haunt social relations. The recent publication of 'Why public schoolboys like me and Boris Johnson aren't fit to run our country' also indicates how the positioning of the mainly white sons of some of the most affluent people in the UK includes taken for granted intersectional histories of privilege and othering. We laughed at anyone not like us, and the repertoire on repeat included gags about slaves and nuns and women hurdlers. One September, after a boy came back from a holiday in Australia, we had jokes about Aborigines. We internalised this poison like a vaccine, later making us insensitive as witnesses to all but the most vicious instances of discrimination. Everyone who was not us, a boy at a private boarding school from the late 70s to the early 80s, was beneath us. Obviously, we too were a minority, but of all the minorities we were the most important. Of course we were. We'd end up running the country. Beard's autobiographical writing argues that the emotional culture of boarding school produced a sense of entitlement, privilege, and untouchability. It gives insights into how racialisation is intersectional , taken for granted and passed down the generations. --- In conclusion The unexpected transformations produced by the conjunction of COVID-19, the murder of George Floyd and the resurgence of Black Lives Matter, highlight the importance of social psychological understandings and the need for a step change in theorisation of the social. This paper argues that, while social psychology has made vital contributions to the understanding of group processes and of racisms, there remains a need to humanize racialization by producing holistic analyses of black people's intersectional identities ). This requires analysis of how personal and historical understandings are sedimented into, and haunt, contemporary life in ways that frequently exclude black people from the characteristics that constitute humanness. It suggests that personal histories are linked with longer histories of, for example, enslavement and colonialism. Social psychologists, therefore, need to understand how events that become historically significant affect everyday practices and permeate social relations, constraining and/or facilitating possible futures. It is not, therefore, possible to understand the impact of histories on any one person without knowing the constellation of relations and effects that have been passed down the generations and sedimented into everyday practices. Research done on the intergenerational transmission of different forms of trauma, provides a good starting point for humanising racialised subjectivities . Equally important, we need to know more about how racism and racialised positioning is part of the present and the affects and relationality they enable. There are some creative psychosocial attempts to conduct such intergenerational analyses . A media interview is not designed to address particular research questions. However, the interview with Lewis Hamilton allows the analysis of one of the many public testimonies produced following the murder of George Floyd by black people and others subjected to racisms. It contributes to understanding of the conditions that produce group identifications and political action. It contributes to answering Wynter's question 'What is it to be human?' The paper argues that an intersectional approach is crucial to the project of humanizing racialisation and normalising black people's visibility. Taking an intersectional approach to Hamilton's newspaper interview allows us to examine his processes of meaning making. It shows the importance of recognising the contradictory nature of power relations and the contingent nature of the groups to which Hamilton belongs. His narrative contradicts notions that political action is only possible on the basis of a simplistic and reductionist identity politics since it is possible to see different groups with whom he identifies at different times and some of the identities that he does not refer to are clearly taken for granted in their intersections with racialisation. Hamilton's account confirms the plethora of social psychological work that shows that identities and the meanings people make from them are important to what they do . Hamilton's narrative illuminates processes by which crisis can accelerate change , forcing recognition of inequalities and fuelling: personal testimony, recognition of intersectional commonalities/differences and claims to humanity through resistance and political action. Hamilton constructs a process of change away from silence about racism to a politics of refusal of racism, renunciation of complicity with it and visible resistance to racism. George Floyd's murder and the resulting resurgence of Black Lives Matter demonstrates how old racialised relations haunt the present. In enabling a complex view of Hamilton and his actions, intersectional analyses facilitate the humanising of racialisation. It provides theoretical resources to make holistic sense of everyday events and the processes involved in identity change and the narrative construction of identities as well as claims to a different, liveable, future. All these issues are important to future social psychology research designed to promote social justice. --- Acknowledgements: The COVID-19 pandemic has led to collaborations and publications that have rejuvenated social psychology and highlighted the creative productivity of collaboration . Equally, the multiethnic protests, critical engagement and testimonies impelled by George Floyd's murder and the resurgence of Black Lives Matter have revitalised thinking about social psychological work and hopes for society. These developments, together with the explosion of creative new methods and foci in multiple disciplines, are helping to produce new ways of understanding racialisation. I am deeply indebted to the numerous people who have illuminated this landscape. Thanks to the editors, Stephen Gibson and Laura Smith, who were exemplary in being forbearing, encouraging and incisive. I am particularly grateful to Jill Bradbury and Nick Hopkins for their careful and detailed comments on the first draft of this paper. The final version is much better for their generous inputs.

The unexpected transformations produced by the conjunction of COVID-19, the murder of George Floyd and the resurgence of Black Lives Matter, highlight the importance of social psychological understandings and the need for a step change in theorisation of the social. This paper focuses on racialisation. It considers issues that social psychology needs to address in order to reduce inequalities and promote social justice. It draws on theoretical resources of intersectionality and hauntology to illuminate the ways in which social psychological research frequently makes black people visible in ways that exclude them from normative constructions. The final main part of the paper presents an analysis of an interview with the racing driver Lewis Hamilton to illuminate ways in which it is possible to humanise racialisation by giving recognition to the multiplicity and historical location of racialised positioning. The paper argues that, while social psychology has made vital contributions to the understanding of group processes and of racisms, there remains a need to humanize racialization by conducting holistic analyses of black people's (and others') intersectional identities.This paper calls for the humanising of racialised subjectivity in social psychology through engagement with processes of racialisation. It argues that social psychology needs to move beyond recognition that people's experiences are racialised to analysing the relational impacts of racism, both historically and in the present. The paper discusses the theoretical concepts of racialisation, racism, intersectionality and hauntology as central to a social psychology that promotes social equality. It argues that the unexpected conjunctions of the COVID-19 pandemic and the resurgence of Black Lives Matter have increased public understanding of the utility of social psychology and provoked new ways of thinking about racisms and racialisation. They, therefore, have the possibility to be positively transformational. The paper first considers the ways in which the COVID-19 pandemic has illuminated social psychological contributions to addressing social problems. There has not yet, however, been a sustained social psychology contribution to considering racisms and racialisation in the wake of George Floyd's 2020 murder in the US and the resurgence of Black Lives Matter in many countries. The second section introduces the concepts of intersectionality, racialisation and hauntology, which inform the paper before the third section considers current social psychological approaches to racialisation. The fourth section draws on Sylvia Wynter's interrogation of what it is to be human and how an engagement with those who have been cast out of the category as colonized and worthless can provide new understandings. The final main section analyses an example from a major UK newspaper to illustrate the ways in which intersectionality, racialization and hauntology are simultaneously central to the humanising of racialised subjectivities. The paper argues that social psychology needs to retheorize the social, to encompass intersectional complexity and to recognise that personal and social histories are part of the everyday and of imagined futures if it is to make a step change towards transforming racialisation and racisms for the better.

Introduction Injury were one of the most common causes of death in children and adolescents around the world [1][2][3]. In China, injury was ranked as the leading cause of accidental death of about 50,000 cases among the 0-14 age range [4]. Unintentional injury, as the most common component of injury, included drowning, accidental falls, road traffic injuries, poisoning and burns [1,5]. According to reports of the world health organization, the most common unintentional injuries were road traffic injuries, falls, burns, drowning and poisoning [2]. Previous studies determined that the rate of unintentional injury was from 11.34% to 13.86% and that the rate was higher in rural regions than in urban regions [6][7][8]. Additionally, unintentional injury in adolescents accounted for 60% of all deaths [9]. According to the Chinese comprehensive survey report of health-related risk behaviours in adolescents, we defined unintentional injury-related behaviour as some common risk behaviours could directly cause or activate the occurrence of unintentional injury in daily life [10]. To our knowledge, there was a close relationship between unintentional injury-related behaviours and unintentional injury. However, due to the variations of the economic level and living environment in different countries, potential risk behaviours were also diverse. Thus, investigations of possible risk behaviours would be discrepant in different countries. Combined with available literature, this research chose the common risk behaviours that led to unintentional injury in China, including unsafe swimming, jaywalking, illegal bicycling and not wearing seat-belts [10]. Considering the large sample size of our survey, it had the merit to recognise the true risk behaviours in the adolescent population, particularly in China. In this study, we performed a cross-sectional survey to identify the epidemic pattern of unintentional injury-related behaviours in adolescents in Ningbo Municipality, China, and analysed the potential influential factors attributed to these unsafe behaviours, which all could contribute to providing a scientific basis for developing an effective intervention strategy. --- Materials and Methods --- Study Design and Sampling Method This cross-sectional investigation was carried out from October to December in 2015 and covered 11 regions that included five districts and six counties in Ningbo city. The respondents were students enrolled in middle school, senior high school and vocational-technical school in Ningbo city. A two-phase stratified cluster random sampling method was used to select the participants, and self-administered questionnaires were applied to investigate unintentional injury-related behaviours. The first phase initially extracted all schools classified as middle school, senior high school and vocational-technical school in 11 regions in Ningbo city. All of these schools were sorted by geographic region in north to south orientation and west to east orientation. After that, three middle schools, two senior high schools and two vocational-technical schools were randomly selected in each county or district. The second phase extracted the classes from the included schools. Classes in the same school were sequenced from junior to senior by class number. Then, a simple random sampling method was applied to randomly select eight classes in each middle school, six in senior high school and six in vocational-technical school. Finally, there were 24 classes from middle school, 12 classes from senior high school and 12 classes from vocational-technical school in each district or county. All students from the selected classes participated in the questionnaire survey. Eventually, 10,806 students from the selected classes participated in the questionnaire survey, and 10,790 valid questionnaires were used to analyse the characteristics of unintentional injury-related behaviours. --- Ethics Statement All participants were informed of the research purpose and informed consent was obtained from each participant before carrying out the research. This study was approved by the Ethics Committee of the Ningbo Municipal Center for Disease Control and Prevention . --- Personnel Training and Questionnaire Survey The self-administered questionnaire survey was anonymous. Local Centres for Disease Control and Prevention initiated the unified training and organised the implementation of this study. Specifically, all respondents were convened in their classroom simultaneously and the questionnaires were completed independently without a school teacher being present. The researchers were onsite when the survey was conducted. Survey contents included basic characteristics and information regarding unsafe swimming, jaywalking, illegal bicycling and not wearing a seat belt. --- Indicators Unsafe swimming referred to having swam without safety measures in the past 12 months. Jaywalking referred to having run a red light in the past 12 months. Illegal bicycling referred to the behaviour of running a red light, taking over a vehicular road, riding double, turning without a gesture or retrogradation by bicycle in the past 12 months. Not wearing a seat belt was defined as not having worn a seat belt while sitting in the front seat of a car in the previous month. --- Statistical Analysis All data were recorded using EpiData 3.0 software by two individuals. A descriptive analysis was carried out to illustrate the distribution of characteristics among participants. A multiple logistic regression model was used to determine potential influential factors for unintentional injury-related behaviours. Ultimate results in tables were displayed as an odds ratio with its 95% confidence interval correspondingly. All available data were analysed by SPSS Version 13.0 . A two-sided p-value < 0.05 was considered as statistically significant. --- Results --- General Characteristics of All Participants We had recruited 10,806 students, of which 10,790 completed the questionnaires. There were 5485 male respondents and 5305 female respondents with a sex ratio of 1.03:1. Of all included respondents, 5824 respondents were from urban areas and the rest were from rural areas. According to the school types, 5272 were from middle school, 2948 from senior high school and 2570 from vocational-technical school. The details were shown in Table 1. --- Analysis of Single Unintentional Injury-Related Behaviours According to the definition of each indicator, the proportion of four unintentional injury-related behaviours involving unsafe swimming, jaywalking, illegal bicycling and not wearing a seat belt among respondents was 6.35%, 33.08%, 18.10% and 15.73%, respectively. Concerning the distinct gender, the proportion of unsafe swimming, jaywalking and illegal bicycling was significantly higher in male than in female. With regard to different area, the prevalence of not wearing a seat belt was lower in urban areas contrasted to rural areas, whereas the prevalence of the other three behaviours was significantly higher in urban areas compared to rural areas. In various school types, adolescents attending vocational-technical school tended to have the highest prevalence of all unintentional injury-related behaviours. These information was displayed in Table 2. --- Analysis of Combined Unintentional Injury-Related Behaviours All respondents with two potential risk behaviours accounted for 14.59%, of which 4.27% had three potential behaviours and 0.57% had four potential behaviours . Besides, the results suggested that the proportion of two combined behaviours was higher in male than in female, and were higher in urban areas than in rural areas. Regarding the diverse school types, students who attended vocational-technical school had the highest prevalence of two combined behaviours while those attending middle school had the lowest prevalence. Furthermore, findings from three combined related behaviours analysis were similar to the results of two combined behaviours analysis. However, a four combined behaviours analysis only identified the difference in gender that males had a higher prevalence. --- Analysis of the Potential Factors That Influenced Unintentional Injury-Related Behaviours In this study, multiple regression analysis was performed to identify the possible influential factors attributed to unintentional injury-related behaviours. The finding showed that male adolescents, living in an urban area and attending a vocational-technical school might contribute to the occurrence of four unintentional injury-related behaviours. Besides, married status of parents and father with a college degree or above were negatively associated with the adolescent's behaviour of not wearing a seat belt. These details were shown in Table 4. --- Discussion To the best of our knowledge, due to the discrepancies of economic level, living environment, study population and lifestyle in different countries, studies for the probable unintentional injury-related behaviours were inconsistent. For instance, in America and Thailand, unintentional injury-related behaviours commonly involved whether a bicycle helmet was worn or not, a seat belt was worn or not, driving when drunk or driving with other irrelevant behaviours [11,12]. In China, unintentional injury-related behaviours generally included unsafe swimming, jaywalking, illegal bicycling and not wearing a seat belt [10,[13][14][15]. In this study, we found that of the four evaluated behaviours, jaywalking had the highest proportion in Ningbo , followed by illegal bicycling, which were similar to previous studies conducted in Yinchuan City, Yuxi City and Shandong province [13][14][15]. With advances and development of social economics, this phenomenon could be explained by the sharp increase in the number of motor vehicles but without relevant improvements in other aspects such as related traffic safety education, suggesting that lack of traffic knowledge and weak safety awareness were still existed in student population. Besides, walking and bicycling in China were still the main trip mode for the student population. Thus, strengthening traffic safety education was an urgent need in juvenile groups. Although available literature reported that the proportion of drowning deaths in rural regions was significantly higher than in urban regions, our study revealed a lower proportion of unsafe swimming in rural regions than in urban regions [16,17]. Yet it was not a contradiction for the results above. Due to poor protective implements and limited resources in rural areas, the occurrence of drowning was more common in these areas. Thus, the government and local schools tended to focus on drowning safety education that resulted in a decline in the prevalence of unsafe swimming, which was also proved in a study conducted in the Pudong area of Shanghai City [18]. Another interesting finding of our study was the gender difference in unintentional injury-related behaviours. The occurrence of unsafe swimming, jaywalking and illegal bicycling was significantly higher in schoolboys than in schoolgirls. Moreover, schoolboys also showed a higher prevalence of combined unintentional injury-related behaviours. Compared to schoolgirls, the odds ratio and 95% CI of three and four combined unintentional injury-related behaviours were 3.08 and 8.95 , respectively. This might be associated with the boys' natural behaviour. Therefore, more concerns should be highlighted on schoolboys in these aspects. Additionally, vocational-technical school tended to have the highest prevalence for four unintentional injury-related behaviours, which might be caused by weaker safety education. Thus, more courses about safety education should be considered in vocational-technical school. Previous studies indicated that parents with high educational level could reduce the child's injury rate more obviously [19][20][21]. In our study, multiple logistic regression analysis only showed that father with a college degree or above was negatively correlated with the behaviour of not wearing a seat belt. This outcome might be explained by the phenomena that in China most drivers in the family were male. That is to say, a father with higher educational level might have higher traffic safety consciousness, being a role model for their children. However, our study had some limitations: Given we did not collect data of unintentional injury from respondents, the association between unintentional injury-related behaviour and potential injury was not further analysed; Partial indicators collected from the past year might cause an unavoidable recall bias in our design; Other potential factors including hiding some unsafe behaviours and disturbances from other classmates might also influence the prevalence of risk behaviours in our study; Considering the limited respondents with two or three unintentional injury-related behaviours simultaneously, we did not conduct the analysis in different combinations; In this study, we only performed our research in one city, the representativeness of which could be influenced by some specific factors as economics and education level. Therefore, some potential bias might disturb the extrapolation of our conclusions. --- Conclusions Overall, considering the diverse epidemiological characteristics of unintentional injury-related behaviours among adolescents, targeted interventions such as enhancing self-protection capabilities and strengthening safety consciousness in the adolescent population, should be implemented by family, school and other related departments to effectively lower the occurrence of unintentional injury-related behaviours. --- Author Contributions: Hui Li designed the study. Wei Feng collected, screened, analysed the data and drafted the manuscript. Qinghai Gong collected and screened the data. Kui Liu revised the manuscript. All authors read and approved the final manuscript. ---

The purpose of this study was to explore the epidemiological features of common unintentional injury-related behaviours and to identify possible factors that lead to these unsafe behaviours among adolescents. A representative sample of 10,806 students was recruited from 77 schools by using the two-stage stratified random sampling method. All participants took a self-administered questionnaires and data were analysed to estimate the prevalence of unintentional injury-related behaviours and to identify the influential factors for these behaviours. The prevalence of unsafe swimming, jaywalking, illegal bicycling and not wearing a seat belt was 6.35%, 33.08%, 18.10% and 15.73%, respectively. The proportion of students who had two, three or four unintentional injury-related behaviours was 14.59%, 4.27% and 0.57%, respectively. Multiple regression analysis showed that male adolescents, living in an urban area and attending a vocational-technical school might contribute to the occurrence of four unintentional injury-related behaviours. In addition, the marital status of parents and father with a college degree or above were negatively associated with the adolescent's behaviour of not wearing a seat belt. Considering diverse epidemiological characteristics of unintentional injury-related behaviours among adolescents, targeted interventions such as enhancing self-protection capabilities and strengthening safety consciousness by family, school and related departments should be implemented to lower the occurrence of unintentional injury-related behaviours.

With a focus on socially marginalised Greenlanders in Denmark, this study explores the significance of the concept of social su ering for the concept of total pain. Greenland is a former Danish colony and Greenlanders retain the right to Danish citizenship with all the benefits of access to the resources of Denmark as any other Danish citizen. However, Greenlanders are overrepresented amongst the most socially disadvantaged in Denmark. They have a disproportionately high risk of early death, often undiagnosed and untreated. This study reports on research conducted with socially marginalised Greenlanders and some of the professionals who work with them. It interrogates the concept of total pain as developed by Cicely Saunders, the founder of modern palliative care. Saunders noted that pain at the end-of-life was not adequately explained by symptoms of a disease process because it was more like a situation that engulfed every aspect of the patient and those close to them; it included physical, psychological, spiritual, and social dimensions. We agree with other scholars that the social dimension of the total pain experience is underexplored. By drawing on the theoretical and methodological lens of intersectionality, our work with marginalised Greenlanders has enabled us to describe the multiple and intersecting social forces that create social su ering for this group. This leads us to conclude that social su ering is not entirely an individual experience but a product of social harm and disadvantage, poverty, inequality, and the various legacies of colonialism, which combine to place some citizens in a harmed condition. Our findings also draw us into a discussion with the concept of total pain and its neglect of the socially constructed nature of social su ering. We conclude by indicating ways in which the concept of total pain can be informed by a more thoroughgoing concept of social su ering. We conclude, with others, that there is a problem of inequity in the way that --- Introduction For more than 6 years, the first author has researched palliative care in Danish hospices where she found that paying attention to total pain often helps people to die peacefully and also to 'live until they die' following the ideals of Cicely Saunders. However, patients were mainly middle-class white patients with cancer, a picture lately confirmed statistically . One, among several, wake up call that changed her focus from this cohort towards 'socially marginalised Greenlanders' was the comment from a palliative care practitioner describing her visit to a Greenlandic woman with cancer she described as marginalised: As soon as I go through the door I can feel in my own reaction that all my prejudices are confirmed. You know, you enter a flat she has only had for a year, but totally smoky and yellow and you stink when you leave. Extremely dirty and sparsely furnished. She sleeps on the sofa surrounded by cigarettes and bottles. She is so lonely and somehow trapped there, she has a sparse network I don't know if she was sober when we arrived, but we could only talk briefly and about simple issues. When I left I doubted whether she knew why we were there working with the most socially marginalised people is complex and I experience it as personally frustrating because you would like to come with something that can help them. Apart from the problem of equity as to who receives appropriate care at the end-of-life, the narrative also points to the complexity of managing care at the end-of-life. Cicely Saunders, a nurse, social worker, and medical doctor, established St Christopher, the first modern hospice in 1967. She founded the hospice movement and has been a major inspiration for the development of the discipline and the culture of what today is internationally recognised as palliative care . Palliative care is defined by WHO as relief of serious health-related suffering, be it physical, psychological, social, or spiritual . Saunders developed the concept of total pain inspired by what she learned from terminally ill cancer patients. Saunders explains that total pain 'is a situation rather than an event' and requires attention to the individual's physical, psychological, spiritual, and social pain. Clark ) points to how the concept challenged the dominant biomedical discourse and he gives an analogy of the sociological distinction between disease and illness. The concept of total pain focuses not only on somatic pain but also on the individual's narrative or subjective awareness . Even though 'the social' is included in the definition of total pain, this dimension is the most underexplored . In palliative care research, the social dimension is often conjoined as the 'psychosocial' which arguably does not capture the social dimension fully. We will take a wider and more structural approach to the social dimension as now recognised in the International Association of Hospice and Palliative Care definition of palliative care with its focus on reducing barriers to palliative care . Thereby, we challenge the idea that social pain refers only to the experience of interpersonal connexions between the dying individual and the social world. The impact of social inequality and inequity is an increasingly important focus for palliative care . Inequity is defined as the unfair and avoidable differences in health status . In line with what Richards describes as the 'equity turn in palliative and end of life care research' , we will argue that the social dimension of total pain requires attention to suffering produced by the social world through inequality, marginalisation, injustice, powerlessness, and other structural forces. Cicely Saunders developed total pain in response to a health system that had failed to manage the pain of dying people for whom nothing more could be done . However, in the same vein as Saunders ' original provocation, there is a class of people excluded from palliative care. It is recognised that people at the end-of-life who also face social and structural inequities are less likely to access palliative care . International research has identified some of the physical, psychological, social, and existential needs that characterise the end-of-life of socially marginalised people. Intersecting factors such as homelessness and substance use overlie issues such as fear of losing control, being a burden to others, dying alone, or remaining unidentified after death . There are also concerns that professional services do not regard end-oflife care as a priority, lack the competence to manage complex pain alongside substance dependence, and the challenges of caring for socially isolated people who lack insight into their condition and have little or no family support . Investigating the nature of 'social suffering' provokes a new challenge for the concept of total pain. Within the field of sociology, the connexion between suffering and social is theoretically well established. The concept of social suffering is used to describe how lived experiences of inequity and suppression can cause suffering. As such, the suffering is conditioned by the social that reaches beyond the subjective experience of suffering . Understanding social suffering thus involves directing the gaze towards the social conditions that cause suffering . As discussed by Wilkinson and Kleinman : 'The concern to understand human suffering in social terms brings critical debate both to the moral meaning of suffering and to the category of 'the social" . Including the perspective of social suffering in the approach to total pain invites an exploration of the social forces that affect total pain. In addition, a focus on social suffering is a springboard for questioning current understandings of total pain. With a focus on socially marginalised Greenlanders in Denmark, this study explores the significance of social suffering for total pain. In our study, socially marginalised Greenlanders are people whose pain must be explored in the shadow of a complex of social suffering understood as the sum of multiple structural factors including inequality, racism, and psychological and cultural trauma . Greenland was colonised by Denmark in the 18th century and not until 1953 was selfgovernance considered. In 2009, a new law established Greenland's autonomy within the Danish Commonwealth, recognising Greenlandic people as people in their own right, with the right of self-governance . Greenland is the largest island in the world with a widely dispersed population. However, the indigenous Greenlandic population is not one homogeneous group; there are three main language groups, namely, the dominant West-Greenlandic , the minority East-Greenlandic, and Thule languages . In 2021, 16,740 people born in Greenland lived in Denmark . Greenlanders are Danish citizens with the same right of access to healthcare services as all Danes . This includes a right to a translator if the doctor considers a translator necessary . Greenlanders in Denmark have been discussed as an invisible and neglected group despite their status as Danish citizens because this very status means that their ethnicity is not registered in national registers which renders systematic assessment of their social and economic conditions difficult . However, surveys and mappings reveal an overrepresentation of Greenlandic people among the socially marginalised people in Denmark . The average life expectancy of socially marginalised people, in general, is 19 years shorter than the average population . It is well established that the experience of colonialism, and the long-term effects of being colonised, causes inequalities in the health status of indigenous people . Subica and Link point to cultural trauma as an additional causal factor in health disparities . Several studies reveal that mainstream health services fail in meeting the needs of indigenous peoples and that indigenous people are often excluded and marginalised from mainstream health services . This picture is also true for socially marginalised Greenlanders in Denmark who are overrepresented amongst those experiencing poverty and homelessness and those who are problematic users of alcohol and drugs . Several Danish studies reveal structural barriers such as language, culture, ethnicity, and an inability to navigate Danish healthcare and welfare systems as factors in Greenlanders problematic access to welfare benefits . A national survey of marginalised Greenlanders found that 50% of the respondents were homeless compared to 23% of the general population of people who are socially marginalised. The survey also found that 19% reported hunger due to poverty and almost half of the respondents had attempted suicide and that their contact with health and social care was primarily under acute circumstances . The aim of this study is to explore how ethnicity, socioeconomic status, and illness and death culture intersect and impact the end-of-life experiences of socially marginalised Greenlanders in Denmark. Palliative care research is beginning to recognise the cultural aspects of illness and death culture, and how suppression or ignorance of cultural identity can influence suffering at the end-of-life . These aspects are still underexplored and there is no research on these aspects for Greenlanders in Denmark. Recognising that total pain unfolds in the relationship between the dying individual and their social world, the study also aims to shed light on how the concept of total pain can be informed by understanding social suffering. --- Methodology and theoretical framework Methodologically, the study is inspired by Layder's adaptive theory drawing on the idea that social reality is a complex, that cannot be captured by one explanatory principle, theoretical approach or methodology . The adaptive theory focuses on ties between the agency and social structures and the connexion between macro and micro levels of analysis in its approach to combining theory and empirical research. Adaptive theory transcends the gap between an inductive and deductive approach to research. The analysis is an ongoing process steered by theoretically informed 'orienting concepts' that guide without being deterministic . Initially, our study used total pain and intersectionality as orienting concepts to focus our empirical work. However, following empirically informed insights into the social aspects of total pain, we expanded these to include 'social suffering' , 'stigma' , and 'othering'. These orienting concepts are briefly introduced below. --- Intersectionality Intersectionality provides a theoretical and methodological framework for analysing mechanisms of social suffering at the end-of-life. Using intersectionality enables an exploration of how lived experience intersects with the processes of power, structural vulnerabilities, and social identities . As palliative care researchers increasingly recognise, the lens of intersectionality enables an exploration of the multiple levels of intersecting inequities that shape the end-of-life for marginalised people . To gain an understanding of social suffering as well as the role social structures play in marginalised Greenlanders lives, the methodology draws on Christensen and Jensen approach to intersectionality which employs life-storey narratives and the analysis of everyday-life . --- Total pain and social su ering The concepts of total pain and social suffering work as orienting concepts for exploring how the social constitution of total pain is rooted in social situations and conditioned by cultural circumstances. As Wilkinson notes, 'social worlds are inscribed upon the embodied experience of pain and there are many occasions where an individual's suffering should be taken as a manifestation of social structural oppression and/or collective experience of cultural trauma' . --- Othering and stigma The concepts of othering and stigma are also relevant to our research because they are contributors to social suffering through processes of differentiation. Jensen's definition of othering includes the notion of 'intersectional othering' , that othering is a discursive practise where powerful groups ascribe subordinate groups' problematic or inferior characteristics . Othering is relevant to intersectionality since both concepts touch upon several different forms of social differentiation, thereby leading to intersectional othering . Othering helps us to understand how power structures condition agency and to reflect on how historical symbolic meanings frame the possibilities for negotiating identity . Therefore, the theoretical framework of intersectional othering allows us to explore how power structures and culture such as illness and death culture, socioeconomic status, and ethnicity intersect and influence total pain at the end-of-life for Greenlanders in Denmark. To understand how socially marginalised Greenlanders experience stigma in everyday life, we draw on Link and Phelan's theory of stigma. According to Link and Phelan, the four components that can produce stigma are: labelling, stereotyping, separation, and status loss and discrimination, when all four components are present in a power situation that permits them, stigma occurs . Discrimination can occur in two ways. First, it can exist as individual discrimination, where a labelled person can experience social inequalities in life circumstances and life chances. Second, it can occur as structural discrimination, where institutional practises directly or indirectly exclude labelled persons . As Link and Phelan observe, people are often socialised to absorb cultural stereotypes and therefore reproduce them . Moreover, it takes social, economic, and political power to stigmatise, and the stigma that people experience is shaped by the individual's power to resist stigma . The outcomes of stigma are different for different people, but stigma can lead to lower self-esteem, reduced access to social and healthcare services, and have an impact on life chances . Combined with intersectionality and othering, this stigma perspective allows us to grasp stigmatisation both on a structural and individual level. Hence, having a devalued status in the wider society can lead to stigmatisation enhancing social suffering at the end-of-life . --- Methods Adopting an ethnographic approach enabled us to become immersed in the field and explore the social circumstances in which marginalised Greenlanders live and experience social suffering . This approach involved studying peoples' actions and accounts by participating in people's daily lives for an extended period of time . Fieldwork took place in two cities in Denmark and involved participant observation and semi-structured interviews. VG conducted fieldwork over 18 months and was the responsible contact in all settings. She accessed the field in collaboration with local social services, NGOs, the Danish Cancer Society, and Greenlandic Houses . Practitioners from these organisations acted as gatekeepers to the field. Participant observation focused on gaining knowledge of implicit beliefs, norms, and behaviours within the group . VG conducted participant observation in a social care-home , a social accommodation for people who need support from a social worker to manage everyday life , plus informal conversations with and observation of marginalised Greenlanders on the streets. This involved participating in and observing daily life such as meals, drinking, conversations, and professional care situations. For over a month, TP conducted participant observation by volunteering in two NGO drop-in centres that provide support for vulnerable peoples' basic needs plus social support. MBA conducted participant observation by volunteering over a month in a social café in a Greenlandic House. Practitioners who were gatekeepers to lay participants were asked the 'surprise question' used in palliative care as a test for identifying patients nearing the last phase of life: 'Would you be surprised if the person died within a year?' In this way, we were able to determine whether to include people in our study. Eleven narrative interviews with lay participants and 13 semistructured interviews with practitioners were conducted guided by bespoke interview guides for each group. The two participant groups consisted of lay Greenlanders who, in professional jargon, are labelled socially marginalised due to their socioeconomic status and social functioning; and professionals and volunteers working in contexts in which Greenlanders with social problems live their everyday life . Participants' names and professions are changed, and ages are collated in groups. Table 1 contains anonymised descriptions of participants in the study. Eleven of the participants have participated in narrative interviews whereas two are described based on field notes and fieldwork. Table 2 contains anonymised descriptions of practitioners who have been interviewed. --- Accessing the field and ethical considerations A key challenge is that of identifying and gaining access to seriously ill participants. The usual routes through established social and healthcare systems are problematic; as some seriously ill Greenlanders are undiagnosed or diagnosed so late that they are too ill to participate. Language is also a barrier because many Greenlanders are not fluent in Danish or prefer to use their mother tongue to better express themselves. Finally, amongst our participants, there was a significant issue of mental capacity both to consent to and to participate actively in the study. The factors here include heavy alcohol and drug use, as well as the impact of multiple traumas, all of which add to the problems of capacity but are also Greenland's department for education and social a airs support four Greenlandic Houses in Denmark, providing cultural contact between Greenlanders and Danes. Social, legal, and educational counselling are provided along with social cafes targeted at vulnerable groups. Frontiers in Sociology frontiersin.org aspects of social suffering we seek to explore. Thus, the empirical objective to explore Greenlanders' experience of total pain is difficult to meet via the conventional institutional settings for endof-life . Therefore, the research is conducted within the material world in which Greenlanders live their everyday life . Being morally engaged in people's lives and being emotionally affected are necessary and inevitable aspects of being able to give voice to suffering . However, in practise, it is an ongoing ethical challenge to navigate between the roles of observer and carer. Whilst writing this, VG communicates with Kristian who is desperate about his economic situation, his progressing cancer, and finding the help he needs. Helping him might not only help him as an individual but also gives an insight into the complexity of suffering that can bring public attention to essential components of suffering that otherwise would go unnoticed . The potential blurring between research and personal involvement is an ever-present ethical risk that was navigated based on a pragmatic principle of trying to help people to get professional help. Several of our participants were not fluent in Danish. In three interviews, we used a translator; however, in most cases, this was not an option as it was difficult to coordinate or find a trusted translator. Translations of various kinds were problematic, as Gunaratnam observes, and differences other than language such as class can affect communication and the interpretation of meaning significantly. In a less literal way, translating participants' pain experience through cultural grammar matched the linguistic challenge and we hope that our interpretation is veridical . All participants gave their written or recorded verbal informed consent. The study was approved by the Research and Innovation Organisation, University of Southern Denmark. Journal no. 21/24661. --- Strategy of analysis Data consist of field notes and transcribed interview data. All field notes were handwritten, and interviews were audio recorded and then transcribed . The translation process from non-fluent Danish into comprehensible Danish and then into English has sometimes influenced the nuance of meaning or has lost some of the cultural framing which must be acknowledged as a consequence of this process. When data collection was finalised, all interview transcripts and field notes were coded in NVivo TM to provide provisional code labels associated with orienting concepts or provided empirically generated ideas . We then identified three main themes across data coded into the overarching theme of total pain. Afterward, we turned our focus to orienting concepts, which helped us to structure our analysis. Thus, the elaboration of the themes was a combination of inductive and deductive procedures that enabled us to elaborate on the concept of total pain . --- Analysis The analysis has three sections. The first explores how socioeconomic status and ethnicity intersect in the lives of the participants as a complex web of differentiation associated with othering and stigmatisation. This provides a sociocultural backdrop that contextualises the conditions for social suffering at the endof-life. Second, micro-level aspects of social suffering are explored by focusing on norms, symbolic meanings, and social interaction guiding the participants' interpretations and experiences of illness, dying, and death in their everyday life. Third, structural barriers to accessing and navigating within conventional institutional health and social care are explored in the light of individual and structural discrimination. To connect our readers to our participants, each theme is introduced by a participant narrative . The narratives also trouble the concept of social suffering. Living everyday life in a socially marginalised position involves resilience, and social suffering is not necessarily expressed as individual suffering. Finally, the narratives illustrate layers of complexity of social suffering, that social suffering intersects ethnicity, socioeconomic status, and illness and death culture. --- Othering and stigmatisation as part of everyday life Malik lived in social accommodation, with two rooms, a little kitchen, and a bathroom. In one room, there was a mattress with a dirty blanket. In the other, a few chairs and a little coffee table covered with beer cans and vodka bottles. He lit a candle on a beer can and said: 'I always light a candle if I am here on my own. Then I don't feel so lonely'. He grew up in Greenland with his parents, but his father died when he was a little boy. 'He died of heart failure. I ran after the ambulance and I cried and cried and cried'. His mother died of cancer when he was about 6 years old and he was sent to an orphanage in Denmark. Later, he went back to Greenland and lived with family.' I had a criminal youth; I also stole together with my family. I was the youngest and could get through the little window...we stole wine and beer and such things. I have been alcoholic since childhood; I was about six or seven'. ' And now I have told myself that I should stop drinking, I also smoke cannabis. I am old, I am disabled, and I am also tired of being such a 'sut' Some days ago I could hardly walk and almost called the ambulance. But I tell myself that as long as I can walk I have to walk'. Malik had four children with two women but is not in touch with them. He hopes to make it to 60. 'But if I die, so be it'. Malik said that when he was a child he had a near-to-death-experience and said: 'It is strange, it is completely empty up there and a lot of small lights. They say I have to go back to earth again, and I came back to life. I felt that as a punishment.' VG: 'What encourages you to get up in the morning?' Malik: 'Beer.' He also told that he has his route that includes two drop-in-centres and a bench from which he sells The Big Issue. He would like to end his life in Greenland but he does not know where he would live: 'But I have some friends I can live with'. He would like to get buried in Greenland. Half a year after the interview, Malik had become weaker. If he had the strength, he would still take his daily route. He lost his rollator but found a replacement somewhere. Malik's social worker said that: 'When he drinks, he is in an extremely poor state. Then if he needs money or if there are some days he doesn't drink he seems to revitalise a little. My intuition is that he knows it probably only goes in one direction.' In this section, we first identify characteristics of how social suffering manifests for our participants as compared to Danish socially marginalised people. Next, we describe a complex web of differentiation, intersecting ethnic and socioeconomic categories. This web operates interculturally, between Danish and Greenlandic cultures and within the group of Greenlanders. Some lay participants bear the markers of social marginalisation such as alcohol and drug use, poverty, homelessness or supported accommodation, and an everyday life outwith conventional society. However, intersecting these markers are characteristics associated with being a Greenlander. Practitioners across the field describe the significant role of past traumas experienced by 'their' Greenlanders that are more common and more extreme than in the equivalent Danish populations. One practitioner describes a complex pattern of trauma, abuse, poor social conditions, problematic social networks, and the impact of colonialism: V: How do the traumas manifest themselves? G: Well, I think they do it in the way they live. That it's so massive, the abuse and their complete inability to maintain anything of what the rest of us consider normal: Being in control of their economy, having healthy close relationships. Many of them have close relationships but they are connected to [substance] abuse. Many Greenlandic citizens say that you lose a network if you stop drinking. Many of them have lived a really hard life, also with physically hard labour. Also, they haven't had much money, also just to be homeless in Greenland! You get the feeling that there is a lot caused by their history in Greenland -that Denmark took over Greenland. Malik's storey demonstrates this complex history and patterns of trauma, but this is not only an individual storey, many participants also describe experiences of bad deaths such as murder, suicide, shooting accidents, and early death by cancer and TB. A Greenlandic practitioner who describes her childhood in Greenland as the epitome of a socially marginalised background tells how she watched a family member beating her aunt to death and says: 'I have stopped counting the number of people I used to know who have committed suicide'. Another Greenlandic practitioner describes how traumas experienced by some of 'her citizens' cause death anxiety: 'It's the background we carry with us. You become anxious of dying because you have experienced traumas and if somebody has died that way you think that you'll die in the same way.' Practitioners describe how they need to show consideration for the traumas Greenlanders carry with them and avoid opening wounds, for example, by not talking about difficult things and as an explanation for limited participation in everyday-life activities. One practitioner described a traumatised social accommodation client: 'Sometimes she gets these outbursts she throws things around, hops and dances, shouts and screams. And then, when you talk to her it's all about somebody who has died, drowned or some family issues.' Fieldwork also gave an insight into how a history of trauma was used as an explanation for communication barriers and some people's reluctance to participate in social life. VG's field notes from a social accommodation state: ' A woman has several times mentioned that she would like to talk to me. Today she says: 'My head is Greenlandic today. I can't find the words in Danish' and clearly indicates that she wants to be left on her own'. Some lay participants identify themselves with labels such as 'alcoholic' and 'cannabis-user'. Several practitioners comment on the extreme nature of these problems compared to similar Danish groups. As one practitioner comments: 'It simply explodes -four bottles of vodka a day. It's suicide even though it takes time. It's so self-destructive. We had somebody who was found with a blood alcohol level of 5 point something. It really is so extreme'. These accounts show an aspect of suffering that cannot be reduced to individual tragedies. They also illustrate socially conditioned lived experiences of injustice linked with low socioeconomic status and the commonly described effects of colonialism . Interrelated with these characteristics of social suffering are other processes of differentiation, which point to the extreme marginalisation of Greenlanders. Some of these processes are identified below. The stigmatisation and intersectional othering of Greenlanders has racist connotations related to their ethnicity, what Goffman would describe as 'tribal stigma' and compound by their low socioeconomic status. The process of othering is a reductionist discourse that others all Greenlandic participants to a few negative characteristics and by stereotype . For example, othering by stereotype is captured in the phrase 'grønlaenderstiv' . In a candid reflection on her own prejudices, a practitioner discloses a common Danish view of Greenlanders: I have had prejudices, just like everyone else, like: 'don't sit there and get grønlaenderstiv'. And 'Greenlanders are some wretched people, who sit on the benches and stumble around'. I might have said it myself once: 'Yes, that's a typical Greenlander" and shrugged. But after I got to know some Greenlanders outside the abuse milieu, then I got a different storey which I'm able to use in my work. [How] they meet these prejudices even though they may be well settled with husband and kids, house and job and so on you have to fight to get integrated and to be seen as equal. This discourse of othering adds an ethnic stigma to suffering associated with living on the edge of society. The consequence of such stereotyping that link a specific behaviour to Greenlanders is an oversimplification of the internal variations within the whole group of ethnic Greenlandic people that lead to stigmatisation . How this discourse works as a stigmatisation process is indicated by the way all the practitioners with Greenlandic backgrounds describe similar experiences. One practitioner explains: 'When I visit the shopping mall, some Danish people are suspicious of me. They seem to think I want to steal something When I pass by, they turn their noses up, they expect me to smell of alcohol'. Another Greenlandic practitioner comments on how her experience of stigma clashes with the idea of Danish unity: 'If we talk about Danish Commonwealth, there is no common at all because I still meet that gaze.' The practitioners from a Greenlandic background have the resilience to distance themselves from the stereotype and can reject the stigma by being explicit about its simplified view of Greenlanders in ways that our lay participants do not . The othering and stigmatisation of Greenlanders are incorporated into various forms of differentiation. Whether an individual has the resources for resilience means the difference between distancing from or taking on a stereotype such as Greenlandic alcoholic. Several Greenlandic practitioners and translators describe how Greenlanders, who are socially integrated in Denmark and thus able to fight the stereotype, also tend to distance themselves from marginalised Greenlanders who appear to represent the stereotype. Malik's broken relationship with family members is one among several examples of the consequential exclusion: 'My oldest son used to come and visit me sometimes, but I don't know where they live now. And my little brother, he says I'm a freak.' As such, differentiation that creates othering within the population of ethnic Greenlanders in Denmark leads to further exclusion of socially marginalised Greenlanders. How this perpetuation of the stigma works to exclude some Greenlanders was observed in a Greenlandic House to which access is only allowed if you are sober. MBA observed that the rule was displayed on posters and if somebody appeared drunk, they were reprimanded . Several lay participants avoided the exclusion by saying that they found the Greenlandic House 'too posh' and did not like going there. Another differentiation is evident between Greenlandic and Danish socially marginalised people in which Greenlanders are regarded as the lowest in the hierarchy. In drop-in centres and social care-homes, Greenlanders are usually isolated from other groups, which can be seen both as a strategy for avoiding stigma and a consequence of it. Another process of differentiation was identified within the group of our lay participants and reveals that the label marginalised Greenlanders does not refer to a homogeneous group. Several practitioners describe how Greenlanders take care of each other, but also emphasise that it 'depends on whether they are from East or West Greenland" Some of the West Greenlandic participants articulate a stereotype of East Greenlanders as someone who 'steals' , are 'cannibals' , and are 'lower' in the hierarchy and thus more or less excluded from the community. One example is Nora who is from East Greenland and lives in a social care-home. There are three other Greenlanders, and Nora is excluded from their community. As one practitioner commented: 'They don't really want her she is a cannibal, as they say. Tilde and Marie, they just can't get on with Nora. They shout and yell: 'fuck you'. Nora does that too'. As described, various processes of stigma and othering influence social suffering associated with a marginalised status. This constrains the possibilities for agency due to a legacy of trauma plus the mechanisms of exclusion. Intersectional othering is a discourse related to ethnicity and social status defined by Danes that discriminates against and excludes Greenlanders from Danes and marginalised Greenlanders from well-integrated Greenlanders. This complex of stigmatisation and othering leaves the participants in an extremely marginalised position at the end of their lives. The next section focuses on everyday life with illness, dying, and death. --- Approaches to illness and death Anina came to Denmark to find her cousin. She had a 16year-old pregnant daughter in Greenland. When VG first met her she was sleeping rough. She visited a drop-in centre to get a new coat because somebody stole the one she had left with her sleeping bag while she was having breakfast. For now, she wants to stay in [city] because she is having fun and has found a 'boyfriend'. A few months later, VG joins her at a hospital appointment. She says that her boyfriend has beaten her up and she has been kicked out of the shelter for being too disruptive, and she is back on the street but in touch with the street-team and shelter. At the hospital, we had to wait for the scan. Anina was impatient and wanted to get back to the drop-in centre for lunch. However, with VG's persuasion, she stayed. Over the next couple of months, Anina moved in and out of shelters. About 3 months after the scan, Anina died from a brain tumour. A practitioner reported that in the last weeks, she drank as much as possible and died in a shelter surrounded by practitioners but without family. Anina was more focused on maintaining her network with other Greenlanders and keeping everyday routines such as meals in drop-in centres than she was on her health issues. However, her relative contentment with her life belies the conditions of suffering compounded by the social disadvantages that have narrowed her life chances to their current form. Social suffering affects the way illness and symptoms are approached and managed in everyday life . This part of the analysis explores norms, symbolic meanings, and social interaction guiding the participants' interpretations and experiences of illness, dying, and death in their everyday-life contexts. Coping with a life-threatening life Several practitioners describe how Greenlanders cope with illness as a form of fatalism and a sign of their 'toughness'. These attributes can be seen as a tendency to not respond to symptoms until they are severe and thus barriers to conventional treatment and pain relief. Maren, who spent most of her time on the streets, did not seek help when her health obviously deteriorated. A practitioner describes that when she was eventually hospitalised: 'She had an abscess in her pelvic floor measuring 10 by 14 centimetres, severe infection and critically ill, and we do not know if she survives the rest of us would have reacted earlier'. During her admission, she became increasingly introverted and stopped eating. About a week after hospitalisation, she died. Another practitioner comments: X has been hospitalised and was told that he is going to die if he does not take his heart medication, so we ask: 'do you want to get this medication?' , 'No, no' They are calmer in some way. Today I told a man: 'Listen! You need to get the endoscopy, there is something wrong with your stomach' , and he told me that he was feeling fine and that it's unnecessary. There would have been more panicking if it had been a Danish man. Practitioners offer various for this fatalistic approach, as a cultural trait, as a result of stigma, or as a consequence of living a socially marginalised life: My experience is that Greenlanders are very tough. They have to be really ill before they say something or ask for help. It's very often that they are walking around with something, which we don't detect until they are extremely ill now I'm interpreting, but I think that they've been rejected and therefore they need to be very ill before they ask for help. Several explanations point to a cultural identity that is less individualised than Danes and that, combined with social marginalisation, can foster an attitude of not being worth caring for. Anna who is in treatment for throat cancer describes this fatalistic approach in terms of a norm of managing-it-yourself: VG: You like to manage it yourself? Anna: I do. For us Greenlanders you can say that we would rather manage it ourselves. VG: What does it take to accept help? Anna: I don't know. YES, if you cannot walk, then you need to get help. VG: But it's not something that you like? Anna: No, I don't like it. [asking for help] I don't, even though it's nice [to have help]. Malik describes a similar way of coping when he talks of 'nearly' calling an ambulance because he was in acute pain. These behaviours might be seen as evidence of the 'toughness' and 'fatalism' that the practitioners describe. However, as one Greenlandic translator reports, being seen as sick is associated with a loss of social status and people step back from the sick person. We have frequently observed lay participants going to enormous efforts to be seen as active members of their social group even when they are seriously ill. Anina does this even when diagnosed with brain cancer and Malik does this despite relying on a rollator. These acts of 'toughness' can be seen as a way of preserving a social network which is a key source of belonging and identity. --- Social networks Some of the participants are active within a social network. The data reveal a mixed picture of care from these networks when somebody gets ill. Anna is one among several examples of how illness can lead to social isolation and loneliness if it becomes impossible to continue routines and join the usual social network. VG asks Anna if her friends visited her after she began her chemotherapy: Anna: No, no one calls me and asks if I'm feeling all right. VG: Do you wish they did? Anna: Yes, occasionally. Of course, some ask me, my family, that is it. VG: When you finish your chemotherapy, do you hope they'll contact you again? Anna: I don't care, it doesn't matter. Because when I got my chemotherapy, that was when I was in pain and needed them. The social network does not necessarily disappear but the support it provides when somebody gets seriously ill mirrors the resources and role of the network and rarely provides the practical support the sick person may need. For a few days, Aida helps her uncle who lives in social accommodation. He is ill with TB and now has a broken shoulder. She says that she gives him his medication and cleans for him. She comments on his visitors. 'They are just sitting there drinking, fucking and do nothing Then they get grumpy but I tell them I need to clean and that they have to leave: 'Can't you see that he is lying here in his bed'. A couple of days later, Aida gives up. The uncle can now leave his room. VG observes that he is irritated and tries to avoid Aida. Aida says: 'I should help myself instead, I drink too much, my daughter is taking drugs and my son has started cutting himself '. The most common support delivered by social networks was the provision of alcohol, cannabis, and food. A practitioner describes a woman: She had her son who made some Greenlandic food, fish soup and such also just to be there and see if she needed anything. But we also experienced that when she became very ill. There were not as many with her in the apartment. But they still contacted us because they were worried about her. Why we couldn't help. So they really care about each other in a way I don't see with our other citizens with a different background. Providing food is a practical measure but also of deeper significance linked to symbolic ideas of health. Traditional food is important for wellbeing and health, as Tilde says. She misses Greenlandic food and states that it is 'for the body, we need that'. Material resources in the form of cannabis and alcohol are used as forms of self-medication. Kristian says that he believes cannabis might cure his cancer and that it helps with feeling sick after chemotherapy. Practitioners comment that social network is sometimes a hindrance to compliance with medical treatment: P1: They listen to each other. If you have schizophrenia, and need some medication and your friends say 'you do not need that' , then they believe it and rely more on their friends than on the health professional. It might also be that it is about health literacy. P2: It's so integrated in their culture to find support with each other. This final comment alludes to an underlying solidarity between Greenlanders, as Greenlanders abide despite the exclusions. As described in a group conversation with Greenlandic practitioners and volunteers in a Greenlandic House: 'We're cousins all of us Greenlanders. Almost everybody knows each other It can be very far out because the network is so big. Somebody might come here from Greenland to live with a friend's aunt, and they consider that family'. Yet, Nora and Anna are both examples of how different mechanisms of marginalisation exclude them from networking within the wider group of Greenlanders. Even though the participants' social networks are characterised by the mechanisms of marginalisation, fieldwork reveals that funerals are a symbolic ritual that overcomes exclusion. The desire to show respect for the dead person and their family is a unifier across social groups of Greenlanders and activates a wide network. Honouring the dead and loved ones is crucial; as one Greenlandic practitioner says: 'That is something we are brought up with. We have to accompany the dead to the grave no matter if we are family or not. Perhaps we know his aunt or whoever, and then everybody goes to the funeral for the last goodbye'. As such, death seems to dissolve the divide between socially marginalised and integrated Greenlanders. This stands in stark contrast to the funerals for socially marginalised persons with a Danish background where it is rare for more than a few people to attend. --- Dying, death, and spirituality We found a complicated picture of approaches to dying and death. Tilde who lives in a social care-home describes how even though death is a frequent part of everyday life, there is a reluctance to discuss it: Tilde: All you do is to go to one funeral after the other. Who's the next? When one is dead. They just die one after another . They die of drinking and TB, that's because they smoke cannabis, you know. Then they call it TB or cancer. There are so many I have known who have croaked. It's sad and unbearable to think about them. VG: Do you talk about it if somebody is seriously ill or dying. Tilde: No, you can't be bothered to hear that. If that is the problem, then I just leave, then, I don't know, then I just need a walk. The weather is fantastic right now because it's spring, then everything starts to blossom. Tilde's abrupt change of subject is typical of responses experienced in other interviews and several practitioners also comment on Greenlanders' reluctance to talk about dying and death. A Greenlandic practitioner points to a possible cultural aspect of this in her reflection: 'I think that [dying and death] is a touchy subject, but I also think that it's something which isn't a natural thing to talk about and can be a taboo. For example, if we talk about it, then death might come to us far too early. That is why you don't talk about death'. For the researcher, it is not possible to know whether a 'death taboo' is a deep cultural trait or a phenomenon that is observed in a marginalised group traumatised by frequent and sudden deaths. What is clear is that there are practical consequences, which complicate end-of-life care. Some participants have thought about their wishes for disposal but have not disclosed them to the practitioners involved in their care. An example is Kim who has children in Greenland. He says that he would like to be buried with his parents and brother in Greenland and that he wants to be cremated. VG: 'Why is it important to be buried in Greenland -is it to be together again?' Kim: 'Yes, with siblings and children.' Several practitioners are influenced by the idea of openness about dying and death including end-of-life decision-making. However, some say that it is difficult to have such conversations with Greenlanders. Petra has heart failure and her 20-year-old daughter is her only relative. To support the daughter, a practitioner would like Petra to complete an expression of wishes for her death: 'We cannot always talk about such things . That is what I would like to do how can we prepare your daughter to take all this responsibility?' At this point, it is useful to reflect on the practitioner's frustration. As pointed out by Yang et al. , stigma is embedded in the moral life of sufferers. The moral mode of experience is reflected in what matters most in everyday life and practical engagement. Lay participants are extremely marginalised, some of them are ruled by immediate needs given by the demands of substance dependency and thus forced to live life from 1 h to another. For lay participants, these conditions combined with trauma might play a role in marginalised Greenlanders' alienation from the conventions of planning for end-of-life. However, these observations do not mean that there is no awareness or planning for end-of-life as illustrated by Kim. We have found that when asked about dying and death, participants are willing to talk about what is meaningful to them, their feelings about the 'motherland' , their relationship to nature, and the importance of kin and a shared heritage of spirituality. These responses give some insight into the various ways spirituality is integrated into the participants' view of life. However, the lack of recognition given to this aspect also reveals a hidden layer of othering of Greenlanders' cultural identity because it is neither understood nor recognised in some of the contexts where the participants live and die. A telling example is the symbolic meaning of nature. Viktor, who recently moved to a care-home located far from the fjord, where he used to go every day, says: 'The water, nature, is where I connect to everything.' A Greenlandic practitioner also comments: This is where hope comes in. Moreover, nature itself plays an important role in everyday-life for many Greenlanders. This is where you feel yourself, this is where you hear quietness, this is where you come into yourself. This is probably also where you find strength from yourself rather than resignation. Nature, as a way of connecting to the world, echoes the life they used to live in Greenland. However, this is a tentative spirituality, a resource that has been overwritten by Danish culture and religion, as one Greenlandic practitioner states: It's something from generations that has shaped it. Christianity has taken up so much space, but the truth I carry with me is my own is somehow lying there deep in our hearts and can be opened culture and our identity. It's our identity that we belong together and that we come from a place in which spirituality has lived for many years and has been forgotten because of Danish society that overtook our lives in that way. For some lay participants, the adoption of Christianity, the religion of the dominant society, has brought solace and an avenue for expressing spirituality. For others, the connexion to traditional spiritual culture is fragmented but emerges in certain practises such as the tradition of name-giving. Although not a ubiquitous practise, it is important for understanding the wide approach to family and social networks, and for some, an underlying spirituality. A practitioner explains the spiritual meaning of names:'a person consists of three parts: a body, a soul and a name. It's the name that is a particular core part of the person, and that name can continue living in society by being given to a new-born person when somebody dies.' This spiritual approach to names was repressed by the Danish Christian mission . Illness and death culture are adapted to meaning and resources accessible in everyday life. The participant's approaches to illness, dying, and death are interrelated to characteristics of living a life marginalised from conventional society and cultural aspects linked to Greenlandic traditions of family life and spirituality. This includes a reluctance of seeking help, talking about serious illness, dying, and death, plus limited help from social networks, which some practitioners describe as non-compliance behaviour. A spirituality rooted in traditional Greenlandic faith is indicated as an important part of some Greenlandic participants' cultural identity. It is vaguely articulated by lay participants and as a dimension that historically has been suppressed and neither known nor talked about by Danish practitioners, we term it a hidden layer of othering. In the next section, we will explore barriers to accessing and navigating the conventional health and social care system. --- Navigating health and social care systems Kristian had advanced cancer. He moved to Denmark 6 months before the first interview and previously worked as a civil servant in Greenland for 30 years. He did not identify himself as socially marginalised. He had a sister in Greenland and a brother in Denmark. He told me that when his doctors stopped treatment in Greenland, he thought: 'I can't just wait here until I die'. He sold everything he had and moved to Denmark. A friend helped him to obtain his health insurance card, the GP referred him to the oncology unit, and he started outpatient treatment. The hospital referred him to a volunteer programme, which helped him to equip his home. Eight months after the first interview and until submission of this study, Kristian described an increasingly chaotic life. He struggled with financial matters and found it hard to understand who could help him. He stopped seeing the volunteer and his health deteriorated. The oncologist said that he would need chemotherapy for the rest of his life and that he needed to keep up his weight to continue in treatment. He commented: 'I somehow hope they will [stop treatment]. I need a break. I'm going to discuss it with the oncologist today'. Kristian said that he was struggling with nausea and he had no money for food. 'I've only got a little dry bread Some days I cook for my friend and then we eat together. Often I don't get breakfast.' VG asked if he was able to get money for cannabis, he said he needed to prevent nausea. 'I get some from friends and pay back when I've money. I spend too much money on cannabis Somehow; I wish that I had never moved to Denmark. But I've nowhere to live in Greenland. However moving isn't realistic.' After having talked to the oncologist about his treatment, he sent a message: 'I have talked to the doctor and he was busy so we didn't talk about the money issue.' Kristian did not want to go to a dropin centre for food and financial advice: 'I don't want to, it scares me, somehow.' Navigating the Danish healthcare and social care system is difficult for all lay participants like Kristian. The conventional institutions appear complex, with unfamiliar norms and procedures, guided by a logic that excludes lay participants. This section focuses on how this struggle with navigating the conventional institutional framework for health and social care can contribute to discrimination and stigma. --- Accessing the system Kristian's trajectory is a social decline that parallels his progressive cancer. This is an example of how the Danish system is exclusionary and difficult to access for marginalised Greenlanders. This combined with a social network unable to be supportive in navigating the system risks leaving a dying person in poverty and despair. Kristian's financial problems mean that he has not paid his rent and cannot afford food. However, as Kristian reports it, his oncologist focused on his treatment and physical health leaving no time for talking about social problems. Neither has he had help with his social problems from the palliative care team. He describes the system as a 'digital jungle'. A hospital nurse summarises the difficulties that marginalised Greenlanders face when they have to navigate the Danish system: It think that we've systems that actually make high demands for the users to be able to access them and you've to be pretty efficient in society and have electronic access to invitations, knowing who your general practitioner is, having the ability to actually turn up to your appointment, just to be able to access the system, and having the capability to show up in an outpatient facility. I think that you make high demands to some of our weakest patient groups. They need to fit into a box, which they don't necessarily fit into because they for several reasons are so multi-burdened. Some Greenlanders move to Denmark without a long-term plan. As Greenlanders are Danish citizens, they are not enrolled in an introductory programme as immigrants are. Therefore, they often, go under the radar as they do not know how to navigate a complex social system like that of Denmark. As one practitioner explains: Sometimes Greenlanders think that it's fantastic to come to Denmark, and then they sleep on cousin Peter's couch the first year and earn a living by selling bottles. Then, all of a sudden, they have to get NemID and a medical card and get involved in Danish society, and that is really difficult. Some of the practitioners mention simple barriers such as knowing which bus to use to go to the doctor and what side of the road to enter the bus. The challenges of accessing the Danish system are also caused by a clash of temporalities. Some lay participants, like Anina, operate according to a life lived from 1 h to another guided by what appears as meaningful in the present internal time . However, this does not fit well with a system based on clock-time exact appointments, waiting time, etc. In addition, there can be a cultural aspect of lay participants' approach to time. A Greenlandic practitioner describes how when she first started her job in Denmark, she could not see the point of making appointments and organising her day in a diary. Although, she quickly realised that doing so was necessary. Like other lay participants, she refers to Greenlandic hunting culture as incorporated into Greenlanders' identity: 'I think it comes from prehistoric times when the weather was the underlying basis for going out hunting. I think that it is still sitting there, instinctively.' Practitioners across the field report that these different approaches to time often clash. One says: 'For example if they receive a letter about an appointment at the hospital or something without us knowing, well it is 105 percent certain that they don't get there in time.' A hospital-based practitioner describes that often patients discharge themselves and do not get the same support as other patients because they cannot fit in with the way the hospital is organised: 'It's difficult to be allowed to [help them] because they can't function on a ward in which things are normal. I mean the rhythm of the day and the demands they have to face. They get so challenged because they feel that they're misread'. Also, individual discrimination is reported, as one practitioner from a social care-home describes: 'Sometimes in relation to visiting the GP, we need to be very sharp and be their spokesperson, because sometimes we experience that they get dismissed: 'It's because you drink' , 'it's because you smoke' , 'it's because you don't get decent food, you need to eat what they serve". --- Language and understanding Accessing the Danish system is further complicated by communication problems. Language difficulties are one of the most mentioned problems associated with communication between Danish practitioners and Greenlanders. Several participants find it difficult to communicate in Danish and this limits possibilities for understanding the nature of suffering. As a practitioner puts it: 'It's a challenge with language. What I experience is that they pretend to understand in order to be polite'. Valborg describes her strategy when she sees the GP: 'Even though I don't understand it, I just respond with 'yes, yes' and 'bye'. Several practitioners report that they have to speak a very 'simple' language to Greenlanders to obtain some understanding. Greenlandic practitioners within the field point out that a translator is essential for mutual understanding, but it is given low priority. The frequent use of relatives as translators adds serious ethical challenges. A participant from Greenlandic House reports that when her mother was hospitalised, she was told to translate for her mother. For her, it was difficult because she wanted to be in the hospital as a daughter taking care of her mother and not as a translator. Danish-speaking practitioners often act as translators in conversations with GPs for example. A practitioner from social accommodation describes the complexity of communicating and that it requires knowledge about the person. Actually, a lot of them are so bad functioning that they can't communicate with the GP. But often when the GP says something then we translate it, meaning that we repeat what was being said in a simpler way ensuring 'do you understand that he asked if your pain is in the right or left side?' When it is spelled out it usually works. But for some of them we have the perception that even though we try to act as mediators, they don't understand a single thing of what is being said. In situations like that we try to supplement with our observations of the person. 'We have noticed that she does this and this or complains about pain here'. It's not the optimal solution. The importance of using a translator is pointed out by a Greenlandic practitioner: 'When I talk to her, then she talks about all the things involving the problem. But when she has to explain it in Danish it only contours the problem she talks about, we don't get into the core of the problem.' The same practitioner illustrates the complexity of intersectionality in communication: I know how to speak the language, but to understand their cognitive level -where are you placed. And then add alcohol or cannabis abuse. There are more layers, and I've to figure out, where are you, how much do I need to explain. Do I need to explain it to you as an adult, as a teenager or something simpler? Moreover, I have to dig up information about their health. At the end-of-life, difficulties with communicating can leave the dying person alone in their pain. A nurse describes how with Maren it was difficult to know what she needed when she was in hospital. Her illness was advanced when she was finally hospitalised, and she did not speak Danish well and could not or did not express her needs. Another practitioner describes the difficulties of reading pain: 'I remember the Greenlander we had. He was in such pain but he couldn't express it. Yes, he could express it by shouting but not through language.' Link and Phelan describe that even in the absence of individual prejudice or discrimination, institutional practises can work to the disadvantage of minority groups and cause structural discrimination. For the lay participants, the difficulties of access and communication within the conventional health and social care system result in discrimination despite the principle of free and equal access. --- Discussion Drawing on the experiences of our lay participants, we first discuss the nature of social suffering and how it relates to total pain. Second, we offer a more general discussion of how the concept of total pain can be theoretically informed by the concept of social suffering. --- Total pain for socially marginalised Greenlanders Social suffering is an underexplored aspect of total pain. Our findings reveal the complexity of social suffering for an extremely marginalised group and point to how socioeconomic status and ethnicity work as intersecting factors which subordinated our lay participants by stigmatising and othering them . For our lay participants, vulnerability to social suffering is the product of traumas, poverty, and exclusion that leaves them in a harmed condition, which shapes their end-of-life experience . Those who have not already died an early death are likely to end their life undiagnosed, undertreated, and with limited access to appropriate care. Such an end-of-life trajectory affects total pain and we argue that due consideration needs to be given to the drivers of social suffering for marginalised Greenlanders at the end-of-life. Below, we will discuss how intersections of social factors work across micro, meso, and macro levels of society to cause social suffering and affect total pain. On the micro level, social suffering is incorporated into lived experiences of illness, dying, and death . For Kristian, his economic challenges increasingly overshadowed his illness. However, our findings show that social suffering is often unarticulated; obscured by everyday-life challenges. Anina and Maren who lived on the street until shortly before their death were preoccupied with maintaining their everyday life rather than responding to their symptoms of illness. To understand social suffering for lay participants, we need to understand how stigmatisation and othering work in their everyday life. As Yang et al. commented, attending to the moral experience of stigma, both for the stigmatised and the stigmatisers, reveals what is really at stake and what is threatened . The phrase grønlaenderstiv shows not only the stigmatisation incorporated into the common language but also the power of ethnic Danes to categorise Greenlanders as subordinated others. Such is the power of stigma that those categorised often internalise the stigma themselves. Our findings only confirm the well-established facts that stigma threatens self-esteem, mental and physical health, plus access to healthcare systems . As in other studies on socially marginalised groups' end-oflife experiences , we also find institutional meso-level barriers preventing access to health and social care for lay participants. Our study, like other studies on indigenous groups, revealed that such barriers are intermeshed with communication problems related to language and cultural approaches to illness and death. In Denmark, there is a presumption that Greenlanders, as Danish citizens, will know how to navigate the health and social systems. Social forces on a larger or macro scale are also in operation. The long shadow of colonialism and modernisation should be acknowledged as contributing to stigmatisation, othering, and exclusion of Greenlanders. However, some commentators have cautioned against the identification of Greenlanders as a specific target group. Nygaard-Christensen and Bjerge are critical of over-emphasising culture and cultural differences as contributing factors in relation to problems experienced by Greenlandic people . They point to the risk of alienating Greenlandic people even further and overshadowing more diverse representations that foreground the experiences of Greenlandic minorities. However, there is also a risk of not giving due consideration to these factors . We are not claiming a kind of Greenlander exceptionalism but rather seek to shine a spotlight on a highly marginalised group to unpack the concept of social suffering and how it illustrates the intersections of social factors that create such harmed states. Our participants' identify themselves as Greenlanders though their life storeys can be read as lived examples of a cultural/historic struggle with Danish culture. The focus on Greenlanders is important because the character of social problems for Greenlanders is more extreme than for other socially marginalised Danes . Among Danish citizens' attributed low status and power, Greenlanders rank lowest, with all the implications for life chances that entail . Subica and Link's causal model of health disparities identifies cultural trauma as a relevant causal factor that can address the social suffering we have identified . They define cultural trauma: 'as an overwhelming and often ongoing physical or psychological assault or stressor perpetuated by an oppressive dominant group on the culture of a group of people sharing a specific shared identity/affiliation ' . Cultural trauma can be intergenerational and affect health disparities via stigmatisation of affected groups and decreased access to flexible resources. This pattern reflects the characteristics of lay participants in our study, illustrative of what Duran et al. describe as a 'soul wound' . Soul wound refers to a'deep psychological injury that directly compromises people's sense of wellbeing, safety, self-regard, and coping manifesting in problems such as depression, anxiety, shared posttraumatic stress, and substance use' . It is no coincidence that the age range of our lay participants places them in a cohort of Greenlanders born in 1950s onwards for which there is statistical evidence of cultural trauma. The rate of suicide, alcohol use, and domestic violence increased massively in the 1950s, coincident with the decolonisation and modernisation of Greenland . This picture is consistent with the personal storeys of our participants who describe suicide, violence, and alcohol use as some of their earliest experiences. Our point here is that the cultural heritage of our lay participants is relevant for understanding the psychological traumas that are known to affect symptom burden at the end-of-life . Subica and Link also point to the degradation of cultural resources in the form of cultural modes of being including customs values and spiritual beliefs . Our study also points to the suppression and disregard of traditional spirituality that went along with colonialism and missionary activity that were part of the Danification of Greenlanders in the post-1950s modernisation. Most of our lay participants experiencing such traumas exemplify the complexity of individual and intergenerational cultural aspects that affect end-of-life for marginalised Greenlanders. Recognising the cultural trauma of indigenous groups as a factor in social suffering also brings insights into the nature of total pain for marginalised Greenlanders. Physical pain, social isolation, plus psychological and spiritual distress cannot solely be read as the person's symptoms given by a life-threatening illness; for the person with a cultural wound, total pain has to be understood in the light of a person's harmed cultural identity with the consequences it might have for living with life-threatening illness. Drawing on findings from our study, we have shown that social suffering is an important aspect of total pain. This takes us to the question: How can total pain be informed by the concept of social suffering? --- The concept of total pain As a starting point, it is helpful to come back to the origin of total pain. Saunders developed the concept as a way of thinking through the complexity of pain and as a strategy for managing pain. As Clark comments, Saunders recognised that: 'there may be several layers which have to be understood to have a full grasp of the problem of pain in the terminally ill' noting also that pain is indivisible from both body and the wider personality . Even though it has been criticised for its slipperiness, it has been an inspirational tool for palliative care . Inviting the concept of social suffering into the discussion of total pain does not deliver a full theoretical framework for total pain; it nevertheless makes an important contribution to filling the theory gap. As we have discussed, social suffering shapes the nature of suffering that comes with dying. To grasp the core problem of socially produced suffering, it is required to understand how social forces of oppression such as cultural trauma and poverty work. This is a dimension mostly ignored in the literature that deals with the social aspects of total pain . Aligning the social with the psychosocial definition emphasises individual experiences at the expense of the social understood as intersecting social forces. Social suffering is by definition produced by the social and it cannot be reduced to something within the person. This approach to the social has consequences for practise and here we point to just two implications. The first concerns the increasingly recognised problem that palliative care is an unfairly distributed privilege. While it is of course right that palliative care continues to try to understand the individual experience of pain, it is strategically problematic for palliative care to be blind to the social forces that determine who receives palliative care . Studying marginalised Greenlanders' social conditions has revealed inequity that calls for structural changes. We can only agree with Stajduhar that these inequities challenge the very foundations of hospice care because 'it is failing to care for the people who suffer most at the end of life' . The second implication for practise is more abstract but concerns the use of the social lens to try and understand what is at stake when a socially marginalised person, for example, presents to palliative care. Attempting to understand the other in terms of the social forces that have shaped their experience, like Malik, who has been exposed to Danification, trauma, and alcoholism, requires knowledge and recognition. We began this study with a doctor's account of visiting a terminally ill Greenlandic woman at home: 'As soon as I go through the door I can feel in my own reaction that all my prejudices are confirmed '-this comment can be interpreted as an instance of othering. When othering occurs, we already restrict the possibilities of understanding. There is a challenge here that must be addressed: the necessity but near impossibility of understanding the full meaning of pain. When Gunaratnam brings the concept of social suffering into conversation with total pain she points to how the very awareness of the limitations of understanding the other can actually be an opportunity for developing new ways of bridging and communication . Bringing social suffering into the ontology of total pain is perhaps a small step towards bridging the gap in understanding the complex suffering at the end-of-life experiences of socially marginalised people. --- Conclusion This study is the product of a long journey towards a still incomplete understanding of social suffering for people labelled as socially marginalised Greenlanders. As white middle-class Danish-speaking researchers, we have been confronted by the limit of what can be known about the other. An additional challenge has been that our participants are silent on the issue of social suffering, for they rarely complain about their situation not even when confronted by life-threatening illness. Although our participants say little, it is their situation that speaks volumes, as Greenlanders, as Danish citizens, and as socially precarious people, they are situated at the intersection of multiple forces. Significant in our findings are the following: The history of colonialisation and modernisation that has inflicted a cultural wound that cannot be ignored as a major contributor to social suffering. The long shadow of colonialism has, to some degree, alienated Greenlanders from language, spirituality, and cultural identity; Social cultures of care and community have, for our participants, become adapted to a culture of everyday life, frequently dictated by the demands of maintaining fragile networks and marginalised from conventional society; The paradox that as Danish citizens they have rightful access to high-quality health and social care, yet the organisational logic of these systems makes them difficult to access without the appropriate literacy to navigate the system; and The illness and death culture of our participants appear to be tough and stoical, but we need more knowledge to understand the intersecting factors constructing this facade. Our work with marginalised Greenlanders, many of whom died early, often undiagnosed and untreated, has also provided a platform from which to reflect on the nature of suffering and the topic of this volume, total pain. One important observation is that total pain, as currently theorised, says little about the social construction of suffering. As we have noted, some of our participants are silent about their suffering, but pain and suffering are not solely individual experiences. Understanding social suffering requires an awareness of the multiple social forces that intersect to create a harmed condition, which in the case of our participants so frequently excludes them from attention to their needs at the end-of-life. --- Data availability statement The datasets presented in this article are not readily available because the data are confidential to the researchers as consent is not in place for data sharing. Requests to access the datasets should be directed to [email protected]. --- Ethics statement The studies involving human participants were reviewed and approved by the Danish Data Protection Agency, Region of Southern Denmark Journal no.: 21/24661. Written informed consent or oral recorded informed consent was obtained for participation in this study in accordance with the national legislation and the institutional requirements. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

end-of-life care is currently distributed. Finally, we point to ways in which an understanding of social su ering can help to address the exclusion of some of the most vulnerable citizens from appropriate end-of-life care.

Introduction For more than 30 years, criminologists have studied racial differences in criminal justice experiences. Studies have considered both juvenile and adult patterns of arrest , conviction , and sentencing . These studies have traced individuals in the criminal and juvenile justice systems , and a number of studies have considered correlates of aggregate race differences in criminal justice processing . Results vary about how much racial disparity exists and about the proportion of observed differences that can be attributed to legally relevant variables, . Though criminologists do not agree on the proportion of racial differences in criminal justice processing that can be explained by legal differences, they do concur that not all of the observable disparity can be explained by legally relevant race differences . What then explains why people of color have more contacts with police, have higher arrest and conviction rates, and are more likely to spend time in prison? The popular answer is discrimination on the part of police, prosecutors, and judges. Unfortunately, we cannot directly assess with standard quantitative analytic methods either specific acts of discrimination nor racist sentiments on the part of criminal justice officials while performing their duties. Not surprisingly, officials generally do not reveal on questionnaires or in interviews that they are prejudiced. Laboratory research on implicit bias has made important strides measuring unconscious reactions to race by studying physiological changes produced by controlled stimuli, but this methodology does not apply well to field studies to definitively draw conclusions about discriminatory attitudes or behaviors in criminal justice settings. Qualitative studies have reported subtle forms of racial biases on the part of court actors . So discrimination remains a viable interpretation of legally unexplained racial differences in criminal justice decisions; but one difficult to measure in large field studies. Another possible explanation is seemingly racially neutral criminal justice practices, such as bail standards that emphasize the importance of defendant characteristics, disadvantage some segments of the population . Here we focus on racial differences in police contacts for juveniles. Obviously one reason that young people come into contact with police is delinquency. We address two questions: first, are there racial differences in police contacts which are not explained by criminal involvement, and second, are there features of social life that help to explain these differences. Differential contacts with the police are important because they potentially put young people who have these experiences at risk in other aspects of their lives. We more fully discuss these risks below. Garland argues that in the wake of the upheaval, riots, and criticisms in the 1960s and 1970s, police strove to become more professional and used more technology, and moved away from close community contact. Because that approach did not produce reductions in crime, police departments changed their strategies, leading to widespread adoption of community policing practices in the last decades of the 20 th century. The earlier period, where professionalization was stressed, led to poor interaction with the populace, especially poor and minority communities. The newer community policing approach brings officers into more daily contact with the public. This means that young people are now more likely to encounter police in a broad array of their social environments, including in their neighborhoods and schools. If the social environments of young people of color and White juveniles differ, the former may be at greater risk of having contacts with police independent of their involvement in delinquency. If the former are more likely to live, go to school, and work where there is greater police presence or where there is more surveillance, they may be more likely to come to the attention of the criminal justice system. For instance, Krivo and Peterson report that racial residential segregation consigns African Americans to neighborhoods where they are more likely to experience victimization. To the extent that police departments focus patrols where there is more crime , or where police perceive there to be more crime , African Americans are at greater risk of encounters with police. In this paper we will examine several important "environments" to assess the extent to which, in the face of community policing, they increase the likelihood that young African Americans will be more likely to have police contacts than Whites above and beyond what would be expected by observable differences in delinquency. Racial differences in criminal justice processing go to the heart of our collective notions of fairness and justice. When the questions are about juveniles they take on added importance. Labeling theorists, more than 3 decades ago convinced us, and continue to affirm the formative influence of interactions with agents of social control. Braithwaite , in making the case for reintegrative shaming, criticized normal justice procedures for engaging in "shaming" that is disintegrative . Here we take their invocations to heart in our effort to study social environment factors that contribute to race differences in police contacts. When teens have adverse interactions with police, the consequences can be longer lasting than for adults . If they are not treated fairly, or if they perceive that they are not treated fairly, their future behavior and interactions with law enforcement might be negatively affected. The question of the propriety of the behavior of police and other justice actors is also important to communities of color. When communities observe differences that cannot be explained, it potentially gives credence to an injustice narrative that harms community-police relations. So, as Garland described, officers engaging in community policing seek to have more interactions with citizens, but this presents risk for perceptions of police, and for the youths themselves who interact with officers. Here we do not test Garland's assertion that community policing leads to more public/police interaction. He more than adequately makes the case, as does the continuing existence of the Federal COPS Office in the Justice Department, whose website states their mission as to "advance the practice of community policing in America's state, local and tribal law enforcement agencies" . Our purpose here is to examine the extent to which, in the context of contemporary policing, there are racial disparities in juveniles' contacts with police that cannot be accounted for by differential delinquency involvement. And, if there are unaccounted for disparities, how do the social environments experienced by juveniles explain them? Here we are interested not in arrest or subsequent decision points in the juvenile justice system, but in contact with police. Of course, contact is a necessary condition for arrest, but we should no more presume that the former has only benign effects on children, than we should assume that all contacts with police are negative. While some effects of contacts appear immediately , others may show up later, some much later. The nature and consequences of such effects is an empirical question that gains importance if there are racial differences in likelihood that juveniles have contact with officers. Skogan reported that while negative interactions between citizens and police have the predictable effect of increasing hostile feelings and cynicism, positive interactions do not help, they have no substantial effect at all. One wonders if such patterns are exacerbated in minority communities where predominantly negative narratives about policeminority community relations are present. --- Racial Disparities in Crime and Criminal Justice We are not the first to analyze social environments and police contacts. Stewart and colleagues studied neighborhood context and Black youths' perceptions of police treatment. The environment did matter. Black youths were more likely to have adverse interactions with police when they were in White neighborhoods, especially those with growing Black populations. Similar patterns were uncovered by Brunson and Weitzer in a qualitative study of both Black and White juveniles in low-income communities. Tapia found that both race and social class were significant predictors of juvenile arrests, but interestingly, being "out of place," a minority person in a non-minority neighborhood, resulted in stronger effects for higher income Blacks than for those of lower SES. Here we add to this line of research by examining the experiences of young Black and White juveniles in two important social environments, as well as individual and family characteristics, while accounting for self-reported level of involvement in delinquency. Offenders, including juveniles, are arrested because they do something wrong. Perhaps people have contact with police because of their suspicious behavior or because they are in places where they probably should not be. Though some people concur with such statements , research results have long shown that it is not so simple. For nearly 40 years, criminologists have periodically explored the factors which lead police to make an arrest once they have made contact with a suspect. Piliavan and Briar stressed the importance of suspects' demeanor. Black and Reis found that the seriousness of offense and the wishes of victims influence police the most in their decisions to make an arrest. Recently, Beckett and her colleagues reported that police drug enforcement practices affect who is arrested more than behavior patterns of users and sellers. Other researchers have also reported that suspects' demeanor and non-legal factors such as race, social class, and gender affect police decision making when they encounter citizens . And, Brunson and Weitzer have documented that citizens, particularly in inner-city neighborhoods, try to prepare their children for "unwelcome" contact with police because of the potential for escalation during encounters. Generally, researchers have found that not all of the observed racial differences in police contact, arrest, or incarceration can be explained by patterns of criminal behavior. Three coordinated studies, in Denver, Rochester, and Seattle, used existing longitudinal datasets to examine the extent to which arrest and charging decisions could be accounted for by selfreported involvement in delinquency . While there were modest differences across the sites, the investigators all found that the racial differences in arrest were attenuated by differences in self-reported offending, but the race differences in police contacts remained statistically significant and substantial. These researchers, like others who have studied racial disparities, acknowledge that discrimination may be a part of the explanation of higher arrest and charging of Black youth, but they cannot dismiss other unmeasured reasons for the unexplained race differences. By examining critical environments that may lead to more police contact for Black juveniles beyond what would be expected by differences in offending behavior, we can specify important, previously unmeasured forces affecting racial differences. The two environments that we focus on are school and community, while also considering individual, familial, and peer influences. --- Race and Environment When discussing discrimination, the general public and many criminal and juvenile justice practitioners mean that criminal justice agents make racially prejudiced decisions to arrest, prosecute, or to sentence. It is possible, even likely, that this happens sometimes. But this alone is too limited of a conception of how race can negatively affect outcomes. In the United States, African Americans are at greater risk of victimization, school dropout, unemployment, and a host of other problems, in part, because they live in racially segregated communities . Racial residential segregation is a direct result of discriminatory practices in real estate markets and in mortgage lending . The consequences of segregated communities are examples of institutionalized racism, not just individual prejudice. The environments in which African Americans live are the consequences of continuing racialized patterns of social life . Peterson and Krivo advance what they call "a structural race perspective" to explain how stratification affects those living in Black and Latino neighborhoods. They write: "… the internal structural conditions of neighborhoods and the circumstances of surrounding areas are fundamentally racialized in ways that specifically reinforce and maintain superiority in the local contexts of Whites, and widespread deprivation in the contexts of African Americans, Latinos, and sometimes others" . In the current analyses we integrate a basic tenant of labeling theory, that particular social statuses , help to determine who is subjected to social control, with the structural race perspective of Peterson and Krivo, to add a micro-level component that allows us to specify elements of environments that may cause the negative outcomes from contemporary policing practices written about by Garland . And these negative outcomes add to the structured inequality faced by people of color. We expect that characteristics of juveniles will in part explain variation in their contacts with police. Those engaging in delinquency, reasonably, will have more police contact, but several other individual-level patterns may also increase the probability of contact. While these are not legally relevant factors that can be said to "account" for observed racial differences in police interactions, familial and peer characteristics may lead to differential contact. We discuss our reasons for these predictions below. We also expect that two particular environments, school and community, are especially important for explaining racial differences in police contacts. Important racial differences in these environments make it more likely that African Americans will experience more police contacts than would be expected by any observably higher involvement in self-reported delinquent behavior. Police presence in schools, particularly in schools in "trouble areas," and officer engagement in "problem communities" increase the likelihood of having contact with officers. Clearly, these two environments, schools and communities, are neither mutually exclusive nor independent of one another or from familial or peer group characteristics. Several careful qualitative studies demonstrate this in examinations of school violence. Brunson and Miller found that violence in school is linked to the disadvantage and violence that occurs in communities where students live. Mateu-Gelabert used ethnography to study the bi-directionality of school and neighborhood violence. And in a subsequent study he reports that schools that foster commitment to education can counter the negative effects of street codes that students are exposed to in their home communities . Sullivan argues that to understand school violence, the individual, school, and community factors must be addressed within a nested conceptualization. Hagan, Hirschfield, & Shedd demonstrated that there are links between neighborhoods and schools where serious violence has occurred. Also, families select neighborhoods based on multiple factors, including cost, and in doing so make choices about schools and the peers to which their children will be exposed. And Hellman and Beaton report that school violence in middle school is not linked to communities, but violence in high schools is. Why do we predict that family factors will influence police contacts independent of selfreported delinquency? Family factors, including parental monitoring, parent-child attachment, and the number of siblings influence children's behavior and experiences. We have included both problematic and non-problematic family functioning as potential contributors to differential contacts with police. If a household has an abusive or alcoholic parent, the familial interaction and the child are affected, and such factors may especially increase the likelihood that members of the family, including children, will come to the attention of authorities. How else might "bad families" lead to more police contacts? We can begin to answer this question by turning to old-school criminology, where the Jukes and the Kallikaks, two 19th-century families, were cited in very early works about the intergenerational transmission of criminal behavior . Presumably, local authorities of that day would have looked at the younger members of the Juke and Kallikak clans when there was trouble about because they were so frequently involved in trouble. A more compelling argument, though, for how families might increase police contacts, independent of the behavior of children, is provided by Anderson . Members of "street families," adhering to the "code of the street," will get in trouble with the law . In this era of community policing, it is likely that effective officers will have knowledge of who are in such families. In these situations, they may be more likely to seek out members of those families for questioning, just as 19th-century constables looked to the Jukes and the Kallikaks when something bad happened. Families will not only affect the behavior of juveniles, but we predict that, independent of that behavior, children from "problem families" will get more police attention, warranted or not. Family influences might be considered those most proximal to the teen, whereas peer influences are only slightly less so. Associating with deviant peers is one of the primary risk factors for engaging in problem behavior , but may also lead more directly to contact with law enforcement. As might be the case with family members, peers may also have had previous contact with law enforcement. Teens who associate with other teens who engage in illegal activity are more likely to experience a police contact regardless of their own illegal activity. No institution other than the family so dominates children's lives as their school. A major portion of their waking hours are spent there and, at least for those who do homework, school affects even the time not spent there. School also heavily influences young people's behavior. Those who are educationally successful are more likely to develop attachments to school and have positive feelings towards their teachers, and as a result are less likely to be delinquent . School environments affect children's social lives in positive directions through extracurricular activities, friendships, and budding love lives; and in negative directions through rivalries, bullies, and gangs. Peer relations in school are important determinants of young people's social status . Some students go to schools that must be equipped with metal detectors, others do not. Some schools have guards or police on hand while others have the luxury of not worrying so much about security. It is likely that these and other environmental differences will affect both delinquency and police contact. And because of the racial residential distribution of the American population, school environments vary by race. School environments are likely to affect the probability that a student will have police contact, above what might be expected based on behavior, in two ways: directly in school and indirectly in the community. First, if there is a police presence in the schools, officers may become involved in school discipline issues that would not have provoked a police response in the community. During the period when the data that we are using was collected, all Seattle middle schools had officers assigned to them. 1 Typically, one officer was assigned to two or three schools. The actual police presence in schools depended on the perceived need. Likely those schools with more discipline problems or those with more "at-1 The agreement between the Seattle Public Schools and the Seattle Polce Department was described to the authors by an official of the school system and was confirmed by a representative of the Seattle Police Department. risk" students had more actual 'officer present' times. The officers had some discretion in how they allocated their time. It is easy to imagine that schools may have used the officers in 'get tough' discipline strategies, but Seattle Police Department policy was that they not be involved in standard, noncriminal problems. But readers should recognize that officers may have intervened in the belief that early intervention would deter wayward children from a life of crime. This is, of course, the philosophy of the early juvenile courts . A second means by which school environments might produce racial disparities in police contacts is school discipline. Kirk reports there are common social control linkages between school suspension and juvenile arrest. And other researchers report that African American students are more likely to be suspended or expelled from school . Seattle Public Schools, like other districts, has a history of racial disparity in school discipline, including suspensions , and in a recent national study, the U.S. Department of Education found that African American and Latino students are more likely to receive harsher punishments . So if students are suspended, they may end up being on the street during school hours where they could potentially be viewed with suspicion. Racial residential segregation means that the neighborhoods from which young White and Black people come will be different in many important respects. It is well established that race and social disadvantage of neighborhoods-characterized by poverty, unemployment, welfare dependency, etc.-are related . Black children live in neighborhoods with more negative social and economic influences . They are likely to be exposed to criminogenic forces, are more likely to be victims of crime, and more frequently live where there is greater fear of crime . Consequently, there will be a greater police presence in the neighborhood environments of Black juveniles, as well as in their school environments. The neighborhood or community environment can affect police contacts in two ways as well. As the criminological literature has long documented, there are important social forces that lead to higher levels of delinquency among some segments of the population , but here we are interested in forces that will increase the probability of police contacts beyond those produced by criminal involvement. Because police know which neighborhoods have higher crime rates, enforcement is concentrated there, increasing the odds that juveniles will experience police contact. A second means by which juveniles may be put at risk of involvement with the police in their community is by exposure to "risk-producing elements," specific people who officers may watch out for. Families who embrace the "code of the street" are not just known by members of the community, but also by law enforcement. Juveniles who spend time with carriers of the code of the street or who spend time where such people frequent will have more opportunities to encounter police . Other researchers have studied social environments and contact with justice system actors. What this study uniquely contributes is the opportunity to study the impact of multiple social environments, here measured with considerable detail, on racial differences in police contact, net of Black and White differences in self-reported delinquency. The current study complements qualitative research on school violence and extends quantitative examinations of peers, neighborhoods, and police contacts by adding simultaneous consideration of families and schools. --- Method Sample Parents of eighth-grade students during the 2001-2002 school year in the Seattle school district received a letter describing the study, and the parents were contacted by phone. Families were included if the teen and one or both parents consented to participate. Eligibility included self-identifying as African American or European American, speaking English as their primary language, and planning to live in the area for at least 6 months. Recruitment stopped when an adequate number of African American and European American males and females had agreed to participate. Forty-six percent of families who received letters consented . The parents who refused were more likely to be European American, married, and had a higher education on average than those who consented. Other ethnic groups were not recruited. The sample was stratified by teen race and gender. There were significant differences by race in several demographic variables. European Americans reported higher per capita income and parental education, and African Americans reported higher prevalence of single parenthood . Some teens in each race group self-identified as mixed race , but were included in these analyses. Most primary caregivers were female , with 71.6% being the adolescent's biological mother. Caregiver gender and relationship were similar across race with one exception: more African American youth had another female caregiver as a primary caregiver than did European American youth [χ 2 = 13.95, p < .001]. Data collectors went to the families' homes. Teens and their parents completed selfadministered questionnaires in their homes using laptop computers while the data collector was present. This ensured that parents did not monitor their teens' responses. Prior to parent/ teen interaction tasks, a trained research assistant set up video equipment, provided oral and written standardized instructions to each family, then left the room while the family completed the task. Upon completion of a warm-up task, the primary caregiver and the teen completed two structured interaction tasks: a 10-minute problem-solving interaction and a 5-minute recognition task during which the parent and child complimented one another. Family members received $15 each time they completed questionnaires. The family received $50, and each participant received $15 each for completing observational measures. --- Measures Police contact was collected on the teen surveys. Contact was coded as having occurred if the teen responded affirmatively to any of the following questions. Have you ever been picked up or stopped by the police, but not arrested; been in trouble with the police for something you did; been arrested by the police; spent time in a juvenile detention center for something you did wrong? Of the 331 eighth graders in the sample, 70% had no contact with police. Ninety-nine had some contact with police, 76 had "only contact" but were not arrested, 23 had been arrested, and 6 had been placed in detention. No questions were asked about where contacts took place or any specific information about the context of the actual police contacts. Teen self-report of criminal activity was measured separately for property and violent crimes. Each was computed as the mean of two items measured on a 5-point scale of frequency in the past 30 days . Property crime includes arson and theft. Violent crime includes carrying a gun to school and hitting someone with the intent to hurt them. Race was based on parents' reports of their child's race on school enrollment forms . Gender was reported by teens on the survey . Age was measured in years calculated from birthdates reported on the survey and the date the survey was completed. Household per capita income was calculated from parent's endorsement of 1 of 11 categories for annual household income . We assigned the midpoint of the range and then divided by the number of people in the household as reported by parents on the survey. Single-parent status was reported on the parent survey . Parent juvenile delinquency was an index based on parent retrospect accounts of their own teen years, "Before you turned 18 did you …." Response options were 0 = No, 1 = Yes. The five items were skip school, get drunk, run away from home overnight, use a weapon in a fight, and often start physical fights. Sibling criminal activity was measured with a single item, "In the past year, have any of your brothers or sisters done anything that could have gotten them in trouble with the police " 0 = No, 1 = Yes. Family conflict was computed as the average of 13 items taken from the Strauss Conflict Tactics Scale . Response options for eight items were a Likert-type scale from 1 to 7. One item ranged from 1 to 4, and four items were dichotomous. The item scores were standardized and then averaged. Items included "Family members often criticize each other," and "In the past month, how often did you yell, insult, or swear at your teen when the two of you have disagreed about something?" Observed rewards for positive or negative behavior were computed as composites of item responses from trained raters reviewing videotapes of parent-teen interactions. Variables were measured using the Social Development Model-Observational Coding System . Eighteen raters completed an average of 93 hours of training. Ratings were made using 5-point Likert-type scales . Twenty percent of the videotapes were double rated to check inter-rater reliability. Inter-rater agreement was high . Rewards for positive behavior was computed as the mean of seven items, including "Caregiver was warm and encouraging of teen's opinions" and "Caregiver reinforced or rewarded teen's prosocial behavior or attitudes." Rewards for negative behavior was computed as the mean of four items, including "Caregiver failed to respond to teen's negative or antisocial behavior or attitude" and "Caregiver reinforced or rewarded negative behavior or attitudes." Delinquent peers was measured with teen report. The teens were asked to name their three best friends , and were then asked a series of questions about each of those friends, including having done anything in the last year that could have gotten them in trouble with the police. A dichotomous score was created, with 1 indicating at least one of the friends had engaged in the behavior. Academic success/grades were based on reports from teens: "Putting them all together, what were your grades like last year?" Responses ranged from 1 = very poor to 6 = very good. School discipline was computed as the average of four standardized items from the teen survey, including "In the past year, how often have you been sent out of the classroom for doing something wrong?" . These items did not ask about police involvement in school discipline, although it is possible that some incidences of school discipline may have led to police contact. Community/neighborhood resources and cohesion was measured using 10 items from the parent survey. Item responses were from 1 to 5, indicting how accurate the description or how frequent the activity.2 High scores indicate positive neighborhoods. Items include "How often do your neighbors visit each other's homes" and descriptors such as "nice parks and playgrounds" and "crime" . Neighborhood environment was assessed using the average of nine items from the teen survey. All items were measured on a 4-point scale , coded so high scores indicate safer, less deviant neighborhoods. Items included "If a kid carried a handgun in your neighborhood, would he or she get caught by the police?" and " Do you feel safe in your neighborhood?" . Teen report of deviant adult network was computed as the mean of four items measured on a 5-point scale. The items were, "About how many adults do you know personally who have … in the past year?" The deviant behaviors were: used marijuana, crack, cocaine, or other drugs; sold or dealt drugs; gotten drunk or high; and done other things that could get them in trouble with the police, like stealing, selling stolen goods, mugging, assaulting others, etc. --- Analysis Preliminary analyses included chi square and t-tests to examine simple race differences in police contacts, and the 18 predictors from six domains . Logistic regression analyses were conducted to determine the unique contribution of predictors to the probability of reporting a police contact at any time before the eighth-grade survey was conducted. Predictors were entered in blocks to reduce the number of separate models to be tested and to examine the effects of predictors within a domain simultaneously. Seven models were tested in a hierarchical fashion. The order in which we added predictors was based on our specific questions. First we wanted to know if there were race differences in the likelihood of police contacts after controlling for criminal involvement based on self-report. We added demographic variables in the next step because they were least likely to be influenced by having had a police contact and were therefore most exogenous to the outcome. The rest of the predictors were added starting with the environments most proximal to the individual and working out to the least proximal. Teens originate and live their lives in families first, then eventually have some choice about their peers. Schools and communities/ neighborhoods are typically not directly chosen by teens, but are to some extent chosen by their parents. Step 1 included the teen's self-report of criminal activity , as well as race predicting police contacts. Step 2 included self-reported criminal activity, race, and demographic variables . In order to determine if variability in other domains accounts for race differences, predictors were added in successive steps in this order: parent and sibling criminal activity, family interactions, delinquent peers, school/grades, and community context. Although it is not possible to calculate a true R 2 for logistic regression models, a measure of the variance explained by the model can be calculated using the likelihood-ratio index , comparing the log likelihood of the intercept-only model to that of the model including predictors and then adjusting for the number of predictors in the model. To examine whether predictors of police contacts were different for African American and European American teens, we tested the interactions of each predictor by race in separate models. The clustering of families within schools was addressed using SAS GENMOD . Missing data were imputed . Missing data ranged from 0% to 4% across variables, with slightly less than 19% missing across variables. Forty imputations were calculated and appropriate standard errors were computed using MIANALYZE . Models reported in Table 2 used multiple imputations. However, because adjusted pseudo R 2 is only an approximation of variance explained by the model and has not been validated for estimation across multiple imputed datasets, pseudo R 2 was calculated using only those participants with complete data. --- Results Means and standard deviations are presented by race and for the total sample in Table 1. African American teens are almost twice as likely as Whites to report having had a police contact . Significant race differences were evident for 10 of the 18 predictors. African Americans reported more property crime , but not more violent crime than White teens. As reported above, African American families had lower income and were more likely to have single parents than White families. No race differences were apparent in parent-reported juvenile delinquency. African American teens were no more or less likely to report sibling criminal behavior than Whites. Family conflict was higher , and observed rewards for positive behavior were lower among Blacks than Whites. No race differences were evident for rewards for negative/problem behavior. Blacks in the sample had fewer delinquent peers than Whites . African Americans reported lower grades and more school disciplinary events than White teens. Although parent reports of neighborhood resources and cohesion were not different, African American teens did report less positive neighborhoods than Whites , and reported more adults in their network with recent deviant behavior . Odds ratios are presented in Table 2. Self-reported property crime and violent criminal activity did significantly increase the probability of ever having a police contact in Step 1 . After controlling for self-reported criminal activity and race in Step 2, Black teens were more than twice as likely to have a police contact as White teens. Girls were half as likely to report a police contact as boys . In Step 3, parent juvenile delinquency increased police contacts by about a third , but sibling criminal activity increased the likelihood of police contact by over 4 times . Race differences were significant after controlling for family criminal activity. In Step 4, higher levels of observed parent rewards for negative behavior were significantly associated with police contacts , while family conflict and observed parent rewards for positive behavior were not. In this step, self-reported criminal activity was not significantly predictive; however, race differences were. In Step 5, associating with at least one close friend who broke the law more than doubled the risk of police contact , and this effect was statistically significant before school and broader environmental factors were included. School factors were added in Step 6. Teen reports of school disciplinary contacts significantly increased the likelihood of police contacts , but grades did not. In this step, race differences were not significant. In the last model , community contextual measures were added. Teen and parent reports of neighborhood quality and cohesion were not related to police contacts. However, teen reports of knowing adults who drank, got high, or committed crimes in the past 12 months were significant. Knowing adults who exhibit these deviant behaviors significantly increased the likelihood of police contacts . Pseudo R 2 statistics for these models should be interpreted with some caution. There are no statistical tests for the incremental change or differences between nested models. Furthermore, they have been calculated without addressing missing data and are therefore based on subsamples of the original 331. Adjusted pseudo R 2 increases from 3% to 11% with the addition of predictors from each domain. The last model, including 18 predictors, explains just 41% of the variability in the risk of a police contact. Tests of interactions between race and each predictor separately produced two statistically significant interactions: neighborhood quality and academic performance . More cohesive, safe, resourced neighborhoods significantly reduced the probability of police contact for Blacks , but not for Whites . Higher grades reduced the probability of police contact for Whites , but not for Blacks . --- Discussion We examined several social environments that influence the likelihood of police contact, although not all of them accounted for racial disparities. Interestingly, low income and single-parent household were not significant predictors. But we did find that family patterns contributed strongly to police contacts. Having parents who have a history of involvement in juvenile crime increased the odds of police contact by about one third. Parents who were delinquent in their youth are perhaps more tolerant of similar behavior in their children and view this behavior as just normal teenage activity that their children will grow out of. Youth with a sibling involved in criminal activity were also more likely to have police contact. We suspect this may be in part due to police knowing the older siblings and being aware of the household. We did not examine if these families are, in the words of Anderson , "street families," if the police had such a conception of them or other views of them, but these results are consistent with what would be expected if police were aware that parents, brothers, and sisters have been in trouble. In addition, families were important because of observable behaviors between parents and teens in which parents encouraged or failed to discourage negative behavior. This finding points to the importance of parenting practices-or the practice of everyday parenting-in which parents have the opportunity to influence their children in both positive and negative ways. In video observations, we observed things like parents laughing or teasing about the child's misbehavior or ignoring deviant behaviors such as fighting. For example if a parent talked about doing something fun together when their child was suspended from school it was coded as a reward for negative behavior. Although this is not the main thrust of this research, we note that the effect of parent's own juvenile delinquency is no longer significant when observed rewards for negative behavior are included. This is consistent with the notion that parents with a history of delinquency are more tolerant of their children's problem behaviors which in turn increases the likelihood of a police contact. Other family predictors, conflict, and observed rewards for positive behavior were not predictive of police contacts. This finding is consistent with families socializing children in a counter-normative direction, exposing their offspring to increased scrutiny from law enforcement. Turning to the two environments that are central to our analyses, school and community, we found that the relationship between grades and police contacts was significant for White teens, but not for Blacks, suggesting that high academic performance, although important in other ways, does not protect Black teens from early contacts with law enforcement. Selfreported school discipline was higher for Blacks than for Whites and accounts for racial differences in eighth-grade police contacts. During the time of the study, schools in Seattle had programs which linked police officers to middle schools. This practice is consistent with community policing as described by Garland , which we expected might lead to greater risks for African American youths. Even though Seattle Public Schools' policy stated that officers were not to participate in school discipline, one cannot but help but speculate that these data may be reflecting police involvement in student disciplinary actions. Some officers may have believed that by intervening early they would decrease the likelihood of future police contacts and arrests. But, analyses by Hirschfield and Sweeten , which find that arrest and involvement with the juvenile justice system increases school dropout should give these officers pause. In general, there is reason to believe that further criminalizing school infractions will not only have deleterious effects on future behavior, but also on school performance . Of course it is also possible that youth with a police contact are the ones who are more likely to act out in school, which is not illegal, but can lead to more frequent and serious responses from school authorities. These cross-sectional data do not allow for a clear causal interpretation. Finally, teen reports of neighborhood quality were not predictive of police contacts; however, parent report of resources and cohesiveness were, but only for Black teens. Parent's views of the neighborhood may be more objective and reliable. Cohesion and resources are related to crime rates, and poorer neighborhoods have a higher police presence, increasing the chances of contacts for Black teens. The same may not hold for Whites. Black teens may also live in a wider range of neighborhoods, with the poorest Blacks in poorer neighborhoods than poor Whites. If this is true, the relationship between neighborhood quality and police contacts may be stronger when comparing poor to moderate neighborhoods than when comparing moderate to high-quality ones. Juveniles' reports of deviance in their broader adult network did significantly predict higher probability of police contact in both groups. Contact with such adults in the community may place teens in proximity to crime. Black youth report more deviant adults in their network. They are put at greater risk because association with deviant adults may lead police to consider the teen to be a lead in their investigations, or even a witness to illegal activity. Or police may simply associate the juvenile with adults involved in crime, and assume the teen is also involved. The primary limitation of this study is the use of cross-sectional data which does not allow for disentangling cause and effect. Factors we include as predictors could be predicted by unmeasured variables which also predict police contacts . In some cases the predictors might be influenced by having had a police contact . We recognize potential problems with the reciprocal effects of police contacts, delinquency, and the environments that we study . The analytic steps adding groups of predictors could be conducted in any order to reflect different levels of explanation. We chose an approach that reflects our specific hypotheses about how community policing might account for race differences beyond delinquent involvement. Regardless of the ordering in which predictors are added, once all of the predictors are included, the model is the same. Although we observed race differences in almost all of the predictors, most of them did not predict police contacts and therefore cannot account for race differences in the likelihood of those contacts. The importance of school and adult network variables may be reciprocal in that kids who interact with police make lower grades, run afoul of school authorities, and are acquainted with more deviant adults, which leads to further police contacts. A second limitation is the lack of information about where and how specific police contacts occurred. More detailed information would help to illuminate how and why schools, siblings, and deviant adults contribute to police contacts beyond self-reported delinquent activity. Although the families in the African American sample are an accurate representation of the Black community in Seattle, nonparticipation by more affluent Whites could reduce our estimates of race differences and income disparities. In spite of this limitation, the race differences in income are striking. Finally we should note that our data collection began 10 years ago. Seattle Police Department practices have not changed appreciably since then, but we acknowledge that some community responses may have shifted. It is our view that there is value in documenting the patterns and processes that we have studied here, even with 10-year-old data; and such secondary data analyses are in keeping with National Science Foundation and National Institutes of Health efforts to encourage researchers to maximize the use of existing data sets. We should note that Seattle's ethnic composition is different from some of the cities where a great deal of sociological and criminological research has been conducted . The African American population is relatively small, and Seattle has a sizable Asian American and Asian immigrant population, as well as growing African immigrant and Latino populations. In fact, after the 2010 U.S. Census of the population, one of Seattle's south-end zip codes was found to be the most diverse zip code in the nation. For three reasons we believe that our findings are worthy of note even with Seattle's ethnic composition. First, While Seattle differs ethnically from some large cities, it is much like others , so while we would argue for caution in generalizing our findings, we believe that Seattle is not so unique as to raise serious concerns. Second, our sample, is drawn from that portion of Seattle that, both in ethnic and economic composition, is much more like cities with entrenched disadvantaged populations. Third, others have used the city of Seattle to do important research on policing and crime . What are the implications of our findings? On the positive side, some may argue that early police contacts should be seen as early intervention, particularly for children having problems at home, in school, or who are associating with bad elements in the community. Perhaps this is so, but the contrary may also be the case. Forty years ago, labeling theorists described the negative possibilities of contacts with agents of social control. More recent publications in the labeling tradition continue to affirm these negative consequences . The theory would predict that "unwarranted" contacts lead future problems, including perhaps to more crime and future involvement with the criminal justice system. Perhaps other negative outcomes, such as school failure, employment irregularity, or drug involvement may increase as well. This is particularly troublesome because of the observed race differences in police contacts. Here we have seen that African American juveniles are more likely to have had police contacts and these cannot be explained by differential criminal involvement, at least not as measured by self-reported delinquent behavior. Nor do they appear to be the result of differential income, single parents, family conflict, delinquent peers, or overall neighborhood quality. There exists a dominant narrative in the Black community that police, and the criminal justice system more broadly, discriminate against African Americans. Even those who believe that early intervention will sometimes "nip problems in the bud" should recognize that racial differences in criminal justice contacts that are not justified by criminal behavior will seem to confirm the accuracy of that narrative. Additional research is needed to explore the consequences for individual juveniles of the differential police contacts that we have observed, but it is now safe to say that it is very likely that the race differences that we do see, in confirming the discrimination narrative, are not good for the relationship between African American communities and law enforcement.

Criminologists agree the race disparity in arrests cannot be fully explained by differences in criminal behavior. We examine social environment factors that may lead to racial differences in police contact in early adolescence, including family, peers, school, and community. Data are from 331 8th-grade students. Blacks were almost twice as likely as Whites to report a police contact. Blacks reported more property crime but not more violent crime than Whites. Police contacts were increased by having a parent who had been arrested, a sibling involved in criminal activity, higher observed reward for negative behavior, having school disciplinary actions, and knowing adults who engaged in substance abuse or criminal behavior. Race differences in police contacts were partially attributable to more school discipline.

Introduction --- C ontemporary cross-national studies demonstrate that, in most countries, the majority of young people report a very good health status. This is reflected by indicators of general health, 1 healthy dietary practices, 2 engagement in physical activity 3 and abstinence from substance misuse. [4][5][6] Yet, inequalities in health persist. These are especially disadvantageous to adolescents with a lower socio-economic status . 7 Socio-economic gradients in adolescent health have persisted over recent decades. 4,8,9 Among adolescents, such inequalities have been identified for self-rated health; psychological health; academic performance; 10 mental health problems; 11 life satisfaction; 12 overweight and obesity; and its behavioural determinant of physical inactivity. 13,14 There is evidence of an increasing socio-economic gap in such health indicators during the last decade, 8,9 although the available research base is still scant. Consideration of these inequalities and their impacts remain a high priority for health policy. 15 Relatively few studies have focused on international trends in adolescent health inequalities. 8,9 Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes relative to others. Focus on adolescents at risk for poor health is important as such inequalities tend to perpetuate into adulthood. 15 While economic arguments suggest that investing in the early childhood years is an efficient strategy to build a productive future workforce, 19 similar arguments are seldom heard with respect to adolescence, although this remains a crucial developmental period for a healthy transition to adulthood. There is also evidence that interventions that are implemented during adolescence can mitigate adverse effects of poor well-being during earlier childhood. 20 In a large cross-national study that spanned 12 years and involved some 700 000 adolescents from 31 European and North American countries, we studied the extent to which their SES, relative to their peers in the same country, was associated with being in the bottom-end of different indicators of health behaviours. Here, we focused on indicators of physical activity and diet. Physical inactivity and the absence of a balanced diet are key reasons for increases in childhood obesity 21,22 and other indicators of short-term and chronic morbidity. [23][24][25][26][27][28][29][30] Our primary objective was to estimate the magnitude of the association between family SES and poor adolescent physical activity and dietary behaviours across 34 countries. Moreover, if there was evidence of such an association, we aimed to determine whether the magnitude of this socio-economic gradient changed over time in the different countries . This foundational information would inform social policies and health promotion strategies, internationally. --- Methods 'Health Behaviour in School-aged Children' is a crosssectional school-based study that has been conducted every 4 year since 1982. HBSC aims to increase the understanding of adolescent health and well-being as well as the health behaviours and social environments that contribute to such outcomes. 5 Data are collected from 11-to 15-year-old school children according to a common international protocol. Each member country obtained ethics clearance to conduct the survey from a university-based review board or equivalent body. A detailed description of the aims and theoretical framework of the HBSC study can be found elsewhere. 4 The current analysis used data from the last four HBSC cycles in --- Socio-economic status The Family Affluence Scale is a measure of material family wealth developed as an indicator of absolute level of SES. 23,24 Between 2001/2002 and 2009/2010, the HBSC FAS scale was comprised of reports for the following items: ownership of a family car ; own bedroom ; family holidays during the past 12 months ; and family computer . These items are combined to produce a composite score ranging from 0 to 9 . In 2013/2014, two more items were introduced to this scale, numbers of bathrooms and ownership of a dishwasher. 24 As per precedent, 31 cycle-specific measures of FAS were transformed into a continuous proportional rank score ranging from 0 to 1, separately by country, with the country sample means set at 0.5. These country-specific ridit scores reflect the proportion of respondents with lower family affluence, with higher values reflecting higher levels of SES relative to others within the country. In regression models of health, one unit on the ridit scale refers to the difference between the least affluent and most affluent adolescents in the country. 8 --- Physical activity Moderate-to-vigorous physical activity was measured by asking on how many days over the past seven participants were physically active for at least 60 min. The survey item defined MVPA as 'any activity that increases your heart rate and makes you get out of breath some of the time' and includes examples such as running and brisk walking. 24 Response options ranged from 0 days to 7 days. The item has been used in the HBSC since 2001/2002 and correlates highly with a general question about physical activities. 3 The measure reflects current policy recommendations for children's physical activity. 25 --- Healthy diet Following precedent, 26 two items on fruit and vegetable consumption from an abbreviated food frequency questionnaire were combined into a healthy eating index. Each item was first re-coded from an ordinal to a ratio scale as follows: 'never'= 0, 'less than once a week' = 0.25, 'once a week' = 1, '2-4 days a week' = 3, '5-6 days a week' = 5.5 and 'once a day, every day' and 'more than once a day, every day' = 7. The two items were then added together so that the index ranged from 0 to 14, with 0 corresponding to never eating fruit or vegetables and 14 to eating both fruit and vegetables at least once a day. --- Unhealthy diet Items on sweets and sugar-added soft drinks consumption similarly contributed to an unhealthy eating index. Items were first reversecoded as follows and then summed together: 'never' = 7, 'less than once a week' = 5.5, 'once a week' = 3, '2-4 times/week' = 1, '5-6 days a week' = 0.25 and 'once a day, every day' and 'more than once a day, every day' = 0. In the composite 0-14 scale, 0 denotes consuming both sweets and sugared drinks at least once a day and 14 refers to never consuming sweets or sugary drinks. --- Indicators of 'bottom-end' health To identify adolescents who have substantially poorer health behaviours relative to their peers, we estimated the mean in the lower half of the distribution for each indicator separately, by country and year. Adolescents who fell below this threshold reported worse outcomes than an average respondent among the least well performing half of their population. This method has been applied in past HBSC, 16 UNICEF 17,18 and other international reports. 27 The present study focuses on the individual determinants of scoring in the bottom-end group rather than on the country-level dispersion of health scores in the lower half of the distribution [see 27]. --- Analysis Our analysis focused on the strength of associations between the country-specific ridit-transformed FAS score and bottom-end health outcomes. We used linear probability models separately for each indicator, country and survey year. Such models are computationally straightforward and easy to interpret in terms of differences in probabilities of the outcome, but since they can produce biased estimates 28 we also replicated all analyses using logistic regression. The findings were qualitatively identical . All models controlled for age and gender and were fitted separately by country, using sample weights. Standard errors were adjusted to reflect the complex sampling structure of the surveys, with schools identified as primary sampling units and regions as strata . We report on the effects of a one-unit difference in the ridit score for family affluence on the probability of being in the bottom-end group for each indicator, controlling for age and gender. To investigate whether the effects of family affluence have changed significantly over the last four survey cycles, we included a trend variable with interaction terms between this trend variable and each of the three other predictors . --- Results There was substantial cross-national variation in the size of the bottom-end group of adolescents to define the bottom-end group. The share of adolescents that 'fell behind' on MVPA ranged from 10% in Portugal to 27% in Belgium, while the proportion of those that 'fell behind' on healthy eating varied from 16% in Belgium and Switzerland to 26% in Hungary and Spain. While one in four adolescents scored poorly on unhealthy eating in the combined sample, this varied from 12% in Sweden to 33% in Norway. Table 1 shows large and widespread socio-economic inequalities in physical activity and healthy eating, with mixed results for unhealthy eating. In all 31 countries in the 2013/2014 analysis , adolescents from more affluent households were significantly Notes: Percentage point difference between the most and the least affluent adolescents; Standard errors adjusted for clustering at the school level and stratification by region ; Statistically significant effects in bold. less likely to fall behind in MVPA, with the FAS gradient ranging from À8.3% in Finland to À21.9% in Luxembourg . Thus, the probability of scoring in the bottom-end group in MVPA is 21.9% points lower for adolescents from the most affluent families in Luxembourg than for their peers from the least affluent families. Similarly, adolescents in the lowest SES strata were more likely to fall further behind in healthy eating in all countries except Malta and Romania, where there were no statistically significant differences identified. The United Kingdom showed the largest negative gradient in healthy eating: the poorest adolescents were 21.3% points more likely to fall furthest behind. Hungary posted the largest negative gradient in unhealthy eating: adolescents from the most affluent families were 16.1% points less likely to consume excess sugar. In contrast, the most affluent adolescents in Estonia were 7.1% points more likely to engage in unhealthy eating. In the vast majority of the countries studied, the sizeable and statistically significant socio-economic gradient in MVPA remained stable over time . In six countries, it became significantly larger in absolute terms between 2001/2002 and 2013/2014: Belgium, Italy, Latvia, the Netherlands, Sweden and the United Kingdom. The FAS gradient in healthy eating remained stable over time in most of the countries studied . In Canada and the United Kingdom this gradient increased, suggesting that adolescents from lower socio-economic backgrounds were increasingly less likely to consume fruit and vegetables when compared with their peers. In three other countries-Latvia, Lithuania and Romania-the FAS gradient has decreased. Although in Latvia and Lithuania adolescents from lower-affluence families were still significantly less likely to eat healthily, in Romania socio-economic inequalities were no longer statistically significant in 2013/2014. Conversely, in the majority of the countries analysed, there were no statistically significant socio-economic inequalities in unhealthy eating in 2013/2014. In eight countries, adolescents from more affluent households were less likely to report frequent consumption of sweets and sugary drinks, while in three others , they were more likely to do so. While in many countries, there was not a significant FAS gradient in unhealthy eating between 2001/2002 and 2013/2014 , higher consumption of sugar in less affluent households persisted over time in France, Ireland and the United Kingdom. The FAS gradient in unhealthy eating changed significantly between 2001/2002 and 2013/2014 in eight countries. In six of these-Estonia, Latvia, Lithuania, Poland, Portugal and Romania-adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, but these differences either narrowed or disappeared by 2013/2014. In the remaining two countries, Belgium and Hungary, adolescents from less affluent backgrounds were increasingly more likely to report a higher frequency of unhealthy eating. --- Discussion Adolescence is a critical period of transition in the life course, characterized by biological, psychological and relational changes. Such transitions are of fundamental importance to health 15 and set the stage for future patterns of adult inequalities. 29 In the vast majority of studied countries between 2011/2012 and 2013/2014, adolescents from relatively low SES families had a greater likelihood of falling furthest behind in health, particularly with respect to reported physical activity and healthy eating. For unhealthy eating another pattern emerged, with no indication of an association with SES for the majority of the countries and mixed results for the rest. Sweets and soft drink consumption may be associated with knowledge about healthy diets, which is typically higher in more affluent socio-economic groups. 30 Indeed, in eight countries , young people from low socioeconomic groups were overrepresented in the bottom-end of unhealthy eating in 2013/2014. In four of these , the pattern has persisted since 2001/2002. Meanwhile, in another six countries adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, although the association decreased or disappeared by 2013/2014. This is consistent with historical studies that documented a positive association between family affluence and frequent soft drink consumption among adolescents in Central and Eastern Europe at the turn of the century, possibly as an indicator of the ability to afford such luxuries and hence relative wealth. 32 Inequalities in adolescent health were widening in some of the countries studied. Socio-economic inequalities in MVPA increased in six countries during the 12-year study period . In Canada and the United Kingdom, the association between SES and adolescent healthy eating became more pronounced over time, as did unhealthy eating in Belgium and Hungary. Widening national income inequalities tend to increase socio-economic differences in health, perpetuating the socio-economic divide. 8 Conversely, in some countries, inequalities in adolescent health decreased considerably between 2001/2002 and 2013/2014. Adolescents in Latvia and Lithuania from less affluent families were less likely to eat healthily, but this association weakened significantly between 2001/2002 and 2013/ 2014. For Romania, socio-economic inequalities in healthy eating disappeared in the same period. These results suggest that social gradients in health can evolve over time. This study has several potential limitations. First, the crosssectional design of the HBSC study precludes the establishment of the temporal sequence of events, limiting claims on causality. It is unlikely, however, that the outcomes of physical activity and diet would precede of SES. Second, the indicators of adolescent health used here, although standardized and validated for cross-national comparison, 26,33 come with their own inherent limitations. The indicators of fruit, vegetables, sweets and sugary drink consumption are all based on frequency of intake, rather than amounts consumed. 2 Thus, we do not know whether, e.g. adolescents who reported daily consumption of fruit and vegetables actually consumed five portions of fruit and vegetables a day, but our measures are the best proxies for healthy and unhealthy eating afforded by our data. Our measure of SES relies on an assumption that 'the scores of the [FAS] scale can be used to rank individuals and groups along a latent continuum of material wealth'. 31 Although FAS has been validated against other measures of family SES, such as parental occupation, 23 the scale is subject to measurement error; and it is very challenging to assess this construct by self-report in populations as young as 11 years. Finally, not all countries were present in each of the four survey cycles, limiting the cross-country comparability of the results. In conclusion, this study establishes that socio-economic inequalities in adolescent physical activity and dietary behaviours are large and stable across countries and over time. These are important findings because they suggest two common mechanismsphysical activity and diet-by which health inequalities emerge and adolescents become socially disadvantaged within and across different cultures and societies. The results indicate a fundamental unfairness [16][17][18] that affects the most socio-economically disadvantaged adolescents in almost all societies; an unfairness that is consistent and persistent and sets the stage for negative health trajectories. These findings point to a universal need for public health interventions focused on physical activity and diet, including social policies that specifically target these and other aspects of health in our most disadvantaged children, as priorities, nationally and internationally. --- Conflicts of interest: None declared. --- Key points Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes relative to others. This study reveals large, widespread and persistent socioeconomic inequalities in the risks of the poorest adolescent health behaviours with respect to reported physical activity and healthy eating both across 34 industrialized countries and over time . Inequalities in unhealthy eating are less prevalent and may change their direction over time.

37 Cristofar SP, Basiotis PP. Dietary intakes and selected characteristics of women ages 19-50 years and their children ages 1-5 years by reported perception of food sufficiency.

Introduction HIV testing and counselling is the cornerstone of treatment, care, and prevention [1,2]. It is particularly through HTC that the ambitious goal of 90, 90, 90 , by 2020 [3] can be reached. Previously, most efforts were focused on voluntary counselling and testing as the primary means of providing testing and encouraging people to become aware of their HIV status [2]. However, coverage remains low and many infected persons in both developed and developing countries remain undiagnosed. Despite the availability of rapid test with the possibility to have the results approximately in 20 to 30 minutes, in some contexts particularly, there are many who get tested but fail to return for their results [4][5][6]. For example, in the USA, data from HIV testing performed at publicly funded counselling and testing sites using conventional HIV enzyme immunoassay testing from 1999 through 2002 found that 19% to 22% of people with positive preliminary HIV tests did not return for their test results [6]. In 2009, a survey conducted in 12 Sub-Saharan Africa countries with high HIV prevalence showed that only 10% of women and 12% of men were tested and received their test results [7]. In an evaluation of five years of routine program data in Vietnam, Hong et al. found a failure to return rate of 3.5% [8]. In a study of female sex workers in China, Xu et al. found a FTR rate of 47.1% [9]. More recently, in France, Laanani et al. and Pahlavan et al. , respectively, found a FTR rate of 6.5% in a study conducted in a free and anonymous screening center [10] and 14.5% in an HIV-positive population [11]. 2 International Scholarly Research Notices Identifying and targeting these people may improve the return rate for VCT and the proportion of individuals who are aware of their status. Therefore, the objective of this systematic review is to identify the factors that prevent people who are tested for HIV from returning for their results or facilitate their doing so. --- Methods This systematic review examines barriers and facilitators associated with returning for HIV test results in various types of populations and settings. The outcome variable, returning for HIV testing results, is dichotomous. From this point of view, some studies concerned factors associated with returning for HIV test results, while others focused on failure to return . This review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement guidelines [44]. --- Search Strategy. A comprehensive search strategy was developed to identify studies published between January 1985 and June 2015. Four electronic databases were searched , combining terms related to HIV, counselling/testing, and return/failure to return. Retrieved references were imported into Endnote X7, and then duplicates were removed. The detailed search strategy is available upon request. --- Study Selection. All identified records were initially screened by two independent investigators and verified by a third researcher. Eligible studies had to meet the following criteria: be an original research study; be written in English or in French; report adolescents, adults, pregnant women, men who have sex with men , injecting drug users , or female sex workers ; include participants undergoing HIV tests; use a quantitative method to assess return or failure to return for HIV test results; and report a statistical association between a potential predictor/correlate and return or FTR. Study designs of interest were cross-sectional and longitudinal. No geographical restrictions were applied. The reference lists of the relevant articles were also reviewed for additional publications. A short list of records was prepared and the full text reviewed independently by two authors. Citations that were clearly irrelevant were excluded. Uncertainties and disagreements about inclusion were resolved through discussion involving both investigators . --- Data Extraction. Two authors independently extracted data from each study that fulfilled the inclusion criteria using a standard form. Study characteristics as well as key findings related to factors associated with return or failure to return for HIV test results were extracted. Any factors analyzed associated with FTR or return for HIV test results were listed, and the results of multivariate statistical tests for association were noted. For studies where a multivariate statistical test was not done, the results of bivariate analyses were noted. When the result of the measure of association in multivariate analysis was not significant and not reported by authors, the factor was not considered in the synthesis. --- Quality Assessment. The Newcastle-Ottawa Scale for cohort studies and an adapted form of the Newcastle-Ottawa cohort scale for cross-sectional studies were used to assess methodological quality. NOS is a tool for assessing the quality of nonrandomized studies to be used in a systematic review [45]. Each study is judged with a "star system" on three points: the selection of study groups, the comparability of the groups, and the ascertaining of exposure or outcome. Studies for which at least five out of nine items on the NOS were deemed satisfactory and in which appropriate statistical analysis was conducted were considered to be of sufficient methodological quality and included in the review . At each stage of the quality assessment, the reviewers discussed among themselves until a consensus was reached on which studies to include. --- Data Synthesis and Combined Effect Sizes Associated with Return. Factors associated with either return or failure to return may be arranged into barriers and facilitators inspired by the Socioecological Model , which is a framework to examine the multiple effects and interrelatedness of environmental, contextual, and social factors on individual behaviour [46,47]. Recognizing that most public health challenges are too complex to be adequately understood and addressed from single level analyses, the SEM includes a more comprehensive approach that integrates multiple levels of influence to impact health behaviour and ultimately health outcomes. These levels of influence include intrapersonal and interpersonal factors, organizational factors, and structural factors. In this review, due to their small number, organizational and structural factors were grouped as contextual factors. Since we anticipate a potential variability of the methodology across the included studies, we used a random-effects model based on the inverse variance method to estimate the pooled odds ratio for each factor potentially associated with returning for HIV test results and its 95% confidence interval [48,49]. The Higgins's 𝐼 2 statistic was used to quantify the percentage of the variability in individual effect size estimates which is attributable to the heterogeneity [50,51]. This heterogeneity was tested using a chi-squared test [50,51]. Moreover, we performed sensitivity analysis by removing the included studies from the pooled size estimation one at a time. These analyses allowed us to explore the individual contribution of each study to the heterogeneity in the meta-analysis. When we could not explain the heterogeneity, we have interpreted the pooled effect size estimates with caution because these effect sizes would be explained by other factors, which were not taken into account in our analyses. A 𝑝 value of less than 0.05 was considered statistically significant. Analyses were performed in Review Manager . --- Results --- Study Selection. The primary search strategy identified 3,019 potentially relevant citations. After the removal of duplicates and the initial title and abstracts screening, 60 citations were kept for the full-text review. Studies were excluded if they did not report quantitative results or just reported the rate of return or failure to return without assessment of associated factors . The remaining 36 studies were appraised for their methodological quality and included in the analysis. No study was excluded on the basis of quality assessment. A flow chart illustrating the selection process is shown in Figure 1. were cross-sectional studies. Seven of the studies were carried out in Sub-Saharan Africa; seventeen in the USA; three in Australia; six in Asia; and one in Brazil. Populations under study were diverse, including general population , pregnant women , injecting drug users , men who have sex with men , high-risk heterosexual individuals , HIV-positive individuals , factory workers , individuals with psychiatric problems , adolescents , HIV-negative individuals , and female sex workers . The outcome of interest was dichotomous, with 20 studies focused on failure to return and 16 on return for HIV test results. --- Study Characteristics. --- Quality Appraisal Results . Studies were generally of high quality . A total of 3 cohort studies scored 9/9, one study scored 8/9, and 6 studies scored 7/10. For the crosssectional studies, one study scored 9/10, 11 studies scored 8/10, 4 studies scored 7/10, 4 studies scored 6/10, one study scored 5/10, and one study scored 4/10. --- Barriers and Facilitators of Returning for HIV Test Results. In total, 236 factors associated with returning for HIV test results were identified. Among these, 123 factors were reported as barriers and 70 as facilitators. The association was not statistically significant for 72 factors. At the individual level, factors were classified into sociodemographic characteristics , risk behaviours , perceived risk , HIV knowledge , reasons for visit/testing , HIV test results , history of testing , psychosocial factors , and other individual factors . Factors grouped at the interpersonal level were risk partner behaviours , social support , knowledge of person with HIV , domestic violence , and other interpersonal factors such as partner age , years in couple , and communication within the couple . Contextual factors comprised the type of clinic attended , year of testing , and characteristics of the testing center, such as availability of counselling , condom distribution , clinic visit , confidential testing , and location of the testing center in the same city as treatment center . --- Individual Level 3.5.1. Sociodemographic Characteristics. Age was the most reported factor . This factor has been reported as a barrier to returning for results in 7 studies [12,19,27,28,30,36,37] and as a facilitator in 5 studies [13,22,35,36,39]. The association between age and returning for HIV test results was insignificant in 10 studies [10, 12-14, 26, 28-30, 35, 41]. In these studies, being 30 years of age or over was reported as a facilitator in 4 studies [13,22,35,36] and as a barrier in just a single study [27]. On the other hand, having less than 30 years of age was reported as barrier in 6 studies [12,19,27,28,30,36] and as facilitator in 2 studies [22,36]. Level of education was reported in nine studies. In 4 studies [15,17,26,30], it was reported as a barrier to returning for HIV test results, especially for people with no education or a low level of education. In 3 studies [9,34,39], it was - - 2 STD clinic 2 1 - 3 Detention facility 1 - 1 2 Primary care clinic 1 - - 1 HIV test clinic 1 - - 1 Mobile clinic 1 - - 1 Prenatal/obstetric clinic 1 - - 1 Drug treatment center 1 - - 1 Health department 2 - - 2 Outpatient medical service 1 - - 1 Private physician - 1 - 1 College - 1 - 1 Base clinic 1 - - 1 Other type of clinic 3 - - 3 Clinic visit 1 - - 1 Counselling - 1 - 1 Year tested 7 - - 7 1998 1 - - 1 1999 1 - - 1 2000 1 - - 1 2001 1 - - 1 2002 1 - - 1 2003 1 - - 1 2004 1 - - 1 Other contextual factors 1 2 - 3 Condom distribution - 1 - 1 Same city as treatment center - 1 - 1 Confidential testing 1 - -1 Grand total 181 99 127 407 reported as a facilitator for those with a medium or high level of education. In studies with a mixed population when sexual orientation was reported , being heterosexual or bisexual appeared as a barrier to returning for HIV test results [11,36,37]. Marital status was reported in six studies. In these studies, being married or living in a couple [8] and being a widower [8] emerged as facilitators of a return for results. --- Risk Behaviours. Several risk behaviours were positively or negatively associated with a return for results. The number of sexual partners during the last 6 to 12 months was reported in 8 studies. In 5 of these studies [8,10,26,27,37], having more than 5 sexual partners was reported as a barrier to returning for HIV test results. Otherwise, having a single sexual partner during the last 6 to 12 months [22,37] was not significantly associated with a return for results. Using a condom has been reported as both a barrier [30] and as a facilitator [26,41], but in most cases, the association was not significant [14,37,39]. Having a history of STIs was reported in nine studies. However, it was reported equally as a barrier [16,19,29,41] or a facilitator [13,22,25,31] as regards a return for HIV test results. --- Perceived Risk. Perceived risk has been reported in nine studies. In those studies, having low perceived risk [26] and not seeing oneself at risk [10,27] was reported 20 International Scholarly Research Notices as a barrier to a returning for test results. However, this result is somewhat controversial because one of these two studies showed that having high-perceived risk [10] was a barrier to a return for results, and having a medium perceived risk was reported at the same time as a barrier in one study [14] and as a facilitator in another [33]. In addition, the association was insignificant for 4 studies [10,35,38,39]. 3.5.4. Psychosocial Factors. The association between the return for HIV test results and psychosocial factors showed divergent results . For instance, not believing in selfprevention from HIV [26], believing that HIV can be cured [29], and thinking that a medical follow-up can improve the course of HIV [29] were reported as barriers to a return for test results. However, having high self-esteem [33] and positive coping skills [33] appeared as a facilitator for a return for results. Feeling anxious about HIV was reported as a barrier [15,42] to a return for HIV test results. --- Health Coverage. The association between the existence of health coverage and a return for results was studied in two articles. Having health coverage [24] appeared as a facilitator and not having health coverage as a barrier [31]. 3.6. Interpersonal Level. Interpersonal factors were reported in 18 studies. The most common factors were risk behaviours of the sexual partner. The association of these factors with a return for results was investigated in seven publications. The HIV status of the sexual partner [8,19,41] or being a client of a sex worker [8] was identified as a facilitator for a return for HIV test results in three studies. Having a sexual partner who is a sex worker [8,41] , having a partner who drinks alcohol [28] or consumes drugs [41] , having a partner who is always travelling [28] , and having a partner who did not test [28] were all reported as a barrier to returning for results. Domestic violence [28] and rape [16,29] were reported as barriers to a return for test results in three studies. The association between a return for results and the availability of a social network has been studied in six studies. On the one hand, it appears that having one or more gay friends [37], having a counsellor [40], or knowing someone infected with HIV [35] are barriers to a return for results. On the other hand, having social support [40] and lacking a family confidant [38] were reported as facilitators of a return for test results. --- Contextual Level. A negative association was found between the return for HIV test results and having a confidential test in one study [12]. The same negative association was found when the testing was done in facilities such as family planning clinics [27,36] , a detention facility [27] , a primary care clinic [27] , an HIV testing clinic [27] , a mobile clinic [27] , a prenatal/obstetric clinic [36] , a drug treatment center [36] , a health department [27,36] , an outpatient medical service [16] , and a sexual health clinic [20] . However, the association was positive in the case of a physician clinic [36] and a college [36] . Other organizational factors, such as the year of testing [22] , condom distribution during the visit [46] , or having tested in a center located in the same city as the treatment center [46] and not having pretest counselling, emerged as facilitators of a return for results. --- Combined Effect Sizes of Factors Associated with a Return for HIV Test Results. The pool estimates and sensitivity analysis of the association of the return for HIV test results with certain factors, including gender and race for studies conducted in the USA, injection drug use , HIV test results , and HIV testing history , are shown in Table 3. The combined analysis showed that being female [8, 18-20, 22, 30, 34, 35] is significantly associated with a return for results when studies with specific population are excluded. In the studies from the USA, black people tend to return less frequently for their results than white people [12, 18, 22, 27, 34-36, 38, 39]. There is no significant association between returning for test results and HIV test results or HIV testing history. Finally, the association between injection drug use and returning for test results was significant [8,12,14,22,41] when only the general population was considered. Thus, being an IUD appears as a barrier for returning for HIV test results. --- Discussion The objective of this review was to report the factors that were statistically associated with the return for HIV test results in different studies, regardless of the target population, the HIV test method used , the waiting time for results, or the country. Despite these different contexts, periods, and populations, the majority of studies considered the same factors. The vast majority of reported factors are found at the individual level . Very few studies have reported contextual factors, such as organizational factors, policies, economic factors, or social factors. The differences in statistical analysis introduced a great deal of heterogeneity with respect to the studies. Thus, it was not possible to combine effect sizes for all factors. The factors not included in the meta-analysis were grouped into barriers and facilitators based on their statistical association with the dependent variable . Although the factors have been grouped into categories according to the ecological model, it is important to specify, in accordance with the socioecological approach, that the categories are not exclusive but rather influence each other. --- Sociodemographic Factors. Age and level of education acted both as barriers and facilitators. However, the trend indicates firstly that young people and individuals with a low level of education were less likely to return for their results. Indeed, there is evidence that young people are often less informed about HIV and also exhibited a lower rate of HIV testing than adults [45,46,52]. They are unaware of their risk behaviours and are less likely to return for their test results. Conversely, individuals with higher levels of education can better understand the importance of screening [44,53] and are more likely to return for their results. --- Risk Behaviours. The literature has shown an association between the return for test results and risk behaviours [52,54]. In fact, people who display risk behaviours can also develop fear and anxiety with respect to knowing their test results. In these circumstances, they are less likely to return for their results even if they had the courage to get tested. Thus, in this review, the positive test result, injection drug use, a high number of sexual partners, getting paid to have sex, and having symptoms of sexually transmitted infections at a testing visit were reported as barriers to returning for test results. --- Perceived Risk. Studies that have examined the association between perceived risk and a return for test results are sometimes contradictory. Indeed, some studies have shown that people who have a high-perceived risk of contracting HIV were more likely to return for their results [33,55]. Other studies have shown that people with a low perceived risk do not return for their results [10,14,26,27]. This second situation might be explained by the fact that many people at high risk of contracting HIV do not perceive themselves as at risk [56][57][58]. Therefore, they do not see the importance of returning for their results and knowing their status. This is why it is recommended that the education of individuals be intensified in order to foster a high and precise perception of risk. --- Interpersonal Factors. The sexual partner's risk behaviours were the most frequent group of factors influencing a return for test results. Furthermore, having social support, having an HIV infected partner, or being a client of a sex worker have been reported as factors that encourage people to return for their results. In fact, having sex with a highrisk person might increase the perceived risk, which leads the exposed person to learn his or her HIV status [59]. On the other hand, the family and social network provide social support and reinforce social norms [60] that might encourage a return for results. In contrast, being a member of a social network of people at risk, such as a partner of a sex worker, of an alcoholic, or of a drug user, having gay friends, or knowing someone infected by HIV tend to hinder a return for results. These risk groups often experience discrimination and stigmatization [1,61]. Therefore, they are less likely to get tested, to return for their results, to disclose their HIV status to others, to adopt preventive behaviours, or to access treatment services, care, and support. Finally, domestic violence and sexual assault also hinder a return for results. Despite the implementation of strategies that enable women to get tested at opportune moments such as during pregnancy or childbirth, domestic violence remains a barrier for the entire testing process [62][63][64]. First, the female victim of sexual violence is afraid to return for her results and know her HIV status because she is afraid of being rejected by her partner who can blame her for having tested without his consent and for being responsible for his contamination in the case of a positive result [62,64,65]. Second, the feeling of guilt and fear of victimization and stigmatization experienced by a raped woman can hinder her return for results even if she could be tested [62,66,67]. --- Contextual Factors. The HIV testing center and its characteristics were the most frequent contextual factors reported in different studies. Getting tested in most of the sites appeared as a barrier to returning for test results. One reason may be the type of screening test offered at these sites. Many of the studies in this review were conducted before the use of rapid tests. Recently, several HIV testing centers in developed countries and in developing countries have reported an increase in the demand for testing, the proportion of people who received posttest counselling, and the knowledge of status following the introduction of rapid tests [68][69][70][71][72]. Other studies also showed that clients prefer the centers where they can receive their results without delay on the same day [73][74][75]. However, it is also reported that when the testing center is linked with the treatment, the pretest counselling is done well, and there is distribution of condoms, this set of factors encourages people to return for their results [1,7,32,53]. This literature review has some limitations. Firstly, the differences in the measurement of factors and the specificity of certain populations introduce heterogeneity and do not allow meta-analyses for all factors. Secondly, most of the studies were conducted before the advent of rapid testing, but nowadays HIV testing is performed by rapid tests. Therefore, the issue of failure to return for HIV test results is only important in very specific contexts. The use of rapid tests might change the distribution and frequency of certain factors. A majority of the studies were conducted in the USA The countries of Sub-Saharan Africa, which represent 2/3 of infected people worldwide, do not often publish their results, or only publish their results in local journal articles, which are not indexed in most databases. Our search strategy was limited to publications in English and French. Only articles published in peer-reviewed journals were considered; grey literature and conference proceedings were not. This may have some implications for the external validity of our results. However, the review included a large number of studies, covering different regions, a broad range of populations, and barriers and facilitators with respect to returning for HIV test results. Furthermore, to the best of our knowledge, this is the first review to focus on factors associated with returning for HIV test results. --- Conclusion Helping more people learn their HIV status requires the strengthening of counselling and testing services. Returning International Scholarly Research Notices for HIV test results is the gateway for knowledge and acceptance of HIV status. Various recently implemented strategies, such as provider-initiated testing and counselling, community-based testing and counselling, home-based testing and counselling, and the use of rapid tests, might not be effective if the people tested are not well advised and do not accept their results. This review identified important factors that need to be addressed to ensure that people return for their HIV test results. Most barriers and facilitators identified were found at the individual level. These results highlight the fundamental role of counselling. Individuals most likely to fail to return for their results must be identified and targeted by the counsellor and delivered a specific message. --- Competing Interests The authors declare that there is no conflict of interests regarding the publication of this paper. Submit your manuscripts at http://www.hindawi.com --- Stem Cells International Hindawi

This systematic review aims to identify factors that facilitate or hinder the return for HIV test results. Four electronic databases were searched. Two independent reviewers selected eligible publications based on inclusion/exclusion criteria. Quantitative studies published since 1985 were included. Thirty-six studies were included in the final review. Individual level barriers included sociodemographic characteristics, such as being a male, of young age and low education level, risk behaviours such as injecting drugs, having multiple sexual partners, and psychosocial factors. Older age, higher education level, being a woman, having high self-esteem, having coping skills, and holding insurance coverage were identified as facilitators. Interpersonal barriers and facilitators were linked to risk behaviours of sexual partners. Contextual barriers included essentially the HIV testing center and its characteristics. This review identified the most important factors that need to be addressed to ensure that people return for their HIV test results.

Introduction Colombia and Venezuela share an extensive and dynamic border of more than 2,000 kilometers. Since the 1970s, tens of thousands of Colombians have migrated to Venezuela, both for economic reasons and because of Colombia's extensive armed conflict with paramilitary and guerrillas. " Between 1963 and1973, 556,683 people emigrated, 69.5% men, to Venezuela, the United States, Ecuador, Panama, Canada, Peru, Chile, and Bolivia" . Estimates suggest that some five million people-about 10% of its populationhave left Colombia, mainly for the United States, Spain, andVenezuela . Since 2015, however, Colombia has become a host country for immigrants. Due to the political, economic, and humanitarian crisis in Venezuela, for example, the migration of Venezuelans to Colombia has been intensifying. Migratory movements often give rise to xenophobic discourses that demonize and instrumentalize foreigners as scapegoats to explain social problems such as unemployment, labor informality, prostitution, and insecurity . Colombia's new role as a host country for migrants has been an issue of great relevance in political discourse and electoral contests and has also received extensive media coverage. These discourses are part of a black propaganda strategy of some political groups that, in addition, have sold their party as the only solution to prevent the country from "becoming another Venezuela" . These messages follow the logic of hate speeches that justify discrimination and violence while simplifying complex issues, and that can even manifest themselves in physical and verbal violence against immigrants: The media coverage of migrations in the United States and Europe confirm the prevalence of certain negative representations in the construction of the image of immigrants in the receiving societies, in which the labels on exclusion and words like "Invaders of the Nation" stand out, as well as other negative metaphors related to the siege of the national territory of out-of-control forces such as those of natural disasters such as avalanches or waves. These hate speeches are also amplified in the Colombian mass media, as the authors of this paper explained in another article . The objective of this study is to define a set of actions and proposals, in the form of a guide or manual, for journalistic news coverage of migrations, focusing mainly on the functions that journalism should fulfill in a country that receives massive and forced migrations. These recommendations are based on in-depth interviews with migration researchers and practicing journalists, as well as on the Qualitative Document Analysis of international guides for news coverage of migrations. The research questions are: RQ1: What are the functions that journalism should fulfill in a country hosting massive and forced migrations? RQ2: What must the journalist's task be when a predominant political discourse discriminates and labels the migrant as a threat? RQ3: How should journalists overcome negative and emergency framing in the face of massive and forced migration? --- Venezuelan Migration in Colombia In Colombia, for a long time, especially during the hardest period of the war against drug trafficking and the armed conflict that had the country and the cities in check and kept extensive territories out of State control, the country's image was that of a dangerous territory and many governments recommended their citizens not to visit it . This situation has improved, especially since the strengthening of the Colombian state against the armed rebels and since the signing of peace treaties with the FARC, the main guerrilla group, in 2016. Colombians have even become increasingly accustomed to receiving tourists. For 2019, the country received a record number of foreign visitors . In February 2022, the Colombian government agency for migration estimated that there are about 2.5 million Venezuelan migrants. In contrast, in 2014, around 23,000 Venezuelans resided in Colombia . This is, therefore, an unprecedented situation for both countries. --- Journalism Responsibility Journalism is one of the pillars for the construction of a more democratic society, as it provides citizens with information that can be useful for making better decisions and selects an agenda of issues that should be of interest to both citizens and public institutions. The media's role as guardian of public power requires the audience to believe that the media are their representatives, based on a relationship of accountability and trust. The public must believe that the media are not complicit with political or economic powers and, furthermore, the public must trust that the media can represent the world in a reasonable, fair, and impartial manner . Lamuedra argues that the metaphor of the press as an "impartial window" is challenged and qualified as a kind of "imperfect window" by considering the approach to news as a construction affected by the working conditions, routines and values of journalists and their media in prioritising a media agenda as a framework in which a dominant framing of facts, issues, sources of information and certain points of view is presented. This situation is reinforced by the concentration of media ownership in monopolies and oligopolies, which causes a ventriloquist effect , whereby the same owner has multiple media outlets to disseminate a single discourse, which responds to a uniformity of news criteria, a specific thematic agenda and gives voice to the same predominant sources of information. Thus, media concentration goes against the pluralism of information, infringes the right to information and the freedom of expression and opinion . According to a study by Berganza et al. that surveyed a sample of Spanish journalists to find out what roles are inferred from the professional functions they perform and what importance they attach to the different types of roles they assume, three traditional roles stand out: the broadcaster, the adversary or watchdog and the lawyer or audience trainer. Three emerging roles are also identified that provide evidence of an evolutionary process in the social function of journalism in contemporary society: "the speaker of the citizenry, the entertainer of public opinion and the promoter of the status quo." The study found that journalists most often identify with the role of "speaker for the citizenry." Berganza et al. defined this role of journalism as a civic function of the profession, aimed at promoting tolerance, diversity, and dialogue, through stories about the context of which it is part, especially in "the coverage of citizens' demands, their rights and their duties" . --- Media Influence The mass media influence the definition of the issues that are discussed by the citizenry and how these issues are discussed by prioritizing certain voices and points of view in their coverage of events considered newsworthy . The relationship between media and society has always been problematic in two ways. Firstly, the media control and influence society, through priming, agenda setting and framing, while, secondly, the vast changes and dynamics of information technology render media rules and regulations obsolete . "News shapes audiences" views of people and events beyond their immediate physical environment. Since the mass migration of refugees from Syria represents one of the worst humanitarian crises in modern history, its news coverage necessarily shaped the way global audiences understood the crisis. The media, like other types of organisations, are not exempt from committing misconduct such as fraud, theft, misinformation, broken promises and broken laws or regulations . This behaviour generally results in a lack of social trust towards media corporations, even though many of them invest in public relations strategies . Mistrust is increasing in the digital era as media companies have been gradually growing, and even their ownership has been concentrated in giant international empires, who's financial and ownership structures are opaque in the eyes of society . In the public sphere, there is competition among different social actors to make their proposals and objectives visible in the public space through their influence on the media agenda and in the media's coverage. In addition, the editorial line of the media has an impact on the coverage of social issues of great importance . Furthermore, there are the cognitive biases of journalists described by Entman as slant: "Slant characterizes individual news reports and editorials in which the framing favors one side over the other in a current or potential dispute" . This type of framing can contribute to polarization by reinforcing readers' or viewers' prior convictions: "People tend to select attitude-congruent news media, which would mean conflict has the potential to continuously magnify polarization" . Conflict is one of the predominant types of informative framing in the media . Eberl et al. describes how this applies to migration: "While migrant groups are generally underrepresented, when they are present in the media, they are often framed as either economic, cultural, or criminal threats and thus covered in a highly unfavourable way." Although we analysed the issue of Venezuelan migration from the perspective of its journalistic coverage in Colombia's media, it is necessary to recognizethat the country, although part of the Global South, is integrated into the logic of globalization and that its constitutional order and institutional framework safeguard the right to information as a fundamental right, which brings it into line with Western standards of free journalism. This normative point of view: Generally, privilege Western journalistic norms, which were in turn exported to various parts of the world through training and textbooks, as well as through Western media capital and ownership, and which have been accelerated in recent years by globalization and economic liberalization in many countries of the Global South. According to Dimitrova , the emphasis on the value of freedom must also be contrasted with the social responsibility of both media and journalists: Another way to classify media within a specific country is by evaluating its rankings on a global media freedom scale and then comparing that to the existing levels of media responsibility. While some countries, particularly in the Western world, prioritize journalistic freedom, others put an emphasis on media ethics and social responsibility. --- Method First, we recognize that our positionality shaped our approach to the present study. The authors of this paper are two Venezuelan migrants with a Ph.D. degree obtained in Spain. We are journalism professors who work in academia as full-time faculty members. We work as professors and researchers in two universities, one in Spain and one in Colombia. Therefore, we are part of the international migrant community, potentially affected by the xenophobia described in this article. The present study is qualitative in design and exploratory in scope. To answer the research questions, first, we conducted a literature review on migration journalism, the influence of the media on public opinion and the social responsibility of the media. This literature review was based on a search of documents in databases such as Web of Science and Scopus, taking special relevance to those emerging documents that proposed guidelines or recommendations on the informative coverage of migration issues. To perform this search, we used the criteria and Booleans Journalism [AND] Migration, Journalism [AND] Responsibility and Journalism [AND] Agenda Setting, organizing the results only in journal articles and prioritizing emerging documents from Ibero-American countries. In a second stage, in-depth interviews were conducted with 11 journalists and 10 researchers from Colombia, Venezuela, Chile, Peru, and El Salvador, employing purposive theoretical sampling . Journalists and researchers from these countries were considered because they are the main recipients or historical origin of large migratory movements in Latin America, which has forced the leading media to have, in many cases, specialized action protocols and style manuals to raise awareness of migration among the population. Second, we conducted a Qualitative Document Analysis of 10 reference guides for news coverage of international migration to contrast and compile their recommendations. The Information emerging from the interviews and the documents from the literature review was triangulated, making it possible to contrast the theoretical foundations and the perceptions of working journalists and experts in the field of migration to obtain a sufficient contextualization of the phenomenon under study from the journalistic praxis and the results of previous research. --- In-Depth Interviews The in-depth interview is a very useful technique to learn about the perspectives and experiences of the respondents, as well as their interpretations, feelings, and perspectives . It also seeks to delve into a given situation from a more holistic and broad perspective, as it goes to the details and provides the "big picture" of a series of scenarios, situations, or people . The application of the in-depth interview in this study is justified by the need to provide explanatory elements and indications of possible new prospective phenomena that serve as study variables for future research . We used purposive sampling to select participants since the aim of the study was to open a space for debate in which perceptions are not confined to strict theoretical categories. In this sense, the sampling strategy was intentional and judgmental. The criteria we used to select participants have been prioritized based on conceptual criteria alone, following principles of structural representativeness. Therefore, the researcher theoretically defines the variables that delimit the structural composition of the sample . The sample selection criteria are shown in Table 1. These criteria were intended to reflect the responsibility of the subjects in the field of journalism and journalism research, meeting the principles of suitability and relevance of the profile of the informants exclusively. Potential participants were recruited through direct contact with the researchers via email, on the understanding that for this analysis, the sample selection is non-probabilistic and intentional since we were looking for experts specializing in migration who reside in Latin American countries with migration crises. The Idea of interviewing both active journalists specialized in migration issues and NGO representatives from both countries is based on the fact that while the former can explain what is being done to report on the migratory phenomenon and to avoid xenophobia in the country receiving migrants, the NGO representatives and university professors can explain what else could be done to avoid the stigmatization of migrants, from a perspective more focused on the needs of these migrants. Thirty subjects were initially selected and invited to the research . Of these, only 21 confirmed their intention to participate by providing their informed consent by telematic means. The research focuses on understanding the work experience and specialized knowledge of the selected professionals. The professionals who chose to participate gave their written consent by email. As a preliminary step, they were informed in writing about the objectives and methodology of the investigation, and they were offered the confidentiality and protection of their personal data. As personal data would be collected in the interviews, the IRBs of both institutions of affiliation of the investigators approved this study. --- Table 1 Inclusion Criteria for Informants --- Researchers Journalists NGO Representatives • More than 10 years of experience in international migration research. --- • Residence in a Latin American country with a migration crisis . • Published research on migration, preferably in scientific articles published in the last 5 years. • More than 10 years of experience working in the media . --- • Residence in a Latin American country with a migration crisis . • Experiences dealing with information on migration crises. • Independent media workers • Directors of NGOs working on migration crisis issues. • The NGO they represent has more than 10 years of existence and proven work in the field of aid to refugees, displaced persons and asylum seekers. The in-depth interviewing process is shown in Figure 1. Individual interviews were conducted virtually and audio recorded. The interviews lasted between 40 and 60 minutes. The transcripts of the interviews were anonymized and stored in an encrypted, password-protected digital file to protect participant confidentiality. These interviewees were sufficient to reach the saturation point, understood as the point at which the informants begin to repeat the responses of others , so they do not enrich the research process by contributing novel data. This saturation point was reached between interview 18 and interview 20. We conducted the interviews between October and December 2019. The informants were consulted through five open-ended questions: a. What is the role of journalism in a country hosting massive and forced migrations? b. What should the journalist do when a political discourse that discriminates and labels the migrant as a threat predominates? c. In everyday journalism, a negative framing of migration predominates, presenting the migrant as a threat to security, as a risk to public health, and as a labor competition. What could journalists do to improve the coverage of migration in this sense? d. What specific training should a journalist receive to cover the phenomenon of forced migration adequately? e. Which competencies do you consider most important to cover the phenomenon of forced migration adequately? As seen in Figure 1, the in-depth interview process followed a four-stage protocol. In the first phase, the researchers identified and listed potential informants who met the profile of inclusion criteria explained in Table 1. To make this list, Latin American researchers specializing in migration journalism were sought from migration research groups at different universities and internationally renowned authors who had published articles on migration journalism in scientific journals in the Scopus database or the Journal Citation Reports. To make the list of journalists specializing in migration journalism in Latin American countries with migration crises, we focused on the case of Venezuela, Colombia, Ecuador, and Peru, countries that have been experiencing a massive exodus of Venezuelans since 2016. In this sense, journalists were selected from print and digital media that specialize in migration information and have been writing about the topic for more than five years. Likewise, to choose and list the heads of NGOs, we focused on organizations from Venezuela and Colombia, as these countries have the most cross-border transit points. For the selection of these NGOs, specialization criteria were taken into consideration, i.e., NGOs that work directly with individuals and organizations that help migrants in their adaptation process to the new country and that defend their rights before public and private organizations. While listing potential informants in the three sample clusters, the five questions explained above in Figure 1 were drafted, all focusing on the role that journalism should play in the face of migration crises. The second phase of the process began with sending an email to the 30 informants selected in phase 1, asking them if they wished to participate in this study. This email explained the research objectives, the university institutions linked to the study, and the possibility that the results would be published in reports and scientific articles. Of the 30 informants who were sent the invitation email, 21 responded affirmatively. Next, the 21 informants who responded affirmatively to wanting to participate in the study were sent a document detailing the following: 1. Date and time that the interview would take place and the videoconference program through which it would take place . 2. Personal data that would be kept from the interview, such as name, surname, and position. 3. Informed consent that the interview would be recorded only to transcribe the research results. In the third phase, the researchers conducted the interviews via Skype. At the beginning of the interview, the participants were reminded again that it would be recorded and that if they had any reservations that these recordings be kept only for transcription for this research. Independently the interviews were recorded, which were conducted through the exploration of the five main questions of this investigation, the researchers, as interviewers, took notes of the most important aspects of this to have a more precise reference of the extracts that more could serve when showing results of each one. Finally, in the fourth phase, the authors of this article transcribed the 21 in-depth interviews, with particular emphasis on those responses that were more closely linked to the objectives of this research. --- Qualitative Document Analysis of Guidelines on Migration Coverage To contrast and deepen the experts' answers, a documentary review of 10 international reference guides and manuals was carried out , which aim to guide journalism professionals so that their work in the informative coverage of international migration responds to values such as justice and human rights. According to Wood et al. , Qualitative Document Analysis is a method widely used as a complementary tool to other methods in qualitative research and "provides a systematic methodological process for eliciting meaning from documentary evidence" . According to Ulrich , document-based research has multiple advantages, such as accessibility through digital platforms, credibility of sources, efficiency in the collection and systematization of data, and reduction of ethical conflicts typical of other research designs. After selecting a sample of texts, according to criteria such as authenticity, credibility, representativeness and meaning , the next step was to conduct a thematic analysis of the texts, oriented around the categories of analysis that seek to answer the research questions of this study: social function of journalism , media coverage of migration, hate speech, discrimination and stereotypes , and the role that the journalist must play when dealing with negative news about massive forced migrations. Two coders, who are also authors of this research, participated in this process, described in Figure 2. --- Findings The Role of Journalism in a Host Country of Massive and Forced Migration Historically, Colombia has been a country of emigrants, a territory that expels the population. Recently, however, Colombia has received a large population influx in an intense manner, that is, in a short time. This includes, since 2016, Venezuelan migration estimated at 2,477,588 people , according to data from the state agency Ministerio de Relaciones Exteriore . Several of the informants agree that, to fight xenophobia, it is important to help people understand that the phenomenon of human mobility is dynamic and that in a previous era it was Colombian nationals who had to look for opportunities in other countries. One of the journalists says that in her work she seeks to: [J04] To be an anti-xenophobia instrument. That is my goal and I do my best to create content that leads to that end. We cannot enter into a discourse against migrants or see them as an enemy because they are nothing more than victims of an ideology and a failed political system that led them to this crisis. We have to repeat to audiences that migrants are here because they had no other choice, because they were forced to leave their country [...] as journalists we cannot be indifferent. In our work we must avoid adjectives, look for a counter-current that says the opposite, remember the conditions in which these people are living in our country, and show successful cases of integration. Question politicians who use migration as a xenophobic discourse to get votes. One of the researchers added that journalists who address the issue of migration should do so with empathy and within the framework of recognizing migrants as subjects of rights: [R03] Journalists should always avoid generalising and stigmatising, should be careful not to portray the migrant as a victimiser, and should be empathetic. They could perhaps show "positive or success stories" of foreigners in the host country, focusing on the contribution that the migrant offers or could offer to that culture or society. The journalist should also delve into human rights issues in order to be able to cover them from that perspective. Journalists covering migration issues should think ethically, and fulfil a public service function. The Informative work of the media takes on greater relevance due to the need to help their audiences understand a social phenomenon that is changing the country and, especially, to help combat xenophobia and discrimination against new neighbors: [R10] Migration is a multidisciplinary phenomenon, so it must be approached from many facets. The idea is not to show only one economic side, but to show a social problem that has a number of social aspects. Much of the journalism on foreigners sees the migrant as a different other, ignoring the links they have with that country, but also the fact that a large part of the population that has arrived today is a returning population, that is, citizens from here who are returning. This is where the media must move towards, to show all the nuances and to show the phenomenon in its complexity and in a comprehensive way, and for this, many voices, and different ways of approaching migration are required, obviously with care in the language and a clear look at these processes and with social integration in mind. In this context, the informants highlight a variety of functions that the media and journalists should fulfill , including responsible journalism, combatting racism and xenophobia, and conveying that migration is a complex phenomenon. --- Table 3 Media and Journalistic Functions --- Responsible Journalism -The function of journalism should be the same as always: to inform, but with much more care and empathy for the representation of the otherness. -Inform well and broadly about what is happening. Migratory issues are complex, which requires a plurality of views. -It is crucial to verify and contrast the information, as fake news about immigrants is often created for political interests. -Try to cover the news associated with the topic as rigorously as possible. Use figures and indicators as close to reality as possible, if available, without misrepresenting them. -Consult with specialists in migration issues, both national and foreign, who can give a scientific opinion on this socio-demographic phenomenon. --- Combating 19otos19o and Xenophobia -To make known the characteristics of those who wish to integrate into the destination countries and the benefits. -Avoid stigmatizing the migration process and present the various reasons why people decide to migrate. -Showing life stories helps the population to understand better and achieve empathy for the foreigner. -Establish views without prejudices, clichés, or stereotypes. -Inform professionally and with a diversity of sources without portraying immigrants as criminals. -Integration should be encouraged because migration, in many cases, is not temporary. -The media should not replicate xenophobic or discriminatory language. The role of the media must be formative. -Be cautious with the words chosen to in writing. The use of certain words can negatively affect the entire community of displaced persons and forced immigrants. These words can be, for example: problems, crime, poverty, crisis. --- Helping to Understand Migration as a Complex Phenomenon -Educate and go beyond the immediate news and social alarm. A country that is receiving migration needs to understand the phenomenon it is experiencing and understand the needs of the people who are arriving. -Tell the stories of migration with rigor and depth. The Qualitative Report 2023 -Explore the causes and consequences of the migration phenomenon. -Contextualize and tell the stories with respect and without prejudice, avoiding the simplistic vision of the media spectacle. -Repeat often to audiences that migrants are here out of necessity and because they were forced to leave their country. -Offer information and invite reflection on the causes of migration. -Provide multiple views that account for the complexity of migration in all its dimensions . --- The Journalist's Job When a Political Discourse that Discriminates and Labels Migrants as a Threat Predominates According to Kaufman , hate speech is characterized by referring to historically discriminated groups, by its aggressive or humiliating nature, by inviting others to despise and denigrate against the group in question, and by its clear intentionality to exclude or harm. Following Bustos et al. , currently, the rise and popularization of social networks have caused their dissemination to be practically unlimited, so cyberhate is increasingly explicit due to the difficulty in identifying the authors, who often hide behind anonymity. In contrast, journalistic content must respond to ethical criteria under the logic of a profession conceived as a public service: [R01] Journalists must present other versions and unmask, with solid arguments, the strategies used by the political discourse to discriminate and label migrants and to stigmatise the migration process. In his or her work, journalists should present the different reasons why the majority of migrants decide to make this decision, as well as show the collective and individual changes that this process entails, both for the sending and receiving countries. It is also useful to provide multiple perspectives that account for the complexity of this process in all its dimensions. It is very important for the professional to recognize the role he or she plays as a journalist in shaping an informed and critical society. In this sense, the journalist must be informed and qualified to understand the implications of this process and, above all, to inform about good practice initiatives that show the complexity of the migration process. Both journalists and the media have an essential responsibility in the face of the proliferation of hate speech , especially when developing news coverage on migration: [J05] Journalists must narrate migration stories with rigour and depth. Explore the causes and consequences of the migration phenomenon. Contextualize. Tell stories with respect and without prejudice. Avoid contributing to a biased and xenophobic view: tell the whole story, with multiple perspectives, contrasting and verifying data and facts. The experts highlighted several strategies to avoid framing migrants as threats : -Avoid stereotyped and pejorative charges against the migrant population. -Address xenophobic statements bypresenting other versions and unmasking with solid arguments the strategies used in political discourse to discriminate and label the migrant. -Explain through factual data how the migrant population can become a growth opportunity for the receiving country. -Present the migratory movement as an opportunity for the development of the values of coexistence and integration. Show successful cases of integration. -Avoid qualifying adjectives. -Offer broader contexts than the anecdotal and circumstantial. -Seek truthful information so as not to fall into false interpretations or conclusions, or failing that, try to be neutral until the information is corroborated. -Avoid generalizations so as not to fall into discrimination, discredit, and sow repudiation towards an entire group. -Avoid becoming a sounding board for xenophobic and discriminatory speeches, even when this information is of interest to many media because it generates a lot of traffic and virality in social networks and, consequently, economic return. -If there is a political environment that discriminates against migrants, it is precisely the journalist's role to show the other reality, question these discourses, and play a critical role. In the proposals made by the experts, the need for quality journalistic work that goes beyond the repetition of statements from sources, almost always official, stands out. Most of the journalistic content published every day is classified as "conventional journalism," characterized by its brevity, by the few sources consulted, almost always of an official nature, and, especially, by the fact that it is often limited to simply transcribing what the sources say . In this regard, one journalist emphasizes the primary need to recognize the complexity of the journalistic work on the migration phenomenon: [J01] Mass and forced migrations, such as the Venezuelan one, have become a media phenomenon. Hence, the role of the media and journalists in the face of this phenomenon is key in shaping the currents of opinion and raising public awareness of the arrival of immigrants. The role of journalism in this type of situation is not only to inform, but also to analyse and guide. Journalism must explain the causes of this migration, not stigmatise the immigrant, avoid information bias and, when the case merits it, present the migratory movement as an opportunity for the development of the values of coexistence and integration. Journalists should be formed in data journalism to analyse the figures on migratory flows, as well as the sources, in order to obtain a global overview of the issue that is as close to reality as possible. Also, in solutions journalism to not only expose the problem, but to show the initiatives that exist or are being developed to positively manage the massive flow of Venezuelans and to participate in workshops to tell the specific stories about migration in a different way, to show the faces, how they live, their aspirations and their integration into the country that welcomes them. --- Overcoming Negative and Emergency Framing In contemporary society, the media continue to play a central role as a source of information that potentially reaches the whole of society. Castells explains that information from the mainstream media also influences media of lesser impact and other potential influencers, mainly through three functions: agenda-setting, priming, and framing . This research therefore considers the broad framework of quality journalism studies from a global perspective, which recognizes the influence of the process of economic and cultural globalization on journalistic practices and the values that guide the profession . In this regard, journalists must provide information and invite reflection on the causes of migration, understanding society as a globalized phenomenon: [R09] Journalists, by vocation, owe their work to the facts and, consequently, to the search for the truth. Communicating implies generating communities and this is only possible by favouring the understanding of reality. Offering information and inviting reflection on the causes of migration; making known the identity of those who wish to integrate in the countries of destination, as well as the benefits that this entails; establishing views without prejudice. In a globalised society, the journalist must have the tools to recognise the different sources of information, confirm and contrast them; approach the facts with openness and recognition of the phenomenon in all its factors; as well as openness and ethics in the treatment of information in relation to a situation that generates suffering for the population of the country of origin and concern for the region of destination. The experts made a series of proposals that contribute to improving news coverage of migration and overcoming the typical negative or emergency approach to the phenomenon, which can lead to fearful public opinion. These strategies avoid using immigration as a synonym for social emergency . --- Table 5 Strategies to Avoid Using Immigration as a Synonym for Social Emergency -Avoid, as far as possible, the pressure of immediacy. The treatment of this complex and delicate phenomenon requires a great deal of research and insights from recognized experts in the field. -Guarantee the balance of sources and to avoid opinionated and subjective charges towards migrants. -Pay attention to scientific research on migratory processes in addition to consulting experts. Complex problems require other angles and other sources. -See training to understand the implications of forced immigration and to publicize good practice initiatives. -Learn about refugee issues, pendular migrants, regular migrants, irregular migrants. -Avoid sensationalist headlines that guarantee click-baiting. -Dismantle harmful myths about the effect of forced migrations based on statistics and demographics. -Promote the design and execution of public policies that help overcome any adverse situation from forced migration. -Investigate and report when the immigrant is not allowed to transition to a better life, with a minimum of dignified conditions. -Interview immigrants to contribute to humanizing them, to give them a voice and existence that is denied to them from the generalization of information. As can be inferred from the recommendations in Table 5, high-quality news coverage of international migration requires that journalists have access to ongoing training that allows them to strengthen their skills and competencies so that they can contribute to understanding and avoid hate speech and discriminatory discourse: [J07] Journalists should become aware of the damage they can do to a population with their comments and opinions. On the other hand, they should educate themselves about key concepts such as refugees, pendular migrants, regular migrants, irregular migrants and also stop click bating strategies. Another important thing is to always verify that the facts or data you are giving about the migrant population are real. The Qualitative Document Analysis of the ten guides on media coverage of international migration showed that they identify ten primary normative criteria that guide the work of journalism professionals in addressing the issue of migration, which are visualized, as a summary, in Figure 3 below. All the guides recognize the need to build media content that does not deepen the separation between "Us" and "Them" . According to van Dijk , this distinction between people born in a territory and those born abroad is at the base of racist discourses. --- Figure 3 Recommendations of the Guidelines for Migration News Coverage The guides invite journalists to overcome stereotypes, stigmas, and labels that reinforce negative imaginaries towards migrants. Therefore, they recommend that journalists avoid generalizations when referring to this diverse population and avoid adjectives such as illegal immigrants and terms referring to natural catastrophes. These recommendations agree with the opinions of the experts consulted and are also supported by conclusions that have been noted in the scientific literature . . 3. Know the normative framework of migrations, both national and international, especially treaties on Human Rights. The labour migration guidance for journalists of ILO1 presents all conventions and recommendations concerning international legal standards that are important "for safeguarding the dignity and rights of migrant workers" . 4. Make responsible use of images, both photographs, and videos. From some primary ethical considerations, especially on recognizing minors and people in vulnerable situations . 5. Rigorousness in work with information sources. From the selection of valid sources and, in a special way, highlighting the need to contrast and verify the information collected . 6. Caution against hate speech and xenophobia. Based on recognizing an ethical commitment, journalism can avoid repeating or transcribing insulting statements or generalizations that stigmatize a vulnerable group . 7. Understand the complexity of the migratory phenomenon, the causes that cause displacement, the motivations of migrants, the responsibilities of national and international entities, the desired and undesired effects of human mobility for the host, transit, and origin territories. 8. Be aware of the gender perspective, which allows to adequately visualize women, how they are affected in a differentiated manner, their rights, needs, and struggles for equity . 9. Data and visualization journalism requires the skills to understand the main figures of a problem and translate them for proper understanding by the target audiences. 10. Social responsibility and self-regulation of the profession, following democratic principles, require journalism to ensure the right to information and opinion responsibly and for the common good . According to the guides reviewed in this research, these recommendations should be assumed from empathy with the other, with the social sensitivity of professional ethics. In addition, the need for responsible and critical journalism, oriented from the human rights perspective, is highlighted, especially in a country that has recently become a recipient of migrations. Journalists must carefully prepare headlines for journalistic content on migrations since most users only read the headlines. The use of images must be responsible, ethical, and consistent with text content. When registering the images, journalists must also respect the right to privacy of migrants, especially minors. Migrants may be reluctant to be photographed for cultural reasons or fear of reprisals against them when fleeing from violence. Different guides insist on the need to consult the sources of information necessary to understand the phenomenon to narrate so that the journalists present a variety of points of view, including the migrants themselves as sources of information when trying to create content on international migration, and also hear the voice of host communities. Journalists must always avoid reproducing hate and racist speech, which are the basis of discrimination and violence. --- Discussion and Conclusions The intensity of migration has significant consequences on the host societies since human mobility is a complex phenomenon that generates controversy, polarization, and demonization . The work of the media positions an agenda of issues and highlights specific attributes ranging from how they describe migration, the predominant discourses, which actors are given a voice, how the information is framed, and even in the selection of images used to illustrate the text . According to Theorin et al. , negative portrayals of migration predominate in the media, who depict it as a threat to the security and economic well-being of the receiving countries. Media coverage of migrations "continues a long history of underrepresentation and misrepresentation of underserved racial/ethnic groups" . This undoubtedly affects social perceptions, influences public opinion, and affects relations with migrants' people . Exposure to predominantly negative media content is linked to anti-immigrant and xenophobic attitudes. Negative media "produce negative emotions, encourage attitudes that dehumanize immigrants, sway support for immigration policy, and prompt harming behaviors, under certain conditions" . According to Lecheler et al. , exposure to certain news frames-traditionally conflictivecauses emotional reactions that influence citizens' opinions, attitudes, and behaviors. This means that there are media outlets that seek to portray the negative, the danger, the alert. This news coverage generates damage in the perception of otherness, that is, in this case, of migrants. According to interviewed participants, journalistic coverage of migration, especially of Venezuelan migration and its complex humanitarian crisis, has been an enormous challenge for journalists in all the countries involved. This challenge is related to their capacity to fulfill the main functions that journalism must fulfill in the social context of a country that receives forced migrations: the commitment to responsible journalistic work that meets the quality standards of the profession. These standards require that their work be oriented towards the search for the truth through verification and contrast of all data and statements collected and conducted with honesty that allows identification of the limitations of the results obtained. In this context, the informative work of the media takes on greater relevance due to the need to help their audiences understand this social phenomenon and, especially, to help combat xenophobia and discrimination against foreigners. The fulfillment of this social commitment requires responsible journalism, in which information is verified with honesty so that journalists do not play a role in the face of the proliferation of hate speeches when developing informative coverage on migrations. As confirmed by , in daily journalism, especially in the news, there is a predominantly negative framing of migrations, which presents the migrant as a threat to security, as labor competition, and as the recipients of social aid that could be allocated to the native population. The experts consulted suggest that it is necessary to improve the news coverage of migrations and overcome a negative and emergency frame, doing investigative journalism, slow journalism, and, above all, works that deepen the humanization of traditional frames and encourage empathy, from the understanding that, as humans, we are all migrants. This work requires an awareness of journalists and ongoing training, not only in the mastery of digital skills , which allows them to understand the cultural, aesthetic, and social dimension of the communicative processes and fieldwork with people in vulnerable situations, from the values and respect for Human Rights. Through media literacy that reinforces the critical sense of journalists in the face of the statements of the representatives of public authorities and their ability to curate the overabundance of information currently available in favor of themselves and their audiences, journalists will be able to strengthen their handling of languages and narratives and incorporate and adequately manage processes of interaction with users, typical of digital scenarios. The literacy and intercultural dialogue processes are an imperative need and a cross-cutting field for collaborative work between the media, professional associations, universities, and the social sector, to promote continuous training sessions and foster spaces for cooperation, such as observatories on migration and against hate speech. It Is evident that the work of the media positions an agenda of Issues and highlights specific attributes, from how they report on migration, the predominant discourses, which actors are given a voice, how the information is framed, and even in the selection of photography to illustrate the text. This can influence social perceptions and affect relations with migrants . The experts consulted confirm that the primary news sources on migration are government officials or spokespersons of the political or economic powers, while migrants are rarely mentioned in news reports. In media discourses on the recent Venezuelan migratory phenomenon, there is more talk of Venezuelans than of migrants, and even the word "Venezuelan" is presented as a substitute for immigrant, as if it were not possible to find Venezuelans in another situation, as a manifestation of a phenomenon that Cortina recognizes as "aporophobia," whereby the poor are rejected and not so much the foreigner. Canelón and Almansa stated that overcoming the invisibility of migrants is an advance towards an approach that allows recognizing them in their dignity as subjects of rights and an opportunity for mutual enrichment. In scenarios of forced migration, the need for journalism committed to a human rights approach that helps to understand and contrast the risks of disinformation and hate speech is reinforced, following the recommendations of the ethical guidelines of international entities for media coverage of international migration, as part of the ethical commitment of journalism in the fight against speeches that serve to justify violence and discrimination. A line of work of great interest to give continuity to this research, due to its repercussion and, especially, its multiple interactions, could be to contrast the analysis of media agendas with the content generated by users on digital social network platforms, as a reaction to journalistic coverage, due to its connection with the problem of disinformation and the dispersion of hate speech. It is also necessary to compare other territories with forced migrations, such as Southeast Asia, sub-Saharan Africa, Australia, and the United States. --- Article Citation Lotero-Echeverri, G., & Romero-Rodríguez, L. M. . Journalism on forced migration in Latin America: Recommendations from experts and international journalism guides from a qualitative study. The Qualitative Report, 28, 1473-1495. https://doi.org/10.46743/2160-3715/2023.6061

Wars, insurgent groups, dictatorships, and economic crises are the main reasons for forced migration. Displaced persons, asylum seekers, and refugees often face public stigmatization, as they are treated by the media as a social problem and, in many cases, seen as economic and social threats. This article presents the results of in-depth interviews with expert journalists and researchers from different Latin American countries on the phenomenon of forced migration and its journalistic coverage. Their recommendations are complemented by Qualitative Document Analysis (QDA) of international guides on migration journalism. The findings highlight the need for training and awareness-raising in critical skills and social analysis for journalists to understand the complexity of human mobility and approach it from a human rights perspective, as well as to project the positive contributions of migrants to the host society. Findings suggest the need to avoid disseminating and magnifying hate speech, which may be used to justify discrimination and violence against migrants. This highlights the fact that migration issues must be dealt with under slow journalism schemes, with an emphasis on research, thus avoiding falling into xenophobic discourse due to the very immediacy that the digital ecosystem demands.

Introduction Children living with 1 or more smoking parents are at increased risk for involuntary smoke exposure, which can cause many negative health outcomes . Children living with smokers are also more likely to smoke, even after accounting for other sociodemographic factors . A smoke-free environment can not only protect children from exposure to involuntary smoke, but also convey an antitobacco social norm that prevents and reduces smoking behavior in the future . The home, the car, and children's outdoor play areas are 3 primary sources of involuntary smoke exposure for children . Half of US households with underage children and at least 1 smoking parent did not have a voluntary smoke-free home rule by 2007, versus more than 90% of households with children but no smoking parents . However, it is unclear what the prevalence has become in recent years. Furthermore, research has examined voluntary smoke-free rules in cars and has found overwhelming support among the US general population for voluntary smoke-free car policies when children are present . Yet, support for voluntary car rules has not been examined among house-holds with children and at least 1 smoking parent, a vulnerable subset of households. Even less is known about attitudes toward smoking restrictions in outdoor children's play areas like playgrounds and sport fields. The 2010-2011 Tobacco Use Supplement to the Current Population Survey collected information on respondents' reports of voluntary smoke-free home rules and attitudes related to smoking restrictions in cars and outdoor children's play areas. Attitudes are theoretical determinants and good predictors of behaviors and also reflect levels of support for future legislation to expand public smoke-free laws. Using data from this survey, we aimed to update the prevalence of voluntary smoke-free home rules and examine the distribution of parental attitudes related to voluntary smoke-free rules in cars and outdoor children's play areas among US households with underage children and 1 or more smoking parents during 2010-2011. We also investigated parental and household factors associated with these outcomes. --- Methods --- Study population We used data from the 2010-2011 TUS-CPS. The TUS-CPS is a survey of tobacco use that is administered as part of the US Census Bureau and the Bureau of Labor Statistics' CPS. A large sample of households is surveyed during a survey period that provides representative data on tobacco-related behaviors, norms, and attitudes at the national and state levels among the US civilian noninstitutionalized population. All permanent household members aged 18 years or older are eligible for the interview. We included only primary family households with underage children and at least 1 parent who smoked in our analysis. We defined a parent as a household's reference person or his or her spouse, which means this person may not be the child's biological parent. Some items from the TUS-CPS may be self-reported or provided by proxy , whereas the practices and attitudes related to voluntary smoke-free rules in home, car, and outdoor play areas were only self-reported. Therefore, we limited our study sample to respondents who provided self-reported data. --- Measures We defined a smoking parent as the reference person of a household or his or her spouse who had smoked 100 cigarettes or more and was smoking every day or some days at the time of the interview . All self-respondents of the TUS-CPS were asked "Which statement best describes the rules about smoking inside your home?" We categorized the report of "No one is allowed to smoke anywhere inside your home" as a voluntary smoke-free home rule. A smoke-free household is a single-parent household in which the parent reports a voluntary smoke-free home rule or a 2-parent household in which at least 1 parent reports voluntary smoke-free home rules. A previous study indicated that discordance between parental reports of home rule decreased over time and has dropped to below 5% among 2-parent households with children and at least 1 smoking parent by 2007 , making the use of a single parental report reliable. Parents were asked, "Inside a car, when there are other people present, do you think that smoking should always be allowed, be allowed under some conditions, or never be allowed?" Those selecting the first 2 options were further asked, "If children are present inside the car, do you think that smoking should always be allowed, be allowed under some conditions, or never be allowed?" We defined a positive attitude toward a voluntary smoke-free car rule when children are present as the report of smoking should never be allowed when there are other people present or if children are present inside the car. Hereafter, this attitude is referred to as support for a voluntary smoke-free car rule. The TUS-CPS asked respondents "On outdoor children's playgrounds and outdoor children's sport fields, do you think that smoking should be allowed in all areas, allowed in some areas, or not allowed at all?" The response "not allowed at all" was defined as an indicator of support for a rule for voluntary smoke-free outdoor children's play areas. We analyzed the following parental and household factors potentially associated with the establishment of smoking restrictions: household structure, highest parental education level, race/ethnicity, age, smoking status, age of the youngest child, and annual household income. We also included state identifiers to capture unmeasured state-level factors, such as statewide tobacco control policies, which may be associated with the outcomes. --- Statistical analysis We calculated descriptive statistics on the prevalence of parental practice and attitudes regarding smoke-free rules in the home, the car, and outdoor children's play areas. We conducted multivariate logistic regressions to examine associated parental and household factors with the report of smoke-free rules. For the existence of smoke-free home rules, the analyses were performed on the household level; for attitudes toward smoke-free rules in the car and outdoor children's play areas, the analyses were performed on the individual level. Survey weights provided by the TUS-CPS were used, and clustered standard errors were estimated to account for the complex sampling design and clustering to produce population estimates. We performed all analyses using the software STATA/MP version 13.0 , and significance was set at P < .05. --- Results The overall response rate for the 2010-2011 TUS-CPS was 74% . The rate for self-respondents was 56% . Our study sample included parents from 8,083 households, of which 37.8% were single-parent households and 62.2% 2-parent households . Overall, 60.1% of households with children and 1 or more smoking parent established a voluntary smoke-free home rule inside the home. Most parents from these households thought that smoking should never be allowed inside cars when children are present, including 72.2% of those who did not have a voluntary smoke-free rule in their homes . Fewer parents were supportive of smoke-free rules in outdoor children's play areas, which included 61.3% of those not from a smoke-free household. On the household level, children whose parents were single fathers or mothers, had not completed high school education, were non-Hispanic African American, or were aged 40 years or older were less likely to be protected by a voluntary smoke-free home rule than were their counterparts . Households with an annual income less than $25,000 and with a child aged 6 years or older were also less likely to institute a voluntary smokefree home rule . A positive attitude toward a voluntary smoke-free car rule when children are present was more likely to be reported by individuals from 2-parent households, who had a college education, who were non-Hispanic African American or Hispanic, with an annual household income of $50,000 or more, and who had infants living in the home . Similar predictors were observed for a supportive attitude toward smoke-free outdoor children's play areas. Compared with non-Hispanic white parents, non-Hispanic African American and Hispanic parents were more likely to express that smoking should be banned in outdoor children's play areas . The attitude was also more likely to be reported by people from 2-parent households than by those from single-parent households and by those who had a college degree than by those who did not . In additional analyses, we found that support for smoke-free rules was stronger from nonsmoker households with children; 96.3% of parents from these households advocated voluntary smoke-free rules in cars, and 87.4% of parents from these households advocated voluntary smoke-free rules in outdoor areas . --- Discussion We found that more than one-third of US households with children and 1 or more smoking parents had not voluntarily adopted smoke-free home rules by 2010-2011. Children living in these homes are likely to be exposed to involuntary smoke. The existence of a voluntary smoke-free home rule was associated with parental education, parental race/ethnicity, parental age, household income, age of the youngest child, and household structure. The results are in agreement with those of other studies that found persistent disparities by these parental and household factors , and indicate that disparities in home rules have not improved by 2010-2011. Healthy People 2020 calls for increasing the overall proportion of voluntary smoke-free homes by approximately 10% and eliminating health disparities . Therefore, continued work is warranted to encourage the adoption of voluntary smoke-free home rules and reduce disparities in involuntary smoke exposure and tobacco-use-related diseases among smoker households with children. Most parents agreed that smoking should not be allowed in cars when children are present. Strong support for voluntary smokefree rules in cars may reflect the perception that the space in a car is more confined than a home and suggests that parents are more concerned about involuntary smoke exposure for children in a car. Parents may perceive that it is safe to smoke in certain areas of the home or at certain times, as long as children are not present. In comparison, a car offers less ambiguity about likely exposure of children riding with smoking adults. In addition, time spent in a car is usually short and requires less effort from parents to refrain from smoking. More than two-thirds of parents thought smoking should be disallowed in outdoor play areas, which was not as strong a preference as that for car rules. However, a preference for voluntary smoke-free rules in outdoor areas could indicate more antitobacco attitudes, considering that the concentration of involuntary smoke in an outdoor environment may be lower than that of indoor settings . These results suggest widespread support among parents for comprehensive legislation banning smoking in areas where children are likely to be present. The support in states with strong antitobacco programs, such as California, Florida, and Massachusetts, is generally higher than the national benchmark . Attitudes are likely to overestimate actual parental behaviors regarding rules restricting smoking voluntarily in cars and outdoor children's play areas. For example, Cheng et al found that at the national level the rates of voluntary smoke-free rules in home and cars among US households with smokers were similar . However, in our study, the rate of support for voluntary car rules was substantially higher than the rate of existence of voluntary home rules. According to the theory of planned behavior -which posits that personal attitudes toward behavior, subjective norms , and perceived behavioral control together shape the behavioral intentions and behaviors of an individual the gap between attitudes and behaviors may be explained by lack of perceived control over the behavior and subjective norms. For example, compared with parents in 2-parent households, single parents may be limited in their ability to step away from children to smoke outside the home. Likewise, smoking parents may be more likely to be surrounded by others who smoke and this, in turn, may dissuade them from implementing voluntary smoke-free rules in their homes and cars. The high rates of positive attitudes, however, suggest that smoking parents are aware of the harm of smoking around children and possess high level of readiness to implement voluntary smoke-free rules in cars and outdoor play areas. Our findings have implications for public health efforts, including policies aiming to help parents overcome barriers and translate their antitobacco beliefs and attitudes into actions. Health promotion programs should emphasize the threat of involuntary smoke exposure to children and promote the voluntary adoption of smoke-free rules in homes and private cars. Efforts to encourage or even mandate smoking bans in outdoor areas in housing complexes should be implemented. A "complete" voluntary smoke-free car rule when children are present should also be recommended or even required by law, as some states and municipalities have already done. Because residual tobacco smoke contamination that remains after the cigarette is extinguished can continue to harm the health of children , public education about thirdhand smoke and its detrimental health effects should be increased to promote adoption of voluntary smoking bans, even when children are not present. The widespread endorsement of negative attitudes about smoking in cars and outdoor areas in smoker households is notable, considering smokers traditionally hold less negative views about tobacco and are less likely to have a home nonsmoking rule . Our additional analyses showed that support for nonsmoking rules is even greater among nonsmoker households with children. Combined, these results suggest that most parents in households with underage children, regardless of smoking status, would be supportive of legislation to ban smoking in these areas. Most provinces of Canada and all states of Australia have now prohibited smoking in motor vehicles with a minor present . In the United States, 7 states implemented statewide policies to ban smoking in cars when children are present, although the cutoff age for child passengers varies by state . In addition, certain cities or counties in Hawaii, Indiana, New Jersey, and New York also have similar bans. We recommend that all states pass laws to ban smoking in vehicles when there are underage children present. Likewise, policy makers should also consider extending smokefree policies to outdoor children's play areas. Across the country, California has led the way and mandated smoke-free playground spaces designated for children . Its experience should inform similar efforts in other states. Attitudes toward voluntary smoke-free rules in car and outdoor play areas were predicted by parental and household factors similar to voluntary home rules. We found that, compared with non-Hispanic white parents, African American parents were less likely to voluntarily adopt a smoke-free home rule, but they were more likely to report a positive attitude toward voluntary smoke-free rules in cars and outdoor play areas. There is no reason to believe that potential information bias is systematically different between the report of attitudes and behaviors. Therefore, such findings are suggestive that African American smoking parents may encounter special difficulties in setting up voluntary smoke-free rules in the home. For example, they may have less access to other adults who can watch their children when they step out to smoke. The findings may also reflect that African American families are more likely to live in unsafe neighborhoods that make it challenging to smoke outside the home . These results underscore the need to conduct additional research to identify barriers to voluntary adoption of smoking bans among these populations and interventions to support their implementing smoke-free rules in homes, cars, and outdoor play areas. This study has limitations. First, parental practice and attitudes related to voluntary smoke-free rules are based on self-reported data. As smoking continues to decline in perceived normalcy, these findings may be biased due to social desirability. Second, positive attitudes toward voluntary smoke-free rules in cars and outdoor children's play areas do not necessarily equal support for legislation in corresponding places. For example, some individuals may think that it is their decision to implement a smoke-free rule in their cars and the government should not interfere and dictate this behavior. Therefore, future surveys should ask respondents about their own behaviors in these places and about their support for legislation applying to these 2 settings. Third, the response rate for self-reported data was moderate and may have caused selection bias. However, the rate is within range recommended for surveys of high importance about decisions on key policies or resources allocation . We found that children living in more than one-third of households with 1 or more smoking parents were not protected by voluntary smoke-free home rules by 2010-2011. The findings call for tobacco control and prevention efforts to continue to promote voluntary smoke-free home rules among households with both smoking parents and underage children, especially among households with single parents, parents of lower socioeconomic status, and parents without infants. We also found that parents from smoker households are generally supportive of voluntary smokefree rules in cars when children are present and in outdoor children's play areas, providing evidence and encouragement to policy makers to take action to ban smoking in these locations. --- Tables

A smoke-free environment protects children from exposure to involuntary smoke and also can reduce or prevent future smoking behavior. The purpose of this study was to examine levels and correlates of parental behavior and attitudes related to voluntary smoke-free rules in homes, cars, and outdoor children's play areas among US households with underage children and 1 or more smoking parents.We used data from the 2010-2011 Tobacco Use Supplement to the Current Population Survey and logistic regressions to model behavior and attitudes related to voluntary smoke-free rules in 3 settings.Overall, 60.1% of households with children and at least 1 smoking parent had voluntary smoke-free home rules. Approximately 84.6% and 71.5% of parents thought that smoking should not be allowed inside cars with children present and in outdoor play areas, respectively. Positive parental behavior and attitudes related to voluntary smoke-free rules were more likely among households with 2 parents, parents of higher education and household income, Hispanic parents, and parents of infants (P < .05).Tobacco control and prevention efforts are needed to promote the voluntary adoption of smoke-free rules in homes, private cars, and outdoor children's play areas. Most parents from smoker households with underage children were supportive of smoke-free laws for cars and outdoor children's play areas, providing evidence and encouragement to policy makers to take action to restrict smoking in these locations.

INTRODUCTION Since the housing reform in 1998, China's housing sales market has been flourishing, but the rental market has been lagging behind . According to relevant statistics, the proportion of rental housing in developed countries such as the United States, Japan, Britain, and other countries is about 30-60% . However, China's rental population is only 11.6%, much lower than that of developed countries , which is in stark contrast to the recent decline in home ownership in the West . As a result, there has been a serious imbalance between the housing sales market and housing rental market, highlighted by the fact that "the rental market has one short leg and the sales market has one long leg. " For a long time, China's housing rental market and sales market is not balanced, which has led to persistently high housing prices in China's large and medium-sized cities , the normal consumption of young people is overcrowded by housing , and the high investment in fixed assets of the real estate industry makes the development of other industries of the national economy insufficient . Therefore, cultivating and developing the housing rental market is an inevitable choice to revitalize consumption and promote internal circulation under the new development pattern. Most of the existing studies on housing tenure choice are policy-oriented and focus on macro-level studies, such as the impact of economics, institutions, and socio-demographic characteristics . However, a large portion of the change in homeownership remains unexplained. For example, Chinese households continue to tighten their belts to buy homes despite rising house prices , while the United States housing market saw homeownership rise more in 1995 and 2005 but fall more in 2015, which cannot be explained by demographics . The above studies are still lacking in explaining the underlying mechanisms that explain why Chinese people buy houses despite high housing prices. Neoclassical economics lacks explanatory power for the underlying behavioral logic. Therefore, it seems critical to consider other factors that influence homeownership choices. Some studies have shown that culture is an important factor influencing consumers' housing consumption decisions . Particularly in the Chinese cultural context, housing consumption behavior is not only a commodity-rich behavior, but it is also a socioculturally distinct behavior. For more than 2,000 years, Confucian culture is the core of Chinese traditional culture and plays a dominant role, and its customary beliefs and values passed down from generation to generation have a subtle and profound impact on the housing consumption decisions of the Chinese public . However, the existing studies have the following shortcomings. On the one hand, most of the existing studies only consider the influence of a single dimension of Confucian culture, and lack an in-depth exploration of the subdivision dimensions of Confucian housing culture. In addition, most of the existing studies have used subjective reporting methods to obtain data and have conducted empirical studies using economics and statistics to confirm the influence of Confucian culture on housing consumption behavior. However, there are some practical limitations in the existing studies, such as individuals may not be able to express their true opinions and feelings due to face, catering to policies, and other factors, such as the implicit attitude of "applause but not the audience" in their responses. In recent years, the development of cognitive neuroscience has provided objective, visual, and scientific explanations for individual consciousness and behavior . Based on this, this study examines the public stereotypes of housing consumption behavior from the perspective of Confucian culture, and further analyzes the influence of the degree of Confucian cultural influence on the public's cognitive attitude toward housing consumption, revealing the inhibiting effect of Confucian culture on housing rental demand. Qi and Ploeger studied the influence of cultural concepts on Chinese people and found that Chinese people are very face consciousness and face consciousness. Fong found that Chinese people emphasize more on family responsibilities and obligations when comparing the differences between Chinese and Western cultures. Hofstede found that Chinese people are relatively more risk-averse and pursue long-term orientation and collectivism. Based on the above studies, this paper first summarizes the connotation of Confucian culture in Chinese society into four dimensions, including face consciousness, extended family concept, group conformity, and crisis consciousness. Then, by constructing a questionnaire using the dichotomous method, the subjects were divided into high and low impact groups from the overall and segmentation dimensions. At the end of the experiment, the participants' explicit attitudes were measured by a self-reported questionnaire. Finally, we measured the stereotype of housing consumption by Confucian culture using the electroencephalograph method and explored the implicit attitudes of individuals using ERPs techniques. Our study contributes to the existing literature in the following ways. First, this study extends Confucian culture-related research by considering the influence of the Confucian culture segmentation dimension. Second, this study applies physiological measures to the field of urban housing research, and uses research tools from cognitive neuroscience to validate and refine the stereotypes of rental housing among youth group, revealing the variability of cognitive perceptions and decision-making behaviors among individuals at a deeper level. It is the first time to explore the activation model of housing consumption stereotypes, which expands the related research in the field of urban housing. The rest of this paper is organized as follows. Section "Literature Review and Research Hypotheses" reviews relevant literature and raises research hypothesis. Section "Materials and Methods" introduces experimental methods, including experimental procedures and data collection. Section "Results" presents the results of data analysis. Section "Discussion" analyzes the conclusions obtained directly from the results and gives the corresponding solutions. Lastly, Section "Conclusion and Limitations" summarizes the findings of this study and points out the limitations of existing studies. --- LITERATURE REVIEW AND RESEARCH HYPOTHESES Influence Factors of Housing Tenure Choice Housing purchase choice is not a simple investment or consumption decision, but a complex system event . Since the beginning of the research on housing purchase choice in the 1980s, the factors influencing housing tenure choice can be basically summarized as microeconomic factors, macroeconomic environment, socio-demographic characteristics, psychological factors, cultural factors, etc. From an economics perspective, microeconomic factors can be subdivided into uncertainty in residents' income, household savings, credit status, and net household liquidable wealth , and changes in the macroeconomic environment including rising consumer prices, rising mortgage rates, subsidy regimes, and rent controls . Enstrom-Ost et al. uses a two-stage instrument variable logit and probit model to study tenure choices in the Swedish housing market, which shows that financial constraints are inversely related to homeownership. Based on an online questionnaire, Liu and Li researched home purchase decisions in urban China and found that fixed income, for children's education, solid credit, impending childbirth, and marriage are the five most important factors. From a sociological perspective, demographic characteristics can be subdivided into age, gender, marital status, education of the household head, and number of working population . Chen based on household survey data in Guangzhou, found that housing tenure choice was positively correlated with marital status, age, and education, revealing the heterogeneity of socioeconomic status on housing tenure choice. Anderson et al. Using 2004 and 2008 panel data from the Survey of Income and Program Participation , household secured debt, household wealth, and household income play an important role in changes in household tenure choices. Scholars such as Hochstenbach and Boterman argue that the impact of parental housing tenure on children's intergenerational support is also not negligible. From a psychological perspective, personal subjective factors, such as property ownership preferences, can have a significant impact on housing tenure choices . Rubaszek and Rubaszek showed the impact of psychological factors on college students' preferences for home ownership through questionnaires at both economic and psychological levels. Foye et al. argued that homeownership provides intangible psychological benefits and is seen as a symbol of one's social status and home property is seen as a status commodity. Rohe et al. argued that from a psychological perspective, homeownership provides homeowners with a sense of belonging and self-esteem compared to renters. --- The Influence of Culture on Housing Tenure Choice Culture contains a set of enduring beliefs or values that permanently and profoundly influence and solidify people's cognitive preferences and decision-making behaviors . Under different cultural backgrounds, residents' family concepts and intergenerational relationships vary widely, and their housing consumption preferences will also change accordingly, which will lead to differences in their housing tenure choice. In terms of research on culture on homeownership choices, Marcén and Morales analyzed the role of culture in determining whether an individual is a homeowner using data from first-generation immigrants arriving in the United States under the age of 6, suggesting that culture is a homeownership decision important factors. Cultural differences and policy biases affect housing rental and purchase choices . In Germany, rental housing is of good quality and can be used as an alternative to home ownership, usually with a transition to first-time homeownership at a later time after the first birth . In Lagos State, socio-cultural experience of low cost residential areas in Ikorodu influenced resident satisfaction, and cultural background was found to be an important factor influencing consumer housing consumption choices . In Asia, Wei et al. argue that marriage to buy a house is a culture, and marriage triggers demand for home buying as young men use home buying as a strategy to improve their relative position in the marriage market. In the context of Chinese culture, Mu et al. believe that with the deepening of marketization in China, the economic and cultural importance of housing ownership has increased, and the desire of young people to own housing has also increased. Huang et al. explores the causes of high homeownership in China from a cultural perspective and constructs an empirical model to show that cultural traditions stemming from patrilineal marriage and the cultural perception of homeownership as a status symbol drive the high homeownership rate of Chinese residents. Based on the 2011 China Household Finance Survey, Wu examined the impact of Confucian culture on home ownership, and the study showed that household heads with more siblings were more likely to own a house, confirming the interdependence among family members. Currently, housing is the embodiment of culture, representing an important status and bringing various benefits . Thus, rooted in traditional Chinese culture, family structure has become a trigger for entry into homeownership and has gradually evolved to make homeownership a necessary condition for marriage . --- ERP Components Relate to Stereotypes Stereotype refers to the relatively fixed views on specific types of people or things. As a social cognitive schema, it can simplify the process of social cognitive processing and enable people to quickly form the cognition of the corresponding objects . In terms of their valence, stereotypes have both negative and positive aspects. Facts have proved that brand stereotype influences favorable consumer behaviors toward a brand . Country of origin stereotypes affect consumers' product evaluations . In short, stereotype affects all aspects of people's lives. Interactive Memory Systems Account distinguishes explicit and implicit attitudes, revealing the processing system of different components, including cognition, emotion, and behavior . Explicit stereotypes are based on conscious cognitive processing, and fail to account for underlying biases . Different from explicit stereotypes, implicit stereotypes operate without conscious and controlled intentions . Implicit stereotypes are based on fast systems and emotions, making individual instinctive decisions. Thus, in implicit stereotype measures, people are required to respond spontaneously to relevant stimuli in order to examine the individual's underlying associations with these stimuli. Event-related potentials are evoked potentials elicited by stimuli, which reflect EEG signals synchronized with the presentation time of stimuli and show people's objective cognitive and psychological behavior. With millisecond temporal resolution, ERPs technology is able to track the extremely fast processes involved in language processing, allowing real-time assessment of the neural activity of cognitive processes . As a result, ERPs techniques are widely used to explore the neural mechanisms behind human decision-making . In addition, studies have been conducted to explore the cognitive neural mechanisms of implicit stereotypes through ERPs technology. Therefore, ERPs is well-suited for exploring the existence of implicit stereotypes and their causes for excavation. The N400 component of ERPs, first identified as a result of semantic mismatch, is an ERPs signal for semantic processing that peaks around 400 ms after stimulus onset. There is a more-prominent N400 response when the end of the sentence is semantically incongruent with the rest of the sentence . In numerous studies, the N400 component has been considered as an observable indicator of stereotypes, and experimental stimuli with stronger conflict significance can induce larger N400 ERPs wave amplitudes. For example, Hou et al. used the N400 to explore public stereotypes of the recycled water end uses. Grant et al. used the N400 to explore how gender stereotypes affect the semantic processing of male and female speaker statements. In addition, all of the above studies suggest that stereotype incongruity conditions induce greater N400 wave amplitude. Therefore, the ERPs component N400 was selected as an observational indicator in this study to investigate whether the public has implicit stereotypes about rental and home purchase consumption. The above studies show that housing tenure choice is not just a rational behavioral decision, especially since Eastern cultures have more cultural and emotional connotations than Western cultures, which tend to be rational and materialistic. However, the existing research has the following shortcomings. First, most existing studies take homeownership preferences for granted, and the causes and processes of their formation need to be explored in depth. In addition, existing research has shown the influence of culture on housing consumption preferences, ignoring the extent of cultural influence on perceived housing consumption preferences and decision-making behavior. --- Research Hypotheses Under the influence of Confucian culture, what are the public's stereotypes of renting and home ownership? For a long time, Chinese people have a "rent-averse" culture and a tradition of home ownership, and a cultural perception and preference for owning property, and owning their own home is ingrained in Chinese families . People have always believed that owning a home is the only way to have a family, and that owning a home is the only way to feel safe and happy. At the same time, since in most cases, renting is seen by the general public as a transitional housing, it also does not allow them to reflect their status, let alone to have a good living environment and supporting living facilities and rights . Therefore, people have an obsession with the idea that only when they own their own home in a city can they be considered to be rooted in that city. Based on the above analysis, the following hypotheses are proposed in this paper. --- H1a: The public has a negative stereotype about rental housing. H1b: The public has a positive stereotype of home ownership. Face consciousness has been one of the most important ideas for Chinese people since ancient times, and it is also the core of Confucian cultural values . The love of face is a major national character of the Chinese people , which profoundly influences people's value orientation and behavioral intentions . In addition, Chinese people have a strong desire to compare with each other. To some extent, what Chinese people buy is not a house, but a face. As a result, homeowners have higher self-confidence or enjoy higher social status . Therefore, individuals with a high face consciousness will make every possible effort to gain or maintain face. For example, in terms of consumption behavior, consumers with high face consciousness tend to buy luxury goods with scarcity characteristics, so as to reflect their social status and achieve the purpose of self-realization. In terms of face consciousness and housing consumption behavior, since renting does not reflect one's status in the eyes of the public, groups with high face consciousness are more reluctant to rent housing. Based on the above analysis, this paper proposes the following hypotheses. H2a: Face consciousness had a positive effect on negative stereotypes of renting, with greater N400 peak amplitude in the high face consciousness group compared to the low face consciousness group . H2b: Face consciousness has a positive effect on positive stereotypes of home purchase, with greater N400 peak wave in the HFC Group compared to the LFC Group. Group conformity is an important feature of Confucian culture. Group conformity perception refers to subjective perceptions that are influenced by one's own group or reference group and want to be consistent with that group. It is widely believed that Chinese collectivist values originate from Confucian culture, and that collectivism under Confucianism is based on interdependence, group subordination, and group behavioral norms . Therefore, the collectivism-oriented culture in China encourages people to follow group conformity, and individual choice intentions are easily influenced by others . Chinese people are very concerned about whether their behavior is consistent with the group, and there is a strong psychology that is consistent with the group behaviors . In the housing consumption decision, the group unanimously prompts consumers to make the decision to buy a house. Therefore, the stronger the individual's subjective perception of group conformity, the more he or she prefers to purchase a house. Based on the above analysis, this paper proposes the following hypothesis. H3a: Group conformity has a positive effect on negative stereotypes of renting, with greater N400 peak amplitude in the high group conformity group compared to the low group conformity group . H3b: Group conformity has a positive effect on positive stereotypes of home purchase, with greater N400 peak amplitude in the HGC Group compared to the LGC Group. Chinese urban families are predominantly large, and the concept of extended family in Confucian culture is rooted in the sense of responsibility of family members who are closest to each other by blood . In China, housing is not only a carrier of the family, but also provides emotional security for blood relatives under the same roof . In combination with the deep-rooted cultural desire in China, homeownership is also a necessary step in fulfilling the obligation and responsibility to provide security for the family . In addition, driven by China's mother-in-law economy, home ownership is seen as a basic requirement and a major signal of marriage , and as a bond for the continuation of family inheritance. Therefore, the more individuals are influenced by family perceptions, the more they prefer to purchase a home. Based on the above analysis, the following hypotheses are proposed. H4a: Extended family concept has a positive effect on negative stereotypes of renting, with greater N400 peak amplitude in the high extended family concept group compared to the low extended family concept group . H4b: Extended family concept has a positive effect on positive stereotypes of home ownership, with greater N400 peak amplitude in the HEFC Group compared to the LEFC Group. Chinese have a strong sense of risk aversion and tend to make long-term plans. Homeownership offers greater financial liquidity due to secured borrowing. Meanwhile, homeowners can hedge their risk and avoid uncertainty about rent levels with nominally fixed mortgage payments . Homeownership not only provides a place for individuals to settle down, but also brings spiritual sense of belonging and security . In the long run, owning a house is a way for the Chinese to hedge against risk. Therefore, home ownership is a guarantee of financial security, and preference for buying a house is even more profound in China . Based on the above analysis, the following hypotheses are proposed. H5a: Crisis consciousness has a positive effect on negative stereotypes of renting, with greater N400 peak amplitude in the high crisis consciousness group compared to the low crisis consciousness group . H5b: Crisis consciousness has a positive effect on positive stereotypes of home ownership, with greater N400 peak amplitude in the HCA Group compared to the LCA Group. Policy perception refers to individuals' knowledge of policies and their evaluation of the effects of policy formulation, implementation, and enforcement; perceptions determine preferences, which further guide their behavior and decisions. Policy perception plays an important role in farmers' decision making process as meaningful moderators reflecting individual attitudes, environmental influences, and social norms . In addition, legal policy perception positively moderates the relationship between environmental intentions and behavior . The higher the youth group's knowledge and recognition of housing rental policies, the stronger the role of relevant rules in guiding their behavior, and the more likely they are to live in rental housing. Based on this, this paper proposes the following hypothesis. H6: Policy perception negatively moderates the relationship between Confucian culture and housing consumption stereotype. Perceived usefulness influences individuals' behavioral attitudes and has an impact on behavioral intentions. Mwesiumo et al. investigated the impact of public purchasers' attitudes toward public procurement innovation through a selfadministered questionnaire and showed that perceived usefulness is the most important factor influencing attitudes toward PPI in public procurement innovation. In addition, Chi explored external market-oriented consumer technology acceptance from the consumer's perspective, and the study showed that perceived usefulness can influence behavioral intentions by enhancing consumers' attitudes toward use. In the assessment of housing consumption intentions, perceived usefulness reflects the extent to which consumers perceive that the act of purchasing a home enhances various aspects of their life process, such as the ability to buy a home to settle in a household and solve their children's education problems. Based on this, this paper proposes the following hypothesis. H7: Perceived usefulness negatively moderates the relationship between Confucian culture and housing consumption stereotype. The imbalance of housing services in China's housing sales market and housing rental market, the fact that owned housing has more autonomy than rented housing, the lack of rental market policies and the lack of regulation. Coupled with the fact that residential income is still dominated by the growth of house prices, the increase in rent is small compared to the increase in house prices, while the different rights of renting and purchasing make the property rights have a certain impact on the equality of rights, so residents' trust in the rental market will affect the preference for home purchase. This paper assumes that the higher the trust in the rental market, the more inclined to rent a residence. Based on this, this paper proposes the following hypothesis. H8: Psychological factors negatively moderates the relationship between Confucian culture and housing consumption stereotype. Income is an important influential factor in determining housing tenure choice . Most scholars believe that households with higher lasting income as well as better stability are more likely to purchase a home and have higher requirements for housing comfort. Jorgen et al. analyze the consumer housing consumption market through an objective function of consumer utility maximization and show that consumers' preference for ownership leads them to choose homeownership, while those with no preference for ownership are more. The results of this study show that consumers' preference for ownership leads them to choose homeownership, while those who have no preference for ownership prefer to rent housing. Based on the above analysis, this paper proposes the following hypotheses. H9: Income positively moderates the relationship between Confucian culture and housing consumption stereotype. In summary, the theoretical analysis framework of this paper is shown in Figure 1. --- MATERIALS AND METHODS --- Participants and Data Collection As the main force of housing demand, understanding why the youth group prefers home ownership among various forms of housing is conducive to alleviating the strong demand in the home purchase market in the process of implementing "houses are for living in, not for speculation. " It is also conducive to better respond to the needs of the youth group in the process of fostering rental housing and developing subsidized housing. However, there is a lack of research on the attitudes and perceptions of youth group, and the importance of homeownership to young people has been overlooked to some extent . Therefore, in this study, we selected the youth group as the study population and recruited school students as participants to explore the stereotypical housing consumption of the youth group under different degrees of Confucian cultural influence for the following reasons. First, the participants we selected are all senior graduate students or doctoral students who are about to go out of campus to look for jobs, and their decision to buy a house is also one of their most important considerations . In fact, according to the latest survey data from the National Bureau of Statistics of China, the average age of marriage for men and women in China is about 25 1 years old, which is similar to the average age of the participants in this experiment. To a certain extent, there is an understanding of this research question, which can be approximated to represent the youth group. Second, this study adopts the paradigm of psychological experiments. In psychological research, considering the cheapness and the need for a high degree of cooperation in EEG experiments, many studies recruit university students as participants. In addition, in the cognitive neural experiments in the social sciences, students are also mostly selected as participants . We posted an advertisement for participants on the social media WeChat public website and randomly selected participants for the experiment. Then, a total of 36 students from Xi'an University of Architecture and Technology participated in the 1 http://www.stats.gov.cn/tjsj/pcsj/rkpc/6rp/indexch.htm experiment. The experimental data of six participants were excluded due to the number of trial superimposition requirements that needed to be met for each stimulus type . Therefore, valid experimental data from the remaining 30 participants were recorded for analysis. All participants in the experiment were right-handed, in good health, and had normal or corrected normal vision. Indicators of the demographic characteristics of the participants are shown in Table 1. At the end of the experiment, participants were given RMB 66 as a reward. In addition, this experiment strictly adhered to the Declaration of Helsinki and its later amendments . The Laboratory of Neuromanagement in Engineering of Xi'an University of Architecture and Technology approved this study. Each participant filled out a subjective questionnaire after the end of the experiment. The details of the questionnaire for this study are presented in the form of an Appendix. --- Stimulus Materials The experimental material consisted of 256 stimuli, which included 16 photographs × 16 words . The stimulus materials used in our study were determined by the following steps. First, adjectives were used to describe individuals' attitudes toward rental and home purchase consumption. A vocabulary of 32 words was first constructed from the Chinese Affective Words System , and a five-point Likert scale was established to set the questions and distribute the questionnaire. After that, 30 respondents were recruited, who were not involved in the EEG experiment. Subsequently, they were asked to rate their individual attitudes toward rental and home purchase consumption. Then, the eight words with the highest mean scores from the positive and negative adjectives, respectively, based on the adjective scores, were selected as the stimulus words in this experiment. Finally, 16 Chinese words about perception of housing consumption attitudes were selected. The Chinese words used in this study are shown in Table 2. Thus, we have four conditions of stimuli, including owning-positive, owning-negative, rentingpositive, and renting-negative. Participants were provided with a keyboard and responded by pressing either the "1″ or "3″ key on the keyboard . Before the formal experiment started, all participants were given a brief introduction about the experimental process. After they fully understood the process, the experiment started. --- Experimental Procedure The stimulation process was presented using E-prime 3.0 programming, and the entire experiment was consisted of 256 trials. The initial eight trials were pre-experiment, and the stimulus materials in the pre-experiment were used only for practice, not in the formal experiment, to promote participants' understanding of the experimental procedures. In order to alleviate the fatigue of the participants during the experiment, the 256 trials of the formal experiment were randomly separated into two blocks, with each block consisting of 128 trials, and the participants rested for 2 min after each experimental step of one block was completed. The size of each picture was 1,920 × 1,080 pixels. Prior to the experiment, the participants were told that they would need to view a series of real-life pictures with text about renting and owning a house. Next, they would observe either positive or negative adjectives, at which point participants were asked to indicate whether they agreed with the adjectives provided through a keystroke task to describe their attitudes about renting and owning a house. In the experiment, the participants sat comfortably, approximately 100 cm from a computer monitor with the stimulus material presented in the center of the screen. The experimental procedure is as follows. At the beginning of each trial, a 600 ms fixation cross appeared on the screen to indicate the start of the trial. Next, a picture of a rented or purchased house was displayed for 1,600 ms. Then, a blank screen appeared for 600 ms. Then, the phrase "make me feel" appeared for 1,000 ms. After that, a blank screen appeared for 600 ms. Subsequently, a positive or negative Chinese adjective was displayed for 1,000 ms. After the lexical stimulus, a blank screen was presented for 600 ms. Then, an interface asked participants to press a key .. The notification disappeared immediately after the participant pressed the button. If the participant did not press the button within 3,000 ms, the notification page also disappeared. Each experiment takes about 1.5 h from pre-preparation to the end of the experiment. During the experiment, participants were asked to minimize eye and muscle movements. The experimental procedure and details of each trial are shown in Figure 2. --- Electroencephalography Data Recording The electroencephalograph signals were recorded using a cap consisting of 64 Ag/AgCl electrodes and a NeuroScan Synamp2 Amplifier . Bilateral mastoid electrodes were selected as the reference point. The electrodes were placed 1 cm above and below the left eye to record the Vertical electrooculogram , while the horizontal EOGs was recorded from the right and left eye . Subsequently, statistical analysis of the EOG data was performed. After a baseline correction process, artifact activity above ±100 μV per trial was excluded from the analysis. EEG recordings were then digitally filtered with a 30 Hz low-pass filter . Each EEG period was corrected for baseline from 200 ms before stimulation to 800 ms after stimulation onset, using the 200 ms before stimulation onset as the baseline. Finally, the mean value of event-related potentials was calculated separately for each stimulus type. --- RESULTS --- Public Stereotypes of Housing Consumption Decision --- Behavioral Results Two-way repeated measures ANOVA on subjective reported data showed no significant main effect [F = 1.941, p > 0.05, η 2 p = 0.063] of housing tenure choice and a significant main effect of adjectives [F = 44.256, p < 0.001, η 2 p = 0.604]. In addition, there was an interaction between housing tenure choice and adjectives [F = 10.125, p < 0.01, η 2 p = 0.259]. For housing purchase, there was a statistically significant relationship between positive and negative adjectives [F = 49.986, p < 0.001, η 2 p = 0.633]. It was found that, after processing information in the heuristic system, the public tends to think that owning a house is self-contained, restful, warm, and comfortable. --- Event-Related Potential Results Semantic anomalous sentences induce N400 components at the frontal electrode sites . Therefore, four central electrode locations, FZ, FCZ, CZ, and CPZ, were selected to observe EEG waveforms to reflect the public's cognitive attitudes toward renting and owning . Afterward, a distinct negative wave was evoked around 400 ms of stimulus onset. Since N400 waveforms always show a peak amplitude 300-500 ms after stimulus onset, these evoked negative EEG waveforms can be considered as N400 waves . In this study, a two-way repeated measures ANOVA with peak amplitude of N400 was used to explore the public stereotypes of renting and owning a house and whether there was an interaction between the housing tenure choice and adjectives . The observation of the four central electrode locations revealed that the target stimulus in the incongruent condition evoked a greater N400 peak amplitude in the frontal-central region compared to the target stimulus in the same condition. The ANOVA results showed no significant main effect of housing tenure choice [F = 0.0395, p > 0.05, η 2 p = 0.013] and adjectives [F = 1.156, p > 0.05, η 2 p = 0.038]. However, there was a significant interaction between housing tenure choice and adjectives [F = 6.323, p = 0.018, η 2 p = 0.179]. In terms of renting a house, descriptions using positive adjectives evoked greater N400 wave amplitudes in the frontal-central region than those using negative adjectives, and the peak amplitude of positive and negative target adjectives between rentals was statistically significant . This shows that the public has significant negative stereotypes about rental housing consumption, Thus H1a was supported. In other words, compared to homeownership, under the cognitive processing of the fast system, the public perceives renting as constraining, stressful, unhappy, and crisis, while adjectives such as ease, comfort, warmth, and peace of mind run counter to their perceptions of renting. As a result, a larger N400 wave was generated. --- The Influence of Confucian Culture on Housing Consumption Decision First, we divided the participants into a high Confucian culture influence group and a low Confucian culture influence group through a questionnaire. There was statistical significance between HIC Group and LIC Group . The score of HIC Group was higher than that of LIC Group [ vs. ]. Then, a three-way repeated measures ANOVA was used to further explore the HIC Group and LIC Group stereotypes of housing consumption decisions. EEG waveforms of the four electrodes under the grouping are shown in Figure 5. The ANOVA results showed that the main effects of housing tenure choice [F = 0.427, p > 0.05, η 2 p = 0.015] and adjectives [F = 1.138, p > 0.05, η 2 p = 0.039] were not significant, and the main effect of the degree of Confucian culture was significant [F = 4.447, p = 0.044, η 2 p = 0.137]. In terms of different Confucian cultural influence groups, a more negative stereotype of renting exists in the HIC Group compared to the LIC Group [F = 6.109, p = 0.02, η 2 p = 0.179], as shown in Figure 6. In other words, the more influenced the Chinese public is by Confucian culture, the more pronounced the negative stereotypes of renting consumption. Ultimately, the differences in decisionmaking about housing consumption choices among different Confucian culture-influenced groups were examined. The housing consumption decision data revealed that there are differences in the housing consumption decision-making behavior of the Chinese public due to different levels of Confucian cultural influence [Z = -2.408, p = 0.016]. This study shows that the housing consumption choices of the LIC group are more rational and balanced compared to the high Confucian culture influence group. --- The Influence of Confucian Culture Subdivision Dimension on Housing Consumption Stereotype To further investigate the influence of each of the four dimensions of Confucian housing culture on the public's housing consumption stereotypes, a three-factor repeated measures ANOVA was used. The four Confucian housing culture influence degree groupings were statistically significant , with all high culture influence groups scoring higher than the low culture influence groups. The ANOVA results showed that the main effect of face consciousness was not significant in the renting-positive condition, and the difference in N400 peak amplitude data between the high and low face consciousness groups was not statistically significant . In other words, the negative stereotypes of rental consumption among the youth group are not influenced by the face consciousness. The main effect of group conformity was significant , and there was a significant difference in N400 peak amplitude data between high and low groups of group conformity , Supporting H3a. The degree of influence of group conformity positively affects negative stereotypes of rental housing among youth group. The main effect of extended family concept was significant , with a significant difference in N400 peak amplitude data between groups with high and low extended family concept , Supporting H4a. This suggests that the degree of influence of extended family concept positively affects negative stereotypes of rental housing among youth group. The main effect of crisis consciousness was not significant, and the difference in N400 peak amplitude data between high and low crisis consciousness groups was not statistically significant . There was no statistically significant difference in N400 peak amplitude data between the high and low face conscious groups in the owningnegative condition . No significant difference in N400 peak amplitude data between groups with conformity high and low levels. There is a significant difference in the peak amplitude of N400 between the groups with high and low extended family concept , Supporting H4b. In addition, There is a significant difference in the N400 peak amplitude data between the groups with high and low crisis consciousness , Supporting H5b. The study found that group conformity, extended family concept, and crisis consciousness had significant effects on housing consumption stereotypes. Table 4 shows the average peak values and SE of N400 under the influence groups of face consciousness, group conformity, extended family concept, and crisis consciousness. Based on multiple regression for moderating effect analysis, the interaction effect between policy perception and crisis consciousness was significant . In other words, higher policy perception and agreement mitigated the effect of crisis consciousness on negative attitudes toward rental housing. The interaction effect between perceived usefulness and extended family perception was significant . That is, low levels of perceived usefulness alleviated the negative impact of extended family concept on rental housing stereotype. The interaction effect between perceived usefulness and crisis consciousness was significant . In other words, low levels of perceived usefulness mitigated the negative impact of crisis consciousness on rental housing stereotype. The interaction effect between psychological factors and group conformity was significant . In other words, higher trust in the rental market mitigated the effect of group conformity awareness on negative attitudes toward rental housing. The interaction effect between psychological factors and group conformity was significant . That is, higher rental market trust has weakened the group conformity awareness of over-preference for homeownership. The interaction effect between psychological factors and extended family concept was significant . That is, higher trust in the rental market alleviated the effect of extended family concept on negative attitudes toward rental housing. The moderating effects between the remaining two variables were not significant. --- DISCUSSION In this study, the ERPs method was used to carry out an EEG experiment, and the N400 ERPs components representing the implicit attitude of the housing consumption stereotype were collected and counted. Specifically, as follows, the public's cognitive attitudes and decision-making behaviors toward renting and owning a house are first explored at both implicit and explicit levels, and the stereotypes of housing consumption among the youth group are verified at the general level. After that, the differences in housing consumption stereotypes between HIC Group and LIC Group were compared. Then, the differences in the influence of housing stereotypes were explored from the high and low groups of the four dimensions of Confucian housing culture in comparison. Overall, the degree of influence of Confucian culture has a positive impact on the public's housing consumption stereotypes. In terms of the sub-dimensions of Confucian culture, group conformity, extended family concept, and crisis consciousness are important influences on housing consumption stereotypes. Finally, significant factors affecting housing consumption were selected for moderating effect analysis, and the study showed that policy perception, psychological factor, and perceived usefulness can significantly moderate housing consumption stereotypes. In terms of renting, ERPs data show that description with positive adjective can induce more obvious N400 amplitude. In other words, the public believes that renting is restrictive, stressful, unhappy, and crisis. Currently, we can start by breaking through these ideas to weaken the negative public attitudes toward renting. First of all, the state advocates that the social status of renting and owning a house is equivalent, owning a house is not a symbol of social status and financial ability, there is no need to purchase a house for face, renting is also a way to enjoy life. In addition, the government strengthens the supply of smallscale, low-rent rental housing suitable for young people and new citizens, so as to achieve good and stable rent, and effectively enhance residents' sense of well-being and security. Compared with LIC group, HIC group exists a more negative stereotype of renting. In addition, as previous studies have shown, there are differences in the decision-making behavior of the public due to their different levels of influence by Confucian culture . The housing consumption decision-making data show that compared with HIC Group, LIC Group is more balanced and rational in renting and purchasing housing consumption decisions. This may have the following reasons. On the one hand, under the influence of the individual trend, some urban youth values and behavior gradually began to change, rent more easily accepted. On the other hand, the LIC group will choose housing consumption based on its own specific circumstances . Guiding the HIC group to gradually shift to the LIC group and weakening the irrational demand for home purchase will help the development of China's housing rental market and build a housing gradient consumption model. There was no significant difference in the N400 peak amplitude data between the high and low impact groups of face consciousness in both the owning-negative and rentingpositive conditions. This may be due to the fact that the participants in this experiment represent the contemporary Chinese youth group, with generally higher education levels, which improve their ability to accept new ideas , and their living concepts are more independent and cutting-edge . Conversely, extended family concept played a significant role in both owning-negative and renting-positive conditions. As recognized in earlier literature, our housing tenure preferences are culturally mediated and socially constructed , and identified that within-family socialization played a crucial role in this . Interestingly, this study found that crisis consciousness only had a significant effect on the positive owning stereotype, whereas group conformity only had a significant effect on the negative renting stereotype. This may be because Chinese people generally have a sense of urgency. Housing ownership provides housing stability and reduces housing risks by protecting families from future rent increases, which is also a way for Chinese people to avoid risks . In the cognitive concept of Chinese public, group conformity drives the strong psychology of individual existence and group behavior, strengthens the individual's obsession with housing purchase, and then produces a more negative cognitive attitude toward rental housing. The results of the moderating effect analysis indicate that policy perception, perceived usefulness, and psychological factors play a significant moderating role on housing consumption stereotypes. By vigorously publicizing the policy subsidies for the rental market to increase public awareness of macro policies, enhancing life confidence to reduce the impact of crisis consciousness, and promoting a new living culture to improve residents' attitudes toward rental housing, the government can achieve the goal of restraining excessive expansion of housing demand. When we realize that public rejection of rental housing can be transformed by undermining the public's perceived usefulness of the housing services that come with homeownership. In this regard, the state should gradually eliminate the differences between renters and owners in public service rights , and enhance the substitutability between the housing rental market and sales market . By enhancing public trust in the rental housing market, the public's perceived rejection of rental housing is gradually weakened. Therefore, the government should accelerate the formulation and implementation of regulations related to the rental and sale policies, increase financial subsidies and tax relief for rental housing, and strengthen the supervision and management of the rental market . However, there is no significant moderating effect between income and the dimensions of Confucian culture, probably because traditional cultural perceptions have made the Chinese public's perceptions of housing consumption deep-rooted and its influence more far-reaching. Therefore, when the government vigorously develops the housing rental market, it also needs to transform the public's perceptions of housing consumption by means of other policies that govern such cultural perceptions; no amount of policy can achieve the desired effect if it stays in the consciousness and philosophy of the past. --- CONCLUSION AND LIMITATIONS Conclusion Since ancient times, under the influence of Confucian culture, the proportion of urban households in China who rent their homes is generally low, and the housing consumption model of "preferring to own a house rather than rent one" is still common, and the demand for housing rental has not yet been effectively released. The public's attitude toward renting and purchasing directly affects their housing consumption decision-making behavior. Therefore, it is important to study public attitudes toward renting and owning, and to guide and change them in order to finally realize the Rent and Purchase system. This study explores public stereotypes of housing consumption decisions by designing a human-factor experiment on public stereotypes of renting and homeownership consumption, and draws the following conclusions. First, in general, describing rental housing with positive adjectives elicited significant negative brain waves between 400 and 500 ms compared to negative adjectives. That is, the public holds negative implicit stereotypes of rental consumption. However, the subjective report data suggest that the public holds positive outward stereotypes about home purchase consumption. The data on housing consumption decisions reveal that there are differences in the housing consumption decisionmaking behaviors of the Chinese public, influenced to varying degrees by Confucian culture. Second, for groups with different degrees of Confucian cultural influence, the HIC Group exhibits a greater amplitude of N400 for rental housing consumption, and the negative stereotype of rental housing consumption deepens as the public's degree of Confucian cultural influence increases. Third, by subdividing the four dimensions of Confucian culture, the study showed that as the degree of cultural influence increased, group conformity, the extended family concept, and crisis consciousness significantly influenced housing consumption stereotypes, while there was no significant difference in the influence of face consciousness on housing consumption stereotypes. Fourth, this study also considers other factors that may influence Confucian culture on housing consumption stereotypes. The moderating effect analysis indicates that perceived usefulness, psychological factors, and policy perception can be important factors in guiding the public's stereotype of housing consumption, while the moderating effect of income on the stereotype of housing consumption is not significant. --- Limitation and Future Research Although this study provides valuable and detailed insights into public stereotypes about housing consumption decisions, there are limitations to the results. In particular, it must be noted that this paper explores public stereotypes of housing consumption from a Confucian cultural perspective, and both the aggregate and subgroup levels suggest that Confucian culture influences the public's housing consumption decision-making behavior. However, there are deeper reasons that need to be explored. For example, to what extent does Confucian culture contribute to the Chinese public's "preferring to own a house rather than rent one" housing consumption model? In future studies, we will explore the extent to which multiple factors influence the stereotype of "preferring to own a house rather than rent one" from multiple perspectives. In addition, on the one hand, the Confucian housing culture scale defined and developed in this study is not fully representative of the Confucian housing culture as a whole. On the other hand, due to the limitations of the research equipment and conditions, the participants in this study were all students of Xi'an University of Architecture and Technology, and the research sample was somewhat limited. --- APPENDIX The specific contents of the questionnaire. Perceived usefulness Owning a home and being able to choose the style of decoration I like drives my preference for buying a home Owning a home avoids having to move a lot and makes me prefer to buy a home I think rental housing will limit the future compulsory education of children Psychological factor I think it is difficult to protect the rights of renters and landlords in the current rental market I have little trust in the Xi'an rental market I think the rental market in Xi'an is not well-served Policy perception The understanding of "rent and purchase"housing system The attitude of "rent and purchase" housing system which guide housing gradient consumption The understanding of Xi'an City Talent Settlement Rental Subsidy Policy Subjective attitudes toward housing consumption I have a positive attitude toward renting a house I have a negative attitude toward renting a house I have a positive attitude toward owning a house I have a negative attitude toward owning a house --- Housing consumption decision Owning a house Renting a house --- DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. --- ETHICS STATEMENT The studies involving human participants were reviewed and approved by Laboratory of Neuromanagement in Engineering of Xi'an University of Architecture and Technology. The patients/ participants provided their written informed consent to participate in this study. --- --- Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Publisher's Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The ideas of face consciousness, group conformity, extended family concept, and crisis consciousness in Confucian culture have a subtle and far-reaching impact on housing consumption decision among the Chinese public, forming a housing consumption model of "preferring to own a house rather than rent one." The poor interaction between the housing rental market and the sales market caused by the shortage of rental demand and irrational purchasing behaviors has led to soaring house prices and imbalance between supply and demand that prevail in major cities in China. To gain a deeper understanding of public cognitive attitude toward decisions on owning and renting a house, this study divided the subjects into high and low impact groups based on the overall Confucian culture and four subdimensions. It attempts to take a cognitive neuroscience approach for assessing public stereotypes of housing consumption decision with different types based on the analysis of event-related potentials (ERPs). The results are as follows. First, overall, there is an obvious implicit stereotype of renting a house and explicit stereotype of owning a house among the public. Second, ERPs data show that descriptions of renting a house with positive adjectives could evoke more significant N400 responses. In other words, in the heuristic system, the public perceive that renting a house is restrictive, stressful, unhappy, and crisis. Data from subjective reports show that, after processing information in the analytic system, the public tend to think that owning a house is self-contained, restful, warm, and comfortable. Third, a more negative stereotype of renting a house exists in the high Confucian culture influence group (HIC) Group than in the low Confucian culture influence group (LIC) Group, and is more inclined to own a home. Fourth, under the Confucian culture sub-dimension, there are differences in housing consumption stereotypes between high and low groups in terms of extended family concept, group conformity, and crisis consciousness. Fifth, the moderating effect analysis found that perceived usefulness, trust in the rental market, and policy perception can be important factors in guiding public housing consumption stereotypes.

Introduction Ageing has been recognized as one of the most critically important health-care issues worldwide. It is relevant to Asia, where the increasing number of older populations has drawn attention to the paramount need for health-care investment, particularly in end-oflife care. Despite the differences in the context of a 'good death,' patient autonomy has been regarded as one of the top priorities across cultures and religions. This is indicated as essential for quality of dying. From a policy perspective, the advocacy of advance care planning is a mean to honor patient autonomy . Recently, international Delphi consensus studies were conducted in the US and Europe to develop a definition of ACP and recommendations for its application . ACP was defined as the ability to enable individuals to define goals and preferences for future medical treatment and care, to discuss these goals and preferences with family and health-care providers, and to record and review these preferences if appropriate. Sudore et al. in the US has conducted a series of extensive studies on ACP. She has defined and measured the outcome of ACP . She has defined and measured the outcome of ACP, and studied strategies for engaging different populations such as the older adults, people with chronic illnesses, the homeless and prisoners . She has also investigated the obstacles of promoting ACP , designed a web-based tool-'PREPARE' to facilitate the implementation of ACP, conducted randomized controlled trial to demonstrate the efficacy of ACP for general public and cancer patients and investigated the utilization of ACP on electronic record for emergency physicians and safety for patients . However, the aforementioned studies were mainly conducted in the Western context. Since the majorities of Asians are Buddhists/Taoists, Muslims and Hindus rather than Christians/Catholics, the Asian view on life and death is quite different from that of the West especially under the influence of religions; for example, the Buddhism/Taoism are taking a rather naturalistic view toward death and family and physicians are the key person in EOL decision-making under Confucianism and Hinduism doctrines. Specifically, there are disparity in attitude toward advance directive by religions . Roman Catholics were more likely to object to withdrawal of life support, compared with Protestants, Jews and those with no religious affiliations. For Muslims, the responsibility of passing a final verdict regarding life-prolonging therapy rests with the physician; human life is dignified and should be sustained as long as it is sustainable, which should be decided based on the patient's situation. Asian Indian Hindus were more likely to refuse life-sustaining interventions and more likely to engage in autonomous decision-making. For Buddhists, 75% did not want life-prolonging therapy, when considered futile. In summary, except Roman Catholics, Christians prefer a more aggressive way facing AD and EOL issues while the other religions taking a more naturalistic view. In addition, it is generally agreed that EOL communication in Asian countries such as Japan, Korea and Taiwan are usually rare and only a few patients' EOL care preferences can be honored . Since most East Asian countries are influenced by the Confucianism and the concept of 'filial piety ,' patient autonomy is consequently subordinate to family values and physician authority . The dominance from family members and physicians during a patient's EOL decisionmaking is recognized as a cultural feature in Asia. Physicians often disclose the patient's poor prognosis and corresponding treatment options to the male, family member, rather to the patient him/herself. It is assumed that the patient might be unable to accept the truth of poor disease prognosis psychologically. In order to address this ethical and practical dilemma, the concept of 'relational autonomy' and the collectivism paradigm might be ideally used to assist Asian people, especially older adults, to share their preferences on future care and decision-making on certain clinical situations with their families and important others . The definition and recommendations of ACP need to be re-evaluated and redefined to be culturally specific for the Asian societies. In this review article, we invited experts in EOL care from Hong Kong, Indonesia, Japan, South Korea, Singapore and Taiwan to briefly report the current status of ACP in each country from policy, legal and clinical perspectives. --- ACP in Hong Kong The concept of planning ahead for EOL care has drawn increasing attention over the past two decades in Hong Kong due to an epidemiological transition of morbidity and mortality from acute health problems to chronic debilitating diseases. The average life expectancy at birth of people in Hong Kong was among the highest in the world, but its quality of death index lagged behind that of several other Asia-Pacific economies, including Taiwan, Singapore, Japan and South Korea . This raises the concern over how to improve the quality of EOL care rather than merely focusing on life preservation. Hong Kong, which has been a British colony for over a century but with the majority of the population being Chinese, is a place mixed with the cultures of East and West. While the western belief of self-determination and person-centered care is being highly upheld, the traditional Chinese belief of collectivism in major decision-making and medical paternalism is still preserved . These conflicting beliefs add challenges to the implementation of ACP. Following two rounds of public consultation on legislation on ADs launched by the local government in 2004 and 2009, there has been considerable awareness that planning for EOL care involved many aspects beyond the legal concerns and legislation alone and is unlikely to prevent tensions and conflicts between families and health-care teams over patient care . The hospital authority that governs the local public hospitals has recently issued a set of ACP guidelines to provide practical guidance for clinicians on conducting ACP discussions . In this document, ACP is defined as a proactive communication process regarding EOL care. While the primary target group is patients with advanced progressive disease, the guidelines extend the scope to include family members of the mentally incompetent and minor patients. However, before this, a number of initiatives had been developed to promote ACP across care settings to empower patients to share their EOL care preferences and to engage family members in the ACP process. One of the first published ACP studies in Hong Kong found that frail older adults living in residential care homes were open to discussing their EOL care and the ACP discussion enabled them to reflect upon their underpinning values, increased their understanding about EOL care and thereby the stability over EOL care preferences, as well as alleviated their existential distress . With increased confidence on its public acceptance, attempts were made to conduct ACP through home visits so as to reach the home-living patients and engage their family members in this important dialog . These studies provided further evidence on the importance of ACP in increasing family members' understanding about patients' care wishes and increasing preparedness of health-care team for the difficult conversation. In addition, there have been growing interest in developing strategies and tools for promoting the concept in a more systematic manner and building capacity to enhance the sustainability of the practice in various care settings . Hence, the Jockey Club Charities Trust provided a huge amount of funding support for a community-wide campaign on EOL care. Along with this movement, the Food and Health Bureau of Hong Kong also launched a public consultation again in 2019 Downloaded from https://academic.oup.com/jjco/article/50/9/976/5881673 by guest on 07 May 2024 on formulating a statutory framework for ADs and removing legal barriers to dying in place, one of the aspects covered by ACP . All these recent changes revealed that the implementation of ACP involved a wide range of issues at individual, familial, organizational and socio-cultural levels. --- ACP in Indonesia Indonesia is currently facing an epidemiological transition towards an increasingly aging society and advanced cancer population. As the fourth most populous country in the world, Indonesia is a home for over 265 million inhabitants with the average life expectancy at birth of 71.7 years, dramatically rising from 45 years in 1970 and 63.6 years in 1990 . This condition had led to the shift towards an aging population, with the share of the elderly population of 22 million in 2017 to an estimated 61 million in 2050, attributing to 19% of the total population . More than one-fourth of this age group population had reported at least one form of disability . At the same time, cancer prevalence in this country is rising to 29% over the last 5 years, with more than 70% of cases being admitted in the advanced stage . These transitions underlined the urgent needs of quality palliative and EOL care. To ensure that palliative and EOL patients receive care consistent with their values, ACP is required to help health-care professionals understand the patient's wishes and preferences for future care and treatment. Advance care planning in Indonesia remains in its infancy. In terms of regulation, the do-not-resuscitate form is the only recognized and widely used format of written documentation of ACP in Indonesia. It has taken 15 years since palliative care initiation in Indonesia to have palliative care regulation integrated into its national policy in 2007 . In one of its decree, an AD is recognized as a written form documenting patient's surrogate decision maker and wishes and preferences on future medical care and treatment, particularly in relation to life-sustaining treatment withdrawal when it is considered not beneficial. It is recommended that a competent patient is provided sufficient information prior to complete and sign an AD. A palliative care team is suggested to play the role of initiating the discussion with the patient. In a case an incompetent patient has no AD, the family may provide a consensus based on patient's best interest as guidance for healthcare professionals. Health-care professionals are allowed to make decisions of DNR if patient considered terminal and resuscitation could not cure the patient nor improve his or her quality of life based on the most recent scientific evidence . In terms of practice, various conditions may serve as potential challenges for ACP in Indonesia. The lack of integration of palliative care service in most part of the country, coupled with the sociocultural influence on health-care decision making had rendered care in the EOL very complex in Indonesia . Illness is a shared family affair, where communicating EOL to the patient is considered as a taboo . Family plays a central role in health-care decision making in Indonesia, particularly for the elderly population . Although mutual understanding had become a general preference of both patients and physicians, a one-way communication style is still a common doctor-patient type of communication. Healthcare professionals' time constraint, lack of knowledge on ACP, lack of communication skills, discomfort in breaching death-and-dyingrelated communication, together with lack of guidelines, legal and institutional support were among the possible barriers . Meanwhile, a lack of awareness of the concept of ACP may contribute to one of the major obstacles for ACP implementation in Indonesia. To provide basic palliative care knowledge, the government, through the National Cancer Control Committee, is conducting training of trainers to internists and pediatricians of provincial hospitals and general practitioners and nurses in primary care clinics in 34 provinces of Indonesia . Since 2015 an education session of ACP has been integrated into palliative care training programs to several health-care professionals in referral and regional hospitals in Jakarta and West Java, as a collaboration of local non-governmental organizations with Singapore International Foundation. However, currently, no formal training or education available for ACP, nor public education to promote ACP, either on a smaller or on a larger national scale. Not surprisingly, the complex concept of ACP as an ongoing process of value identification is mostly foreign for Indonesian health-care professionals, even though it might have been practiced at a certain level without formal recognition as ACP. Unfortunately, no studies are available to provide information on the perspective and practices of ACP in Indonesia. Currently, there is no consensus definition of ACP available in Indonesia, yet it is translated into living will or last will in the national guidelines for cancer palliative program. In the new draft of the Ministry of Health Regulation on Palliative Care, the term ACP will be introduced as one of the palliative care services implementations, together with other services such as symptoms management, psychosocio-spiritual support, bereavement support and final care of life. However, more detailed guidelines or recommendations on how ACP can be implemented in Indonesia have not existed. It is evident that much remains to be done in Indonesia in relation to ACP. More research is urgently required to provide evidence on influencing factors for ACP practice in Indonesia, whether and how ACP can be helpful in the management of palliative cases in Indonesia, and how ACP can be best integrated into medical care delivery in Indonesia. The future directions for the implementation of ACP in this country include the integration of ACP in the formal education of medical doctors and nurses, a national public campaign to promote ACP, institutional support for ACP and the development of a guideline for ACP implementation in the local context. --- ACP in Japan Japan has experienced a super-aging society since 2007. The number of patients who died of cancer has continued to increase over the past decades . More than 995 000 patients were diagnosed with cancer in 2016, and more than 373 000 patients died of cancer in 2017. To help cancer patients receive quality EOL care consistent with their goals and preferences, it is essential to promote ACP through timely and compassionate EOL discussions with patients and their families. Traditionally, nondisclosure of EOL issues to cancer patients has been a typical practice in Japan. Patients often value 'heart-to-heart communication' , which involves understanding others without words, entrust EOL medical decisionmaking to others such as families and physicians, and expect family-centered decision making . Families used to be the ones who consented to DNR orders in almost all cases . However, findings of recent domestic surveys suggest positive changes in cancer patients' and physicians' attitudes toward ACP. For example, one survey involving Japanese patients who had been informed about the cessation of anticancer treatment revealed that the majority preferred their physicians to respect their sense of values and explain the status of their illness as well as how their life expectancy would influence their future plans . Most Japanese oncologists reported that they would discuss EOL issues such as the option of forgoing anticancer treatment, prognosis, DNR status, hospice referral and place of death when the need arise, although the timing of initiating such discussions might vary widely . Another nationwide survey demonstrated that nearly a half of terminally ill cancer patients in palliative care units had ADs for various LST, and over two-thirds of patients with decisional capacity were involved in the DNR decision . In 2017, the Ministry of Health, Labour and Welfare conducted a nationwide survey to clarify Japanese people's perceptions regarding EOL care . Among the general public , care staff , nurses and doctors who responded, only 3.3%, 7.6%, 20% and 22%, respectively, reported that they knew ACP well; however, up to 65-80% of them agreed with its concept when its meaning was explained. It also showed that 66-78% of the respondents agreed with the concept of AD, although only less than 10% reported that they had actually completed ADs, and less than a fourth agreed that ADs should be enacted. In 2018, the MHLW revised the guidelines on decision-making process on EOL care . The concept of ACP was then officially introduced in the guidelines as a process in which the individual repeatedly discusses EOL care with his or her families and/or other closely related people and health-care professional team in advance. To promote ACP based on the guidelines, a governmentcommissioned educational project called 'Education For Implementing End-of-Life Discussion ' was launched. The E-FIELD provides interdisciplinary HCPs with one-day workshops on ACP that includes lectures, role-plays and group works throughout the country. Furthermore, the Ministry publicized a nickname of ACP and logo mark to raise public awareness. After the introduction of ACP by the government, several professional societies such as Japan Medical Association, Japan Geriatric Society and Japan Pancreas Society published position statements, clinical guidelines and/or brochures for the public regarding ACP. Various communication skills training programs are also available. In addition to the 'SHARE' program that was established and implemented in Japan , those developed in the US have recently been introduced to Japanese health-care professionals . Yet, a standardized model of ACP for cancer patients and their families has yet to be established, and there is no reimbursement system solely for ACP. A number of recent research investigations have explored cancer patients' perceptions of and preferences for ACP discussions . Nationwide surveys among bereaved families of cancer patients also provided useful findings about the optimal timing of EOL discussions, and recommended care strategy to help families maintain hope and prepare for death . A retrospective study indicated that the provision of prognosis information to advanced cancer patients was significantly associated with longer duration between the last day of chemotherapy or the day of confirmed DNR order and death . Currently, several prospective studies are under way to examine the effects of various ACP interventions for patients with serious illnesses including cancer. It is evident that the development of ACP in Japan is still in its infancy. Future policies and education programs should continue to address how to respect patient preferences through shared decisionmaking while ensuring that family harmony is maintained. More clinical experiences, discussions and empirical research regarding ACP at all levels are clearly needed to develop a culturally appropriate ACP in Japan. --- ACP in South Korea South Korea is the country with the population of 50 million and is ranked the 27th most populous country in the world. The life expectancy of Korean is 83-years old and South Korea is a rapidly aging society with 14.3% of the population who are 65 years of age or older . The number of deaths in the year of 2018 was near 300 000, and more than 45% of them were over 80-years old. The most common cause of death was cancer. Among cancer deaths, the rate of hospice service usage has increased from 7% in 2008 to 22% in 2017 since the insurance reimbursement for hospice was launched in 2015 . The change in the place of death has also changed a lot recently. In the past, even in the early 1990s in South Korea, most people hoped to die at their homes , because dying away from home was considered as a tragedy and bad luck, especially in rural areas, but according to a survey, many people in Korea still wish to pass away at their homes, surrounded by their loved ones, but the rate of dying at home was 14.4% and especially in cancer patients, it was 6.2% in 2017 . Also, traditionally, children of elderly patients with terminal illnesses made EOL decisions for their parents, keeping the diagnosis of a life-limiting-illness secret from the patients . In addition, in many cases, the resuscitation order was created almost immediately before the death . Back then, the concepts of ACP, good death and patient autonomy were not widespread. However, there were two significant events in our society that resulted in changes. One is the Boramae hospital case, which a drunken man fell and had an emergency operation for subdural hematoma. The patient's wife strongly asked her husband to be discharged from the hospital against the doctor's advice. Eventually the patient died. After his death, the patient's brother filed a lawsuit against the doctor for being an accessory to murder. After this case, doctors were repulsed against hopeless discharge or discharge against medical advice. The second case is that of Grandma Kim, she had lung cancer and became vegetative state after bleeding excessively from a lung biopsy in 2008. Her children asked to stop LST. Her family sued the hospital for refusing the request and won the trial. She demised in 2010. After these series of events, Koreans felt the need for legislation related to LST and have discussed this issue since 2010. Finally, 'the Act on decisions on Life-Sustaining Treatment for patients in hospice and palliative care or at the end of life' was announced in March of 2016 and this Act was enforced in February in 2018 . The purpose of this new Act is to prescribe matters necessary for hospice and palliative care and LST for patients at the EOL, determination to terminate such LST, to protect the dignity and value of human begins by assuring the best interest of the patients and by respecting their self-determination. Later, the revision was made in March 2018, expanding the types of life-prolonging interventions, the definition of life-limiting illness, and reducing the influence of families that must be agreed when patients are unconscious. The AD and physician orders for LST are formulated as legal forms related to this act, people can write fill up these documents in registration offices and hospitals that also have ethics committees. Each document needs to be uploaded in the National Agency for Management of LST registry. After registration, the institution can get reimbursement for the documents or performance related LST from the government. The legal Act was enforced only for a year, but it brought positive influence toward ACP, protecting patients' dignity and respecting self-determination. However, enhancements to the Act and its promulgation are needed. Many people are still unfamiliar with terms like AD, POLST or ACP . Some doctors may feel forced to register without having enough information or time to think through in advance. Also, it is not applied to all hospitals and people yet. Therefore, advertisement and education in detail should be tailored for patients, health-care providers and for the general public because their attitudes and level of awareness are quite different . There was a report that increased participation of patients in EOL discussions after conducting structural consultations related to AD even before the Act was implemented and the knowledge for advance care plan had a positive impact on quality of life . But, there are few studies that have been published after the Act enforcement. According to a recent university hospital press release, patients have made more decisions on their own than in the past . Although ACP legally has already begun in South Korea, there is still a great need for conceptual propagation of ACP with continual improvement in education and policy , and a culture that respects the best interests and preferences of patients through discussions with individuals, families and closely related people . Also, more research is needed on our clinical experience and the influence of culture. --- ACP in Singapore Advance care planning forms the foundation of precedent autonomy and is a key tenet of good EOL care. As a complex intervention, it has been shown to promote person-centered care through respecting EOL care preferences, decreasing the use of aggressive and burdensome interventions and reducing post bereavement distress amongst family members . The earliest statute relevant to ACP in Singapore would be the Advance Medical Directive or AMD. The AMD was enacted in 1996 with the intention of allowing an individual to sign in advance a legal document to inform one's doctor that one does not want the use of any extraordinary LST to prolong life, in the event that one is terminally ill and mentally incapacitated from making decisions . However, the AMD applies only to a very narrow scenario and as a document, does not adequately address the broad spectrum of advance decisions that can be explored as well as the values and personhood underlying such decisions for any individual. There has also been poor uptake of the AMD by the general public and up till 2015, less than 25 000 people had signed an AMD . Like many Asian cultures, death remains a taboo subject and non-disclosure of diagnosis and prognosis of a life-limiting illness is a common practice in Singapore, particularly towards the elderly. There are likely to be inter-generational differences in the perceptions of receptivity towards talking about death and dying. In a local study done in year 2000 on the preferences and choices of a group of elderly Chinese subjects attending a day care center in Singapore, 43 subjects median age 71-years old were interviewed. Only 16.3% have heard of the AMD. Contrary to the popular approach of collusion, 83.7% and 76.7% of the subjects preferred to be told of the diagnosis and prognosis of a terminal illness, respectively . A decade later, in a study on the knowledge, attitudes and practices of the advance medical directive in a residential estate in Singapore in 2010, 414 residents were surveyed and only 37.9% of the participants knew about the AMD prior to this study . Respondents who did not wish 'to be kept alive indefinitely on a lifesupport machine' and accepted the 'imminence of death' were more willing to sign an AMD. The article concluded that there is a need to increase awareness of the AMD. A retrospective medical chart review of all 683 adult patients who died without intensive care unit admission in a Singaporean hospital in 2007 showed that there were DNR orders for 66.2% of patients. Orders to limit life-sustaining therapies, including ICU admission, intubation and vasopressors were infrequent. Discussions with alert and conversant patients on DNR orders occurred ∼5% of the time, while discussions with family members occurred 80% of the time. The article concluded that there was a lack of commitment by doctors on orders for DNR/CPR and to limit life-sustaining therapies, infrequent discussions with patients on EOL decisions, and excessive burdensome interventions for the dying . A survey on death attitudes of doctors and nurses carried out by Lien Foundation in 2014 revealed that 53% of all doctors are comfortable discussing death and dying with their patients but only 31% of nurses are comfortable discussing death . The Mental Capacity Act was enacted in 2008 allowing an individual to nominate another person or donee to act on his or her behalf in event he or she loses mental capacity, for financial and health-care matters. However, this precludes the donee from making decisions regarding potentially LST on behalf of the individual . Whilst the limited research showed gaps and barriers, clearly there is a need for a systematic approach to training health-care professionals in carrying out this difficult conversation as well as in implementing a system to capture and honor patients' preferences. In 2009, the ACP faculty Respecting Choices of Wisconsin, US was invited by the National Healthcare Group, Singapore to train health-care professionals from various health-care systems in their paradigm of ACP. RC espoused a staged approach to ACP, the constructs of the conversation being dependent on the health status of the individual. Though suitable for healthy individuals as well, RC recommended that Singapore start ACP conversations with people with advanced illness, when ACP is most relevant and more likely to be carried out by health-care professionals. Small pilot projects were started in various groups such as people with advanced heart failure and respiratory disease but these early efforts were stumbled by multiple barriers such as lack of clinician time as well as the perspective that ACP conversations are secondary to clinical treatment. In 2010, the Ministry of Health of Singapore commissioned the Lien Centre for Palliative Care to formulate a National Strategy for Palliative Care that delineated 10 key goals for palliative care in Singapore, one of which is that all patients at the EOL have access to ACP . In 2011, the Ministry of Health injected seed funding and appointed the Agency for Integrated Care as the programme manager for the national ACP programme, which subsequently was aptly renamed as Living Matters . A national ACP Steering Committee was formed to oversee ACP efforts, aided by sub committees, e.g. training and curriculum, forms and documentation and information technology sub committees. Respecting Choices taught that the elements of successful ACP implementation were 4-fold, namely engagement of the community, training in ACP facilitation, building a system to capture and honor preferences and implementing quality improvement and audit checks. At the beginning of the programme, it was realized that the lack of buy in from clinicians is a major impediment and grassroots efforts were implemented at engaging senior hospital leaders as well as the broad swathe of health-care professionals in major hospitals. Outreach efforts took the form of hospital conferences, talks at multidisciplinary forums and patient support groups. Collaborations with media, arts-related groups and national bodies promoted community engagement. In 2018, the Singapore Hospice Council launched the 'Live Well. Leave Well.' campaign to encourage Singaporeans to start early EOL conversations with their loved ones . Systematic training in ACP facilitation took place in various restructured hospitals to build capabilities in ACP facilitation and advocacy. The health-care professionals who attend training come from diverse backgrounds and vocations including staff from nursing homes and community home care services as well as doctors, nurses and allied health providers. Different health providers had different systems of capturing the ACP forms, but this was aligned in year 2017 with the advent of the national ACP system, implemented by AIC. Over the last decade, there was a steady rise in number of ACP documents completed and tracked on a national level, from 3 in year 2011 to a cumulative number of ∼14 000 in 2018. This does not include the conversations that take place between health-care professionals, patients and their loved ones and documented in the electronic medical records apart from completed ACP forms. ACP conversations probably have been taking place in clinical practice though the Living Matters framework was an added impetus to the ACP movement to match the sophistication of medical advancement with conversations about what people really want and value when seriously ill. Though similar in some shared practices, different hospitals drove ACP according to different models of implementation in a real-life experiment. In some hospitals, the programme was clinician led while in others it was driven by social workers. At the beginning of the Living Matters programme, various re-structured hospitals were loosely guided by key performance indicators. Quality improvement and audit checks were not well aligned across institutions. In the international literature up till 2018, outcomes of successful ACP were not clearly defined . Advance care planning in Singapore started out as a hospital centric practice and amongst patients with life-limiting illness . Whilst this has the advantage of diffusing the practice of ACP within well-resourced stakeholders in the health-care continuum, it has reinforced the misconception that ACP is only for the very sick within large hospital settings. Anecdotally, within a hospital system in Singapore, only ∼40% of patients were able to participate in the conversation and the large majority of the conversations were carried out with the loved ones of the patients. In the absence of decisionmaking capacity of the patient at the point of ACP facilitation, the question posed to the patient's loved ones is what they think the patient would have wanted. It has been shown in international and local studies that surrogates' understanding of patient's wishes, that is substituted judgment, is often inaccurate. This also goes against the grain of the true tenet of ACP as an iterative conversation that should start when people are relatively well and evolves over time as the individual's health status and values change. This is especially so as anyone can suddenly become seriously ill and mentally incapacitated. A survey amongst 406 community-dwelling residents living in Housing and Development Board apartments across Singapore was carried out in 2017, this time evaluating awareness of and attitudes towards ACP . 14.4% of participants had heard of ACP, mainly through the media , from family and friends and health-care providers . 26.8% of those who had heard of ACP knew how to begin an ACP discussion and 12.5% of them had a prior ACP discussion. After explanation of what ACP entails, 60.1% of participants were willing to begin an ACP discussion. Being of an older age, having a life-threatening illness, and having more knowledge about ACP were significant factors associated with willingness to have an ACP discussion. Barriers included perceiving oneself as still healthy and preferring the family to make decisions instead. The study concluded that there is low awareness but high-expressed willingness to engage in an ACP discussion amongst those surveyed. More is needed in community engagement and raising awareness on ACP. In 2018, a national evaluation of the Living Matters framework commissioned by the Ministry of Health was carried out. Concordance with ACP preferences was studied in a sample of 1731 decedents aged 21 years and above who completed the ACP documentation . More than 90% opted for no cardiopulmonary resuscitation. Approximately 40% expressed a wish to pass away at home, and ∼30% expressed a preference for death in an institutional setting . Whilst the level of agreement in terms of preferences for CPR and comfort measures was very high among the deceased population who completed their ACP, ∼50% of the sample died at their preferred place of death. A local study of individual, clinical and system factors associated with the place of death concluded that besides cultural and clinical factors, system-based factors including access to home palliative care and ACP discussions were found to influence the likelihood of home deaths . Increasing the reach of home palliative care services and encouraging ACP may facilitate honoring patients' preferences for home deaths. A qualitative study of the implementation of ACP in re-structured hospitals in Singapore described the importance of a cultural and behavioral transformation driven by committed leadership and shared organizational purpose . Current efforts at promoting ACP are aimed at shifting the conversation upstream to the polyclinics and specialist outpatient clinics to patients with chronic disease and more well but middle aged to elderly people. Efforts have also been made to engage private family physicians in understanding their role in ACP. An ACP awareness week was organized across the heartlands of Singapore to reach out to the masses about having dinner table conversations and what 'Living Well, Leaving Well' meant. New task forces under the purview of the national ACP Steering Committee were set up, namely the outpatient ACP taskforce, the community outreach taskforce and the ACP research and development sub-committee. The Ministry of Health of Singapore has been supportive of these broad efforts at ramping up ACP across the health-care continuum as well as in the community. With the surge in use of information technology, there are concurrent efforts at looking at how to educate the public on various forward planning instruments including LPA and ACP and to unite these in a one stop portal as touchpoints of care in a person's health-care journey. Much remains to be done, including improving quality of ACP training, facilitation and documentation as well as aligning quality measures and audits of the process of ACP. There are research gaps in how ACP can be best implemented in Singapore's melting pot of cultures, how to improve shared decision making between patients, their loved ones and health-care providers as well as how to enhance tracking of quality of ACP process beyond tracking number of ACP documents and measuring concordance between clinical outcomes and preferences. With the passing of laws in neighboring countries such as the 'Patient Right to Autonomy Act' in Taiwan and the 'Medical Treatment Planning and Decisions Act' in the state of Victoria, Australia, it remains to be seen how the laws and policies implemented within Singapore will evolve to meet the ethicolegal challenges that inevitably arise from the practice of ACP. In April 2021, when Singapore will host the eighth ACP International conference, it is hoped that this sentinel event will further spur Downloaded from https://academic.oup.com/jjco/article/50/9/976/5881673 by guest on 07 May 2024 --- Definition of ACP Hong Kong A proactive communication process regarding end-of-life care. While the primary target group is patients with advanced progressive disease, the guidelines extend the scope to include family members of the mentally incompetent and minor patients . --- Indonesia No consensus definition on ACP yet. --- Japan A process in which the individual repeatedly discusses EOL care with his or her families and/or other closely related people and health-care professional team in advance . --- Korea The process by which patients with medical staff voluntarily discuss the goals and specific methods of care to be performed in the future so that the patient's autonomy and best interests can be realized. ACP can occur at any stage of the illness and is often initiated at the request of the patient or on the recommendation of the medical professionals in the condition of aggravating the disease. The attending physician will provide the patient with information about the pros and cons of the treatment options that can be selected according to the progress of the disease to assist in the judgment and to record the medical records. Records may also include discussion about physician orders for life-sustaining treatment or advance directive and hospices as required by law . --- Singapore The process of discussion with patients with regards to their values, preferences and wishes for care in event they become seriously ill and unable to make decisions for themselves . --- Taiwan The process of communication between the patient and medical service providers, relatives and other related parties regarding the proper care that shall be offered to the patient and the options he or she has to receive or refuse life-preserving treatments and artificial nutrition and hydration when the patient is in specific clinical conditions, unconscious or unable to clearly express his or her wishes . impetus in efforts at ACP in the country and the region and bring person-centered care to the heart of medicine. --- ACP in Taiwan Under the influence of Confucianism, traditional Taiwanese people take a conservative attitude toward the issue of dying and death. Society treats the topic of dying and death as a taboo; therefore, family rarely discuss it explicitly with their loved one even who is diagnosed with a terminal disease. Collusion is a longstanding tradition, and patients are denied the opportunity of participating in medical decision-making, including decisions with regards to LST. The family usually takes over the responsibility of medical decisionmaking for the patients. Unsurprisingly, the AD completion rate is low at ∼7-8% as a result of poor communication . Studies showed that age, cognitive state, previous discussions of DNR orders with family members or physicians and the nursing home policy were factors related with completion of AD . From the family caregivers' perspective, if the family caregiver is primary, willing to disclose illness with the patient, familiar with the Natural Death Act and takes a less negative perception toward the Natural Death Act, he or she is more willing to help the patient to complete the AD . From the physicians' perspective, prior experience of working in a hospice, attitudes towards the Natural Death Act and knowledge about the Natural Death Act were three important factors that positively influenced the participation in ACP . Taiwan is the first Asian country to legalize the Patient Right to Autonomy Act following the Natural Death Act in 2000. The Natural Death Act guarantees dying patients the right to withhold and withdraw LST; however, the family can decide for the patient and the context is limited to LST and the Act comes into play only when the patient is imminently dying. The Patient Right to Autonomy Act goes one step further compared with the Natural Death Act; it is applicable to healthy people, the content includes issues related to artificial hydration and nutrition as well as LSTs; and it could only be amended by the patient himself or herself. The Patient Right to Autonomy Act was passed by the Legislative Yuan in Taiwan in the end of 2016 and enacted on 6 January 2019. This Act is established to ensure respect for patient autonomy, protect the right of patients to a good and natural death and promote harmonious physician-patient relationship. It advocates AD through the process of ACP consultation. It is aimed for person under any of the five clinical conditions: person diagnosed with terminally-ill diseases; with irreversible coma; in vegetative state; with terminal dementia or incurable diseases. It defines ACP as The process of communication between the patient and medical service providers, relatives, and other related parties regarding the proper care that shall be offered to the patient and the options he or she has to receive or refuse life-preserving treatments and artificial nutrition and hydration when the patient is in specific clinical conditions, unconscious or unable to clearly express his or her wishes. Since ACP consultation is mandatory in order to sign AD, people eligible for the consultation must be 20-years old or above. During the ACP consultation clinics, patients meet the medical team in the company of at least one relative with or without the medical surrogate. The medical surrogate is defined as someone who has no conflict of interest with the patient such as an organ transplant donor except successor to the patient. The medical team will discuss in detail on the possible LSTs and artificial hydration and nutrition under any one of the aforementioned five clinical conditions to the patients for at least an hour. The medical team usually comprises of a medical doctor, a nurse and a social worker/psychologist. In Taiwan, although the majority of health-care services are covered by the National Healthcare Insurance , the ACP services are not supported by the NHI system yet. The ACP participants need to pay an out-ofpocket fee of for the person for the ACP consultation. If there is a second person who wishes to consult the medical team, such as the family member, there would a concession charge of 83 USD for the second person and so on. An unified AD form would be completed to record the patient's EOL care preferences if appropriate. After the patient signing AD with two witnesses, the AD form will be uploaded and stored in an online electronic platform and reported The medical surrogate is defined as someone who has no conflict of interest with the patient such as an organ transplant donor except successor to the patient. The ACP participants need to pay an out-of-pocket fee of for the person for the ACP consultation. If there is a second person who wishes to consult the medical team, such as the family member, there would a concession charge of 83 USD for the second person Downloaded from https://academic.oup.com/jjco/article/50/9/976/5881673 by guest on 07 May 2024 on the personal health-care intelligent card. Thereafter, all hospitals in Taiwan would acknowledge that the patient has completed the AD and will act accordingly. Today, nearly 100 hospitals all over Taiwan have provided ACP consultation clinics. Before the enactment of the Patient Right to Autonomy Act, several indigenous studies investigate the effect of ACP to terminally ill cancer patients. A randomized controlled study concluded that ACP improved psychological distress, but not quality of life or preferred EOL care of the patient . In addition, a largescale ACP communication program in Taipei showed there is an increased AD completion rate to the eligible population as a result of the program and utilization of LST decreased following AD completion . Since the concept of ACP is relatively new in Taiwanese society and the development of ACP is still at a preliminary stage, more evaluation is needed to audit the effectiveness and consolidation of ACP consultation clinic in the future. --- Conclusion As the concept of patient autonomy continues to evolve in Asia, more countries begin to reflect on the issues of AD through the process of ACP to attain a better quality of dying among patients . According to the Asian experiences, we have seen different models of ACP implementation . The Asian Delphi Taskforce for ACP, under the supervision of Asian Pacific Hospice Network, has been working vigorously on the definition, roles, tasks and recommendations for ACP tailored for the Asian culture . The Asian Delphi Taskforce for ACP is currently undertaken by six Asian countries and a more detailed, culturally sensitive whitepaper for the Asian population will be published in the near future. --- Conflict of interest statement None declared.

Ageing has been recognized as one of the most critically important health-care issues worldwide. It is relevant to Asia, where the increasing number of older populations has drawn attention to the paramount need for health-care investment, particularly in end-of-life care. The advocacy of advance care planning is a mean to honor patient autonomy. Since most East Asian countries are influenced by Confucianism and the concept of 'filial piety,' patient autonomy is consequently subordinate to family values and physician authority. The dominance from family members and physicians during a patient's end-of-life decision-making is recognized as a cultural feature in Asia. Physicians often disclose the patient's poor prognosis and corresponding treatment options to the male, family member rather to the patient him/herself. In order to address this ethical and practical dilemma, the concept of 'relational autonomy' and the collectivism paradigm might be ideally used to assist Asian people, especially older adults, to share their preferences on future care and decision-making on certain clinical situations with their families and important others. In this review article, we invited experts in end-of-life care from Hong Kong, Indonesia, Japan, South Korea, Singapore and Taiwan to briefly report the current status of advance care planning in each country from policy, legal and clinical perspectives. According to the Asian experiences, we have seen different models of advance care planning implementation. The Asian Delphi Taskforce for advance care planning is currently undertaken by six Asian countries and a more detailed, culturally sensitive whitepaper will be published in the near future.

Introduction School is one of the most important environments for children and young people. We also know that there is a connection between education and perceived health [1]. Most students report good health, but the number of students reporting mental illness is increasing, as is the number of students who cannot reach their school goals [2]. A systematic review of research on the relationship among health, learning, and mental illness by Gustafsson et al. [1] pointed out several factors related to school performance where school failure and lack of social support were important factors. However, there is less knowledge about the relationship between adolescent adjustments and the role of family involvement in students' school attendance. The present study focuses on students' tardiness within the social and cultural context at school and within families. That is, tardiness can be understood as maladjustment related to various negative effects on academic results and social relations [3,4]. Adolescent adjustment includes attitudes, behaviors, cognitive-and social aspects specifically when it comes to the students' abilities to adapt within the school environment [5,6]. The students' adjustment describes the extent to which the student is committed to the school and accepted in the social milieu [5]. Previous research has found positive relations between a high degree of school adjustment and the results in school and social relations within school [3,7]. Regarding adjustment, scholars have pinpointed the importance of the family [8][9][10]. Parenting characterized by the use of reasoning and warmth can contribute as a protective factor when it comes to adolescent adjustment [7,10]. Support from parents is associated with better results in school [8]. In previous research, tardiness has been understood as the expression of risk for future problematic behavior that could lead to absenteeism, school dropout, exclusion, and later health problems [11]. We will examine tardiness in relation to a broader framework based on a social-ecological perspective [12], meaning that the physical, social and cultural dimensions of students' health evolve in a complex interaction among the environment, organizations, and individuals. The social-ecological theory also appears in the action area of Ottawa's charter [13], where to build a supportive environment for health is described as actions that "offer people protection from threats to health and enable people to expand their capabilities and develop self-reliance in health. They encompass where people live, their local community, their homes, where they work and play, including peoples' access to resources for health and opportunities for empowerment". This action area has been a platform for the development of healthy cities, schools, neighborhoods and other settings for daily life and activities [14,15]. Scholars have found that to be successful as a health-promoting school, collaboration, e.g., with the local community and parents, is crucial [16][17][18]. Hence, we focus not on the specific individual student and his/her behavior but, rather, on the interaction among students, families and school staff from a social-ecological theory perspective related to the school and to the implications useful for developing a supportive environment for students' health and learning. --- Absenteeism Many factors influence student achievement, directly and indirectly. Students who are absent from school are a major problem in Sweden [19] and in many other countries [20,21]. Swedish headmasters reported that during 2015, 1.7 per mille of all students in grades 1 to 9 were reported absent for a month or more, but 18.5 per mille were reported as being absent now and then . PISA reported that nine percent of respondents, aged 15, said they had skipped school one day or more over the last two weeks [22]. The definitions of absenteeism or truancy are not consistent. Kearny [23] suggested the term "school refusal behavior" as an umbrella that covers constructs related to children not going to or coming late to school. He illustrated this as a continuum of truancy going from "school attendance under duress and pleas for non-attendance" to "complete absence from school for an extended period of time" [24], p. 60. The third step along this continuum is tardiness in the morning followed by attendance. However, Kearny meant that a student could move along the continuum and, for example, go from tardiness in the morning to "complete absence . . . " or to "school attendance under duress . . . ". Truancy includes all students that have "successfully" missed school and those who tried to miss school but had not yet reached their goal [23]. Truancy must be understood as the complex result of different structural systems around the child, the family, and the school [25]. It begins when the child is young, often in primary school, and it causes harm, especially to the truants themselves; moreover, it is costly Reid [25] also found that parents and students realized the problem caused by the school environment and that teachers believe that parents' attitudes and the home environment were more influential. In their study, Heyne et al. [26] differentiate between Truancy , School Refusal , School Withdrawal , and School Exclusion . TR captures the problem of a child who does not want to go to school, lacks motivation and has a negative attitude but does not include late arrival. SR differs from TR, whereby the child wants to go to school but finds it too difficult, and this category might include a late arrival, while SW is related to parents versus school-based reasons. However, Heyne and colleagues, differing from Reid's findings, found that late arrival or missing school is not concealed from the parents [26]. Based on Kearny's continuum, one should pay attention to tardiness as an imaginary part of a behavior that can develop into school absenteeism, which in turn increases the risk of dropping out [11], school failure, long-lasting associations with crime, and problematic health behavior [27]. --- Specific on Tardiness Tardiness is not a new phenomenon, nor is it a problem limited to students at school. Tardiness or lateness to work has been considered a facet of undisciplined behavior related to maturity or motivation, e.g., by Edralin [28]. On the other hand, the use of flexible work time has become more common, and this opportunity for employees appears to reduce lateness behavior [29]. Previous studies on tardiness at school are limited, although it has been a research topic since 1930, e.g., Lockwood [30]. More recent studies, carried out nearly 75 years later by Gottfried [31] in the US, show that tardiness is a risk for school achievement in math and reading not only for individual students but also for their classroom peers. In addition, students who arrive late to the classroom pose a risk for others being tardy . In a study of 250 junior high school students, Gottfried [31] found that on average, boys were tardy more frequently than girls, but that half the students in the school had no incidence of tardiness. Fifteen-year-old students from OECD member countries were asked about truancy and late arrival at school [22]. The average percentage of students in OECD countries being late once or twice during the last two weeks was 29 percent. The average percentage of students being late more than five times during the last two weeks was 7.3 percent. The average percentage of students who are never late was 55 percent. Never arriving late was at a lower level in Sweden, as reported by 45.5 percent of the students, and 32.6 percent reported late arrival once or twice during the last two weeks. Punctuality was best in Japan and Korea, where between 80 and 90 percent reported that they never arrive late [22], p. 304. Girls and boys are reported in the same group. In the mid-fifties, D'Amico [32] published an article about teachers' perspectives on and treatment of tardiness. He found that the way teachers handled tardiness varied: some administered punishments, others took no action. The strategies were sorted into two categories: laissez-faire or punitive. Contemporary studies on tardiness interventions at school are in line with D'Amico's two main perspectives: reduce negative behavior or strengthen positive behavior. To strengthen a positive behavior, in order to arrive on time at school, Freeman et al. [33] found evidence in their review of skills training and family support but also of incentive-training, where students earned points for both attending class and arriving to class on time. Earned points could be used for movie tickets, school supplies, and restaurants [34]. Other examples include an intervention through which a US high school has a delayed start time, according to Thacher and Onyper [35]. Ugwuegbulam and Naheed Ibrahim [36] relate that punishment is a common strategy to reduce lateness behavior among students in Nigeria. However, they decided to change this behavior by using a fun perspective by introducing a game they played with students who came late to school four times or more within an observation time of two weeks . The game was presented in a welcoming manner: "Good friends, this group is ours, it belongs to us. We are here to play a game in a purely friendly manner. As we play the game, we discuss. Our discussion, which will be on lateness to school issues should be within us, that is, what we discuss here should be confidential" [36], p. 143. The game method has not yet been evaluated, but this is an example of teachers trying to develop new strategies to motivate students to be on time. To reduce negative behavior, Tyre et al. [37] implemented the START on Time programme , including different kinds of punishments in four steps and levels. The first step is sending parents a postcard about a student being late; the second is the student having no lunch if he/she is late again; the third involves activities such as cleaning desks and vacuuming carpets; and the final step, reached if a student is tardy more than 12 times, involves a conference with a parent or guardian, and the student is required to meet with a panel of tribal elders regarding his or her attendance. The evaluation shows that, compared to the average per day and per month prior to implementation, there was a 67 percent decrease in average daily tardiness rates . --- The Swedish School System The Swedish school system involves one mandatory preschool year followed by nine years of compulsory school or until the child is 15 years old. The assignment to the schools from the government is to effect compensatory work [38]. This means that schools are responsible for providing equal access to education and for compensating for differences in the students' capacity. It also means that the entire school system shall minimize the negative effects ofstudents' backgrounds on achievement and give them an equal opportunity to learn [39,40]. The Swedish school organization is responsible for creating a good learning environment for students' personal development and the development of their knowledge [38]. This includes promoting efforts and projects to increase students' attendance and thereby minimize absence in education. It requires systematic work based on a regular review of their own organization and content based on knowledge of the factors that promote student learning and development. The efforts must include all students. How the school's working environment is designed can in itself promote presence or contribute to absence. A good learning environment, with teaching adapted to the individual student's needs, promotes motivation to participate in education. The participation of students in the work of promoting presence can be effected in many different ways. They must participate in the work of creating a common approach to late arrival and determine what will be considered an invalid absence [38,39]. In summary, we have discovered that there is a gap in research in regard to studies about tardiness in school, even though it is well known that early identification of student tardiness increases the opportunities of detecting and preventing a problematic behavior that could eventually lead to absenteeism, school dropout, exclusion, and later health problems [3]. In addition, there is a lack of research on how absence and/or tardiness among students effects education and the work environment for the other students and the teachers, and vice versa [41]. For example, tardiness is often explained in relation to individual behavior, such as children's sleeping behavior [42], and is seldom related to a school context. We argue that tardiness is overlooked in research and practical work related to health promotion in school. Moreover, we claim that tardiness, as a phenomenon, is overlooked. In terms of a social-ecological perspective on health [43] and a critical perspective on health promotion [44,45], it is crucial to include perspectives on who is most affected: the student. However, most research in this field appears to be based on statistical analyses of absenteeism [19,46]. Thus, there is a research gap in regard to including the student's own perspective on attendance, absence and tardiness, as well as a lack of studies that include stories of those who are part of the social setting of students at risk of absenteeism . Given this background, the goal of the present study is to contribute knowledge about tardiness that can be useful in strengthening practical work at school from a health promotion perspective. Its aim is to analyze students', school staff's and parents' views on students' tardiness. --- Materials and Methods To study students', school staff's and parents' views on students' tardiness, a focus group interview design was chosen. To this end, we created groups of participants and provided instructions regarding discussion themes [47][48][49]. Examples of interview questions to parents and teachers include: "What do you think promotes students' school attendance? and "How do you perceive the responsibility of the school in relation to students' school attendance?" To the students: "Does it ever happen that you arrive late to school? "and "When does it feel good to come to school?" Two researchers participated in each interview: one led the discussion while the other handled the audio recorder, took notes and observed [50]. --- Participants and Procedures This study took place from January to June 2017 in a mid-sized town in the northern part of Sweden. The schools were located in a suburban area primarily comprising middle-income households. The participating schools were both compulsory schools. From the age of six, Swedish children are allowed to start primary school and most of them do so; from seven to fifteen years of age, school is compulsory. The first school , with approximately 200 pupils, went from primary class to grade 5. The second school , with approximately 350 pupils, included grades 6 to 9. The headmaster at the school with the younger children selected nine school staff members who participated in a focus group lasting one hour. At the school for students in grades 6 to 9 , the research group introduced themselves to the staff, presented the study, and asked for volunteers for the focus groups. Staff were also given an information letter and a list was left in the staff room so they could sign up for the focus group. A total of twelve school staff members accepted the invitation. They were divided into two groups: one with the school health team and one with teachers . All parents at school B were invited by e-mail, and three of them accepted. One of them was prevented from participating, so the interview was carried out with two parents. Based on Swedish ethical law, teenagers aged 15 are mature enough to decide on being research participants [51]. Therefore, only students of that age were invited to participate in the study. This was also decided because students in grade nine have the most years of school experience. One of the authors introduced the study to the students in class and asked for participants. The author also spoke with the students about the study during breaks and reminded them of the study and invitation. Twenty-one students in grade nine accepted the invitation. Mindful of Wernersson and Ve's [52] finding that boys tend to dominate girls in groups and of the overrepresentation of boys who wanted to participate, we generated two single-sex groups and one mixed group. The students decided on their own if they wanted to be in the single-sex or mixed group. Group 1 included nine boys, group 2 included five girls and group 3 was a mixed group . All interviews except one were held during the day in a separate room in the school building. The parents were interviewed in the evening in the school library. The two focus groups with teachers lasted from 60 to 90 min, the focus group with the school health team lasted 60 min, and the focus group with the parents and pupils lasted 30 to 40 min. All interviews were audio-recorded and transcribed verbatim by an independent transcriber. --- Analysis The material was analyzed in order to determine which themes best describe the content, by using a process inspired by Braun and Clarke [53]. In the analysis, all transcripts were initially read by three of the authors independently. Based on these readings, themes describing underlying aspects of the stories told were constructed. That is, specific parts of the transcripts were highlighted and grouped as the analysis was conducted. During this process, delimitations were constantly discussed and challenged in the research group. The researchers also noted down their thoughts and interpretations while reading the material. The themes were carefully reviewed several times before arriving at a result. One overarching theme and three subthemes are the focus of the reported results. These subthemes are not mutually exclusive; rather, the subthemes overlap, as one leads to the other. Together they tell a story about students', staff's and parents' views on tardiness. The themes should be viewed as representing patterns of shared meaning throughout the interview transcripts. The original spoken language in the interviews was Swedish. At the end of the analysis process, the results, including quotations, were translated into English. --- Research Ethics At the outset, all participants were given oral and written information about the project and were informed of their right to withdraw at any time without explanation or negative consequences. All participants signed an informed consent. The Regional Ethical Review Board in Umeå, Sweden, found that the study was not covered by the Ethical Review Act and did not consider the study as an issue of ethical concern . --- Results We began our study with the concept of being present in or attending school and class. During an interview with two parents , both mothers, the importance of students being on time for school and class were emphasized. The parents viewed school attendance from a broad perspective and on various levels. That is, punctuality was understood as a skill for future life and a way to be respectful to others; however, they also found that students must bring books, bags and other items with them, in addition to being on time for their lessons: Yes, that they are in place, at the lessons. And on breaks and lunches too, of course. So, they are in place on school days, that's what I think. Yes, that they are on time and are here when they should be. Yes and have what they need to bring to school, like books, clothes for physical education, bags, or whatever it is. The parents' statements on students arriving on time, collected at the beginning of the study, inspired us to further explore how tardiness was viewed by students, parents, and school staff. The analysis relates to how tardiness is viewed in the material. The thematic content analysis showed that perspectives on arriving on time at school are framed by the main theme "It depends on . . . " This main theme is supported by the following underlying subthemes: Signals, Reactions, and Responses. Here, the main theme will be presented first, followed by the subthemes. It depends on . . . Throughout the material, "It depends on . . . " permeates the analysis. Students' actions were viewed as depending on who the teacher was and how he/she behaved when they arrived late. Moreover, students' actions were influenced by many factors: the importance of the school subject, the motivation towards school in general, support from home and time of the day. Similarly, the teachers' understanding and handling of late arrival was related to a number of various circumstances: the teacher him/herself, the student, the circumstances of the late arrival, and lateness frequency. When it came to the parents' views on the issue, the answer was it depends on . . . ; thus, they told their stories about families and teachers. However, the staff on the student health team at school B, did not explicitly mention tardiness but discussed school attendance and truancy in general terms and mentioned the importance of collaborating with parents. That is, from their perspective, it depends on the parents if students arrive at school on time or not. Parents were seen as either supportive or insufficient in their role. In relation to the main theme, and shedding light on the complexity of tardiness, the developed subthemes in the analyzed material is as follows: tardiness depends on what signals are perceived regarding being on time; how tardiness is met by reactions of various kinds; and how students respond to expectations and actions related to tardiness. Notably, these subthemes should be seen as a product of the analysis aiming to describe the material in the best way possible to answer the aim of the study. The subthemes should, therefore, not be understood as mutually exclusive but, rather, as mirroring the complexity of the issue. --- Signals The first subtheme, signals, includes accounts of unclear signals about the importance of being on time. The signals relate to expectations and norms and include issues of structure, predictability, and belongingness. The mothers saw no consensus among the school staff regarding how tardiness is viewed and handled in school. They explained the various ways in which teachers dealt with students as disparities in teachers' tolerance and as normalizing problematic behavior. The mothers requested higher expectations from the school regarding students being on time. They expressed that it is important from a security aspect that the school notice students arriving late. As a parent, you want to know that your child is at school and nowhere else. Furthermore, they believed that the students themselves would like to feel they were expected to be on time and that this emphasis is about respect and routines. A time is usually to be held to for a reason. So if you do not show up or do not bother being on time, then you do not show respect for it. Everything I think is about consideration, respect, and cooperation. We are numerous at the school; it should work for everyone. And then slipping in when you feel like it or not showing up at all, this affects oneself first and foremost but affects others as well. You get into it, there just has to be a routine...In our family, we are useless at being on time, but . . . you have to try. And then there are surely families who, 'but my goodness, come on time, it is well a worldly thing, there are worse problems in life'. But somehow, the school must uphold routines. If you don't have a routine in school...and do not have one at home, where do you have it, if you are a youth? You have these two platforms and then perhaps a hobby. And then there must be security in this. The mother's story illustrates the importance of respect for others through punctuality but also of the school itself as being responsible for maintaining structure, which can be understood as helping families struggling with punctuality. The teachers in the two focus groups, at school A and B, had different views on late arrival. Some of the differences were between the two schools or related to what grades they taught. Some teachers expressed that they expected students to be on time, while others did not seem to care if students were a few minutes late. It seemed to depend more on the situation than on the student. The teachers were said to be more understanding if their class followed, e.g., physical education, as well as if the student was usually on time. If a group of students arrives late, one teacher said, she directs her attention instead to those who arrived on time and praises them. Furthermore, in the material, arriving late seemed not to be viewed as being equivalent to being absent from an entire class. The teachers said that being on time did not seem to be important for some students. This idea was extended to families; that is, teachers in the lower grades said that not all parents found it important to help their children be on time. These teachers also expressed more responsibility and caretaking that fosters a sense of belongingness. One of the teachers said, So that you...And that you get in contact directly, so..., if a student does not come to school, for example, and you have not received an illness report or any other absences, then you must call immediately, of course. The school staff in this school also discussed the importance of being clear with the students about what is expected of them and what happens when they do not live up to those expectations. They begin arriving a little bit late from breaks, linger, don't take things seriously, or the student left home a little bit late. And of course, it affects others too, those who are always there on time [laughter]. So, of course, that is part of the reason why we have tried to work on it. The students' stories also include descriptions of blurry signals regarding expectations about arriving on time. Depending on the reactions from the school, the students maintain there are no clear norms regarding punctuality or being on time for lessons. --- Reactions The second subtheme is about how tardiness is handled by teachers. In the focus group interviews with students, the most common description of what happens when they are late was that no one reacts. In one of the focus groups, the girls were asked what would happen if they arrived late. The reply was that the teachers most often do not respond at all to late arrivals; however, it depends. One of the girls said, "It depends. Sometimes they do not react at all". Another girl continued, "Often" and a third girl said, "I would say that they do not react at all." Similarly, the boys said "It depends", "It is different ". That is, the students pointed out that sometimes being late was met with no reaction but, if handled, the reactions vary from teacher to teacher. Some simply noticed it and said "Welcome", while others punished latecomers by screaming and shouting at them. Therefore, teachers' reactions to late arrivals were seen as being unpredictable in some ways. The material analyzed showed that these variations are understood by the students as relating to a specific teacher or, on the other hand, to a specific student. That is, all students are not treated equally. These variations recurred independently of interviewees . Students expressed that late arrival was sometimes recorded in the electronic system but sometimes not. If the students did not notice such a record, they interpreted it as "no reaction" from the teacher. Registration in the electronic system decided whether the student was judged as being late or not; this was in some ways predictable for the students, e.g., which teacher would register tardiness, but even among those noting it, the handling of the situation varied. Some teachers at school B noted every single tardiness minute while others recorded only if someone was really late, although what was considered "really late" differed among teachers. The teachers at both schools and the SHT spoke about a 20-percent limit to school absence measured from the beginning of the term. When a student exceeded that limit, the SHT initiated a meeting with the parents. Some teachers said they used a strategy to ignore students when they were late to class: they just went on with their lesson without paying attention to the latecomer. Further, some teachers simply asked the latecomer, "why so". The analysis shows that what the students define as 'no response' to a late arrival sometimes means a response but a non-social one. Students relate responses to social reactions in the relationship between teacher and student. One of the boys did not want to be asked for an explanation as to why he was late: "It is good if they ask why you are late, so you can explain . . . There may be reasons". A girl wanted more caring questions from the teachers: "They don't ask us 'Yes, why are you late?' or ask for the reason why we're late. They just note the absence. They are so used to it". The girls had different opinions about the teacher's caring questions; although they do not want to be absent for a long time without the teacher noticing, they wanted to be missed. In addition, in the mixed group, questions about being late were considered preferable. Although a lack of reaction to tardiness was found in the material, the opposite was also interpreted from the stories told, with the analysis showing that powerful reactions from teachers are also important. Teachers who felt provoked or interrupted sometimes reacted aggressively. Their various stories about situations involving late arrival showed how teachers feel about youths. Some teachers described students "flinging" and "kicking on the door", or simply waiting until the lesson began and then coming in. Other teachers, at the same school, based the stories of their ambition to create a welcoming classroom with opened doors and on thoughts about why students arrive late or drop out of class or even from school. In regard to reactions to tardiness, doors played a central role in the material from school B. The students saw a closed door as a clear signal that the lesson had begun; on the other hand, the door could be open even if the lesson was in progress and the students could just step inside. This was experienced by the students as being unclear and showed that lessons could begin in various ways depending on specific teachers and how they set up their own lessons. The girls understood that teachers do not want to be interrupted when they begin a lesson: Respondent girl: She held a review . . . then it is another thing. If she is holding a review. Interviewer: If she has a review, then she reacts? R girl: Yes, then she doesn't want people knocking on the door: it disturbs everyone else and also her review. Other teachers not only close their door but they lock it, which extends the length of the absence that will be reported: Some teachers, they do this... If you are a little late, you cannot get in; then, you get marked even more absent than you really were . . . So, you really were earlier than the report shows because the teacher didn't let you in. And then you want to be on time. You do not want to get an unnecessarily long absence. The teachers lock the door, not just close it. Furthermore, in one focus group with teachers , locking doors was pointed out as being a controversial activity related to security aspects. "The rules about locking doors have been changed", one teacher said, after a crime was committed in a Swedish school two years earlier when a young man killed a student and a teacher and seriously hurt a group of young students. The other teachers in the group took no notice of this comment and continued describing their strategies in handling late arrivals and how they felt provoked by students knocking or kicking on the door to get in. Unpredictably, students mentioned the strong reactions from teachers. In some cases, they said that the teachers reacted more strongly to someone arriving late and knocking on the door. It seemed to depend on what was going on in the classroom at the time. The students said that teachers seldom reacted if the students in the class were working individually while someone arrived late. On the other hand, they pointed out that some teachers not only respond negatively to being interrupted but also to other events in the classroom. There were accounts of teachers frightening students by their behavior, one student reporting, "One may be scared of some teachers, so then one doesn't want to be late. They may be . . . very angry". The students who encountered screaming and shouting teachers said that they could avoid such treatment by being on time, but their narrations also contained accounts of various behaviors among teachers, behaviors that are not always predictable. However, in one of the groups, when 'shouting' teachers were mentioned, this was not considered a constructive way to handle latecomers, even though students found they were seldom late to classes conducted by those teachers: R boy: It is good that they record an absence but it feels excessive that they stand there and scream. R boy: Yes, exactly. It is our choice anyway if we arrive late or not. Then, he can record us as absent, nothing else. It affects us. Not like this: stand and mess just because you are late. --- I: What do you others say about this? R: Same here. Just go in and sit down, and then they record you absent, and so . . . yes, it will be like that. Then, there is nothing more. You shouldn't have to talk about why you are late and so on; it is just unnecessary. I: At the same time, you said that the teachers who get angry, those are the lessons you go to to avoid it . --- R boy: Yes, then it works. Thus, the students will be on time. --- R boy: Yes, I was thinking the same thing. --- I: And then, the teacher has achieved his/her goal? R boy: Yes . From the above quotations, it is clear that the students believe that late arrival is simply their own choice and that they do not see themselves in relation to others. However, some of the students could see the contradiction in wanting the teachers not to react but at the same time coming to class to avoid an aggressive teacher. --- Responses The last subtheme focuses on responses; that is, issues that strengthen the willingness to be on time and that contribute to tardiness and absenteeism. In the interviews with the students, various perspectives on the relevance of being on time were emphasized. Some expressed the view that classes important for high school were prioritized, others prioritized classes with teachers they found kind and interested, and still, others mentioned the importance of being with friends. Students also found that the short time was a motivation for going to school and being on time. For the most part, students wanted to avoid such aspects that create discomfort. They described angry and shouting teachers and angry parents, but there were also aspects that could be a motivation in the long or short term. One boy, who had been truant and tardy a lot, said: "I had to start thinking about getting into high school. So, school started to feel more important". He was motivated by thoughts of his future life. Some students discussed being late as their own choice, according to how seriously they view school and education. They were more motivated to attend some of the lessons. That is, they were found prioritizing various subjects; some were noted as being more important, and others were considered more boring. Thus, students said they were seldom late to those prioritized classes and more often arrived late to other classes. There were also differences in maturity among what was considered motivating aspects. One boy was motivated by more urgent needs satisfaction: to avoid conflicts with his mother, he had coffee at school in the morning; nothing else motivated him: R boy: Yeah, I don't know. Not so damn much. I have to come here; otherwise my mom will not be so damn happy if I'm home every day. I: It is your mom . . . . --- R boy: Yes, she forces me to go to school. --- I: Yes? R boy: Yes. I: So, it's just about force, never any strong desire or anything ? R boy: No, damn it. You get coffee here. I: You get coffee? You don't get that at home? R boy: Yes, damn it, but I have to make it myself; I can't stand it. In order to understand what motivated students to make them be on time for school, or even go to school, we asked them to describe their "dream school". They told stories about starting later in the morning, better teachers, better technical equipment, better food, and more variety and flexibility. The relevance of motivation was found in the parents' stories as well. However, parents' perspectives on motivation were related to the future of their own child. That is, school is important in obtaining a higher education and/or a job. In the material based on the focus group interviews with school staff in both schools, this perspective was not made explicit. When the staff talked about motivation, their stories were mostly related to how they did not find that students and/or parents prioritize school. The staff questioned the students and/or parents when it came to prioritizing being in school and/or on time for class. When the school staff were asked to describe an "attending" student, they mentioned individual aspects such as being curious, motivated, healthy, idea-rich and willing to learn. In addition, aspects related to their social environment were mentioned: being pushed by their parents, living in a family engaged with the school, other activities outside school and having friends. The teachers in the secondary school , on the other hand, noted the importance of being welcoming and showing that they, as teachers, were happy to see the students, and of developing good relations between teachers and students. In terms of pedagogy, they tried to make education "old school", that is, predictable and without too much involvement of the students, in order to make it easier for them. From their perspective, this was a way to motivate their students. --- Discussion The results showed that tardiness, viewed by students, parents, teachers and other groups working in the school, including the student health team, is a complex issue. However, based on various signals relating to how important attendance was understood to be, how eventually tardiness was reacted to, and the students' responses, the pattern we found illustrates the unpredictability encountered by the students, whereas tardiness itself was not viewed in similar ways by the focus groups, nor was it handled equally. These results are discussed below, first in relation to the theme developed from the analysis and then based on our main theme and how it could be understood from a social-ecological perspective. --- Signals Parents and students interpreted the signals from school as indicating that punctuality was not that important. This can be understood as meaning that teachers have no expectations of you to be at school on time. Previous studies show that school attendance is a sign that the school is well structured and predictable for the students [26]. Buhler, Karlsson, and Österholm [54] found it important to promote school attendance by starting and ending the school day in the same way, every day. Teachers ignoring latecomers also indirectly send out signals that you as a person are not important. This result is in line with previous findings [55]. Expectations of how people act in different situations produce norms related to how we behave in a specific environment and indicate what the culture is. Signals can be used to create a sense of belonging or create a feeling that you do not belong to this place or culture. Belongingness has been found as essential in the health-promoting school model described by Rowe and Stewart [56]. To feel included in and connected to school is important for health and school achievement [57]. Wenzel [58] found that perceived support from teachers measured as, for example, "My teachers really care about me" and "My teachers like to help me learn", was a positive predictor for student motivation, school interest, and class interest one year later. --- Reactions Latecomers encountered different reactions. Schools have no shared and clear policy or rules to handle late arrival: some teachers locked the door, others registered the tardiness in the electronic system, some shouted and still others took no notice of the latecomers. The predictability became even more unclear from the students' viewpoints, because sometimes the same teacher reacted differently to the same situation involving different students. This result shows that not too much seems to have happened since D'Amico [32] carried out his research and found no consensus among teachers. On the other hand, treating students unequally might be how teachers interpret compensatory assignments [39]. Swedish compensatory assignments offer support related to the student's needs for equal access to education. With students coming from a vulnerable family and/or environment or with at-risk students, teachers probably want to be welcoming and caring in order to mitigate their vulnerability. The results indicate that teachers from school A saw the students' wellbeing and school achievement as a part of their family and the entire living situation. Teachers who sometimes emphasized the differences among students created an unjust school environment. This was also found in a Swedish [59] qualitative study, where the students reported feeling like "black sheep" or failures, while other students were described as being teachers' pets or high achievers. The "black sheep" described how teachers' behavior, for example, shouting and screaming at them, had affected their self-esteem and interest in schoolwork. In addition, Banfield, et al. [60] described "offensive teachers" who humiliated students, picked favorites, and were rude or sarcastic, which in turn affected the students' wellbeing. Handling tardiness by being offensive to students is in line with D'Amico's [32] description of reducing negative behavior. On the other hand, Freeman and colleagues [33] found evidence that skills training and family support, in addition to incentive-training, to increase the level of student punctuality strengthened positive behavior. The students in our study asked for more caring questions from the teachers and for greater openness to being exposed to unforeseen events. This could be seen as young people's expectation of a relational perspective on teaching [61]; in addition, a relationship with or interaction between teachers and parents was also mentioned by mothers, an idea related to the fact that the primary means of communication was expected to be via the electronic system. Mothers, students, and teachers alike mentioned security aspects but related to different situations. Students and mothers wanted someone to start looking for latecomers because something may have happened to them on their way to school. Teachers in the lower classes also thought along these lines. Secondary school teachers, on the other hand, mentioned security in relation to school attacks and the need to keep the students in a safe place, away from external threats. School shootings or school attacks have been a reality for many years in Western countries [62]. The risk that something would happen to a student on her or his way to school is more likely than a major attack in the school itself. The responsibility of parents and of the school towards the students is an ongoing topic of discussion in Sweden, but in some ways, the rule is clear: parents are responsible for their children arriving at school on time. The school is responsible when the child is at school, but the responsibility is twofold between school and parents [63]. This could explain the differences in the focus of the teachers, parents, and students. --- Responses Predictability, a culture that creates a sense of belonging by giving caring and supportive signals in school, affects whether the students arrive on time or not. Findings [64] reveal that a negative sense of school belonging has a negative impact on intrinsic motivation and perceived learning. The students saw themselves as solely responsible for arriving on time, and this was only related to their own decision. On the other hand, their responses showed that their decisions were influenced by teachers' behavior, parents' expectations, future plans for higher education and, in some cases, simply by the short time period at their disposal. Another finding was the school staff's discussion about motivation. Motivation was described from an individual perspective as a characteristic of the student or coming from a supportive family. None of the teachers mentioned motivation as the responsibility of the school organization or as part of pedagogy, even though a number of previous studies have noted that motivation is part of classroom pedagogy [65]. The teacher-learner relationship is also an important factor in creating engagement and promoting educational outcomes [66]. We interpret this finding to mean that this school had not created a sense of belonging for the students. Nor had they, from the students' viewpoint, created a desire for learning. Both students and parents asked for better interaction and communication with the teachers and the school. The increased digitalization of school involves the risk of losing communication and interaction with parents if all primary communication is effected through texting and through registration in an electronic program. Parental engagement in school has been shown to be important for school achievement [67] and for a health-promoting school [16][17][18]. The results show that teachers do not differentiate between school refusal behavior and truancy or school withdrawal, and neither does the school health team. They focus more on absenteeism percentages and inputting these into the electronic system to begin analyzing the problem if the absenteeism level exceeds 20 percent. The reason for tardiness must form part of the analysis that determines effort. Universal and individual efforts do not contradict each other: both are needed. In this article, we focus on how the school can be a supportive environment for students and their families in regard to students arriving on time. To promote school attendance, specifically punctuality, one must examine the student and his or her individual behavior within a broader social and cultural context in which the student, the family, and the school interact. The main theme, "It depends on . . . ", will now be discussed from a social-ecological perspective. According to social-ecological theory [43], different environments influence our behavior and health. People's thoughts, behavior and perspectives are transmitted among these environments. We move between different environments: our family, neighborhood, school, work and so on. From a broader perspective, the child's home and neighborhood is also part of the system whose components interact with each other [43,68]. Earlier studies have shown that young people see the interaction between the school environment and their home environment as important for their health [59]. This is in line with the idea behind health-promoting schools [69] and the entire school approach [70]. The main theme "It depends on . . . "was developed to describe the various aspects influencing how teachers give signals and react in relation to tardiness and how this reaction elicits responses among students. The interpretation of "It depends on . . . " provides a picture of an unpredictable situation for both students and parents. As mentioned above, predictability and structure during the school day have been shown to be important factors for school attendance, especially for children with school-refusal behavior [71]. In addition, other studies illustrate the importance for all children of consequent, clear and equal rules and norms [72]. The mothers interviewed in the present study stressed the importance of the school as being a place for structure and safety and encouraging an "arriving on time" culture, based on the fact that some children live in a vulnerable situation in families with many social problems. This finding is in line with the directive from the Swedish government about compensatory assignments [39,40]. The social-ecological theory explains how specific cultures develop in different environments. The students move between school, their homes, and places for spare-time activities. These places are culture carriers, but the students affect and are affected by these different contexts when they move and interact within them. The common setting for the students is school, while the other settings differ more or less, given that homes, neighborhoods and to some extent, settings for sports and other activities, differ according to the families' socioeconomic position [73]. Schools in Sweden are obliged to provide equal access to education and shelter, to be compensatory and to strive to equalize inequalities, although sometimes the opposite is the case, even in Sweden [74]. In line with this, the teachers of younger students see their work in encouraging punctuality as a part of their educational assignment, as an interaction between how they organize their schoolwork and how the students respond, and how this, in turn, affects the classroom environment. This example illustrates the ecological system and the constant movement within that system. Individualized education, on the other hand, where teachers say, "Take your books and continue where you are", creates a view of the individual student as an isolated island, without any connection to the environment. This reductionism assumes that a system can be broken down into single components, which is the opposite of a system in which parts interact and must be understood in relation to the whole [75]. However, why should students arrive on time if they do not assume they relate to others or to the environment? Lastly, the results from our study indicate that tardiness is understood by the participants as a problem related to both the school and the family, also observed in research on adolescent adjustment; illuminating not only the students' ability to adapt with school environment [5,6] but also the impact of parenting on these abilities. That is, positively engaged parents can contribute to adjustment [7][8][9][10]. That means, vulnerable families need more support from schools and maybe also social welfare services, on how to become more involved in their children's education [76]. Altogether, the social-ecological theory [43], as well as the school adjustment perspective [5,6,8], illuminates the importance of family involvement. To promote school attendance among students, a social perspective including the whole social context of the adolescent is necessary. --- Methodological Discussion A focus group interview design was chosen for the present study. Students, school staff, and parents were invited to participate in order to provide different perspectives on the topic of our study. School staff were deliberately interviewed in groups based on their profession since we assumed that they would have different views on tardiness depending on their professional role and experience with students. It turned out that they did speak about tardiness in different ways. Two parents agreed to be interviewed. Preferably, one to two focus groups should have been performed to give equal weight to the stories of students, school staff, and parents. This approach would have strengthened the results of the parents' perspectives on tardiness in the present study. The two participating parents provided valuable insights into their views on tardiness. In fact, it was the interview with the parents that introduced tardiness as a topic of interest and led us to study it in more depth. Recruiting students ensures that adolescents from various backgrounds were included since school is mandatory. However, students absent from school when recruitment was carried out may not have received the information and were thereby excluded from the focus groups. Students absent at the time of recruitment are probably more likely to be absent from school in general and, if interviewed, would perhaps provide important insights into the reasons behind tardiness. Conducting individual interviews instead of focus groups would perhaps have decreased the risk of peer pressure among the students when interviewed and made it more likely that they spoke their minds. On the other hand, focus groups with peers could be a more comfortable venue for students than being alone with the researchers. In our opinion, interviewing the students in same-sex or mixed groups based on their own choice made it more likely that the participants were comfortable and could express their opinions. In addition, focus group interviews are preferable when seeking insights into meaningful themes embedded in discussions of the topic chosen. The included schools, school staff, parents and students were not representative, although we included both a primary and a secondary school, and the staff members interviewed represented all key professions. Our study found that a thematic analysis provided the best opportunity to answer our research question. Criticism of a thematic analysis includes individual accounts and language use being lost, and the flexibility of the method making it difficult to know what aspects to focus on. We aimed to present patterns of shared meanings. It turned out to be difficult to create themes that were mutually exclusive, as recommended by, e.g., Braun and Clarke [53]. Nonetheless, we believe the thematic analysis is the best analytical categorization for the material in our study. The present study used social-ecological theory and health-promoting perspectives as a framework to strengthen our interpretations. The involvement of three researchers in the analysis process and a continuous discussion during all steps of the analysis strengthened the process and the credibility of our results. All authors agreed on the final themes. We have also described our methodology and analysis in detail to allow readers to form an independent assessment of credibility. To judge our study's resonance, the preliminary results were presented to and discussed by some teachers and the principal from another school for feedback. The results made sense to them. --- Conclusions The picture that emerges from our analysis shows both differences and similarities among the groups participating in the study. Parents and students shared the same idea about how schools handled lateness, but their views on responsibility differed. Some of the students saw only their own responsibility as a factor affecting punctuality. The inability to see their role as part of a greater whole was more prevalent in secondary school teachers than in teachers in the lower grades. Late arrival is a sign of adolescents' maladjustment from a holistic viewpoint, in a school as an organization created to interact with students, families, and school staff. The results of this study imply the importance of organizing the school day more predictably and with a better structure for the students. Such predictability can be expressed by starting the school day at the same time and in the same manner every single day. Further, the staff needs to agree on a common policy on handling tardiness. The result also indicates that schools must further develop their effectiveness in relation to both students and parents. That is, they must strengthen their work in promoting students' sense of belonging and in relation to students' interests and motivation to come to school. In dealing with tardiness especially, it is important to interpret this behavior as a signal understood in a greater context. In further research, we suggest developing a model of how to handle student tardiness based on predictable standpoints. This model should be developed and tested in collaboration with researchers, teachers, students, and parents. --- Author Contributions: Conceptualization, E.W. and M.W.; methodology, Å.S., E.W. and M.W.; validation, M.W.; investigation, Å.S., E.W. and M.W.; resources, L.T.; writing-original draft preparation, Å.S., E.W., L.T. and M.W.; writing-review and editing, E.W. and M.W.; supervision, M.W.; project administration, L.T. and M.W.; funding acquisition, L.T. and M.W. All authors have read and agree to the published version of the manuscript. ---

Tardiness is a common problem in many schools. It can be understood as an individual risk for future problematic behavior leading to absenteeism, school dropout, exclusion and later health problems. Tardiness can also be examined in relation to a broader social-ecological perspective on health. The aim of this study was to analyze students', school staff's and parents' views on students' tardiness in two Swedish schools. A focus group interview design was used with 21 school personnel, 21 students in grade nine and two parents. The data were analyzed by using thematic content analysis. The results illustrated the main theme-It depends on . . . -regarding what will happen if a student arrives late to school lessons. This finding is further explained by the subthemes about teachers' signals and reactions and the responses from teachers and students. The conclusion showed the importance of organizing the school day more predictably for the students. Late arrival is a sign of shortcomings in a school organization. It is necessary to develop guidelines related to how to handle students' late arrival based on predictable viewpoints but even more so on how to promote students' sense of belonging and their interest in and motivation for going to school.

Background The contribution to the global burden of disease of mental and behavioral disorders is substantial and is estimated to be the first leading cause of years lived with disability in 2010 worldwide [1]. A quarter of individuals will likely present at least one such disorder during their lifetime [2]. The epidemiological literature has demonstrated a high prevalence of these disorders in both developed and developing countries [3,4]. Like communicable and non-communicable diseases, mental disorders contribute to the poor health of populations, and poor mental health has been cited as a hindrance to efforts underway to achieve the United Nations Millennium Development Goals [5,6]. To date, mental health is rarely taken into consideration in the public health policies of low-income countries, particularly in sub-Saharan Africa [7][8][9]. However, as these countries move through the epidemiological transition, the burden of disease shifts from being primarily attributable to communicable diseases and early-life mortality towards a greater contribution to the disease burden from non-communicable diseases . Consequently, among an increasing proportion of adults, NCDs, including mental and behavioral disorders, become more prevalent. Among mental and behavioral disorders worldwide, depressive disorders were the leading cause of YLDs in 2010 [1]. Though they are increasingly common among societies which move away from traditional rural modes of organization into modern and urbanized ones [10], major depressive disorders are not specific to the wealthy and remain prevalent in varied cultural and economic contexts [3,4,11]. Because of the substantial impact of major depressive disorders have on YLDs and the widely established methods for measuring major depressive episodes, we choose to focus specifically on measuring and analyzing the prevalence of major depressive disorder in our study. A major depressive episode indicates the onset of a major depressive disorder. Such an episode is diagnosed on evidence of persistent symptoms affecting a person's mood, perception, and engagement in daily life and normal activities over a specified 15 day period. It is believed that specific symptoms may be less prevalent in some cultures than in others [12], though this view is not unanimous [13]. The symptoms used to diagnose a major depressive episode are identifiable in all cultures [14], and diagnostic criteria have been standardized at the international level [15,16]. The manifestation of major depressive disorder arises from a combination of mechanisms that can be classified into three groups: biological factors, which are relatively poorly identified, psychological factors involving stressful events , and sociodemographic factors [2]. In low-and middle-income countries, poverty has been cited as underlying factor of mental disorders, including major depressive disorder, and there are believed to be interactions between poverty and depression [2,4]. Empirical studies have come to inconclusive findings on the relationship between the occurrence of a major depressive disorder and socioeconomic factors, specifically those related to inequality [17,18]. More consensually, major depressive disorder is consistently associated with individuals' overall state of health [6,19], disabilities [20], risk behaviors such as smoking and alcohol consumption [21,22], and with mortality [20]. Major depressive disorder prevalence remains largely unknown in many sub-Saharan African countries. Studies carried out to date have been single-country studies, almost exclusively in Anglophone countries, rarely conducted at national level, and have produced a wide range of prevalence estimates. Two recent studies that estimated the 12-month prevalence for major depressive episode found low or moderate rates, including a survey conducted by the World Health Organization in Nigeria which found a prevalence of 1 % [7,23] and a survey in South Africa that estimated a prevalence of 4.9 % [24]. A review of surveys carried out in Ethiopia found a higher range of prevalence, from 7 to 11 %, depending on the measurement tool [25]. Conversely, some studies have found much higher prevalence rates, including some studies dating as far back as the 1960s [26] and a 1997 study estimated more than 30 % of women living in urban areas in Zimbabwe had experienced anxiety or depressive disorders in the past 12 months [27]. The difference in prevalence estimates, measurement tools, and variation in periods of observation means there are no generalizable estimates of the prevalence of major depression disorder throughout the region. The rapid urbanization which currently characterizes the African continent is associated with changes in socioeconomic and environmental conditions that are theorized to affect people's mental health, though process has not been sufficiently appreciated [28]. These conditions include frequent migrations and the weakening of social ties, as well as increases in stress and violence levels related to urban living [4,12,29,30]. Moreover, while poor urban residents are on average better-off than rural residents, they may experience a greater sense of social inequity, as they often live side by side with the richest segment of the population that is concentrated in cities [28]. Also, while adults in cities in the global south generally have better health outcomes than their rural counterparts, at a global level they still experience relatively high levels of infectious diseases, premature non-communicable diseases, accidents, and difficulty accessing health care. Despite indications that urbanites in sub-Saharan may have increased risk of experiencing major depressive episodes, the mental health of these urban residents is often overlooked, in part because of a lack of reliable estimates of the prevalence of mental and behavioral disorders. This, in turn, may hamper the establishment services and training of skilled staff to prevent and treat these disorders among the growing urban population. Our study focuses on Burkina Faso, a landlocked country of 15 million in West Africa. Mental health care in the country is provided in hospital psychiatric wards or the psychiatric care units of medical centers but is grossly inadequate in meeting the population's needs, even in the urban centers where such services are concentrated [8]. Like many countries in sub-Saharan Africa, Burkina Faso has seen a substantial increase in the proportion of its population that resides in urban areas. While Burkina Faso's population was only 14 % urban in 1990, but it is estimated by 2015 that nearly 30 % of Burkinabe will live in cities [31]. Ouagadougou, the capital, is a sprawling city with over 2 million inhabitants and estimated to be the second-fastest growing city in sub-Saharan Africa [32]. As urbanization continues in Burkina Faso and across sub-Saharan Africa, an ever-increasing proportion of the population will be found in cities, making understanding mental health in these growing urban areas increasingly important for improving health indicators throughout the region. To our knowledge, no previous study on depression has been conducted in Burkina Faso. The information collected in the 2010 Health Survey enabled us to estimate for the first time the proportion of urban adults in Ouagadougou who had experienced a major depressive episode at the time of the study, as well as to identify associations between a major depressive episode and demographic, socioeconomic and health status characteristics. --- Data and methods This study uses data collected as part of the Ouagadougou Health and Demographic System Site , which has been monitoring the inhabitants of five neighborhoods located on the periphery of the city since 2008. Two of these neighborhood areas are planned formal neighborhoods, while the other three are informal settlement areas [33]. Residents in the informal settlements are on average poorer than those in the formal areas, are more likely to be migrants who were born in rural areas, and have higher levels of mortality [34]. The protocols for the health survey and routine HDSS data collection were approved by the Ethics Committee for Health Research of the Ministry of Health of Burkina Faso. All participants provided their written informed consent to participate in this study. 1 All households from the original census were then visited on average every seven months to record births, deaths, changes in marital status, and migrations for all members of the households. As of January 2010, the Ouaga HDSS was following 76,452 residents, 61 % of whom were older than 15 at that time. In 2010 a specific health survey was conducted on a representative sample of 2,187 individuals among the monitored Ouaga HDSS population in order to collect detailed information about health care, health behaviors, health problems including current major depressive episodes . Prior to the implementation of this survey, interviewers were given a one-week training on the Health Survey questionnaire by a physician and a social researcher. Interviews were face-to-face with the respondents and information was directly inputted on pocket PCs, and this data was managed directly by the Ouaga HDSS team. A multistage stratified sample survey was conducted to select households by grouping people by age groups . After selection of one household, all eligible members were interviewed. Ouagadougou is dominated by the Mossi ethnic group and most inhabitants speak Moore as their first language, although French, the official language, is widely spoken among the educated. The module devoted to mood disorder was translated from French into Moore 2 and offered to respondents in one of these two languages; as during the initial HDSS census, 62 % of the interviews for the 2010 HDSS Health Survey were conducted in Moore and 35 % in French. Ultimately, 2,195 people over the age of fifteen were interviewed for the 2010 HDSS Health Survey, but eight of them did not answer the specific module because their proficiency in Moore or French was considered insufficient by the interviewer, resulting in a final sample of 2,187 individuals. To identify major depressive episodes in respondents at the time of the interview, we used the Mini-International Neuropsychiatric Interview questionnaire, which was developed using the psychiatric criteria from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders [16,35]. Using the MINI format, respondents were asked two questions about 1) overall mood deterioration and 2) loss of interest in usual activities. People who said "yes" to at least one of these first two questions were then asked a set of seven additional questions that concerned appetite, sleep, behavior, fatigue, self-esteem, concentration and suicidal thoughts. Following the MINI criteria, a major depressive episode was diagnosed when a respondent said that he or she had consistently experienced at least five symptoms, including at least one of the first two, during the last fifteen days. We calculated the proportion of respondents experiencing a major depressive episode at the time of the study among the different groups of population. These descriptive analyses were weighted and stratified to take into account the design effect. We built regression models to investigate associations with the risk of being depressed. For the multiple logistic regression analysis, we present three models: the first model includes only demographic characteristics, the second model also includes socioeconomic characteristics and the third model adds in health characteristics which also strongly interact both with demographic and socioeconomic characteristics . Independent variables were selected to identify characteristics commonly cited as linked to major depressive disorder in the literature [2,[17][18][19][36][37][38]. Data on social, demographic, economic, and health characteristics come from two sources. First, we use the basic sociodemographic information collected as part of the HDSS, which includes data on individuals' sex, age, ethnicity, religion, marital status, place of birth, education, type of residence , and a measurement of household standard of living. For the household's standard of living, we created a proxy variable based on an index that accounts for the presence of durable goods and the most expensive mode of transport available in the home. The coefficient attached to each good is derived from a principal components analysis . Four categories of households were defined. The poorest households do not own any of the listed goods , while most of the wealthiest households owned a refrigerator, a television and a motorbike, and half of them have a car. Second, we use more specific data on the same individuals collected from the 2010 HDSS Health Survey. This includes information on reported chronic diseases, obtained by asking the question "Has a doctor or nurse ever told you that you have/suffer from…?" and listing a series of common diseases and health problems . Data was also collected on whether respondents had any accidents during the last 12 months that required treatment, physical violence suffered in the last 12 months, moderate or severe functional limitations , alcohol consumption, smoking, and any household food shortages during the past month. --- Results and discussion One respondent in ten reported feeling either "depressed, sad or empty" or to have "lost interest or enjoyment in things that they used to enjoy" in the fifteen days prior to the survey. Among this group, 65.7 % reported having trouble sleeping; nearly half experienced a noticeable change in appetite or weight, or felt tired; 35.2 % reported a change in behavior ; 25.9 % a lack of selfesteem; 15.6 % suicidal thoughts; and 12 % had difficulty concentrating. Ultimately, the proportion of people aged fifteen and older who were going through a major depressive episode at the time of the study was estimated to be 4.3 % . The odds of experiencing a major depressive episode were associated with a number of demographic and socioeconomic variables, as well as with poorer health status. A greater proportion of women were diagnosed with a major depressive episode than men , but this difference is small and not significant in the multivariate analysis . There were some significant differences according to age, which showed a bell shaped association with major depressive episode, with the highest risk among the middle-age groups , although the significance disappears after controlling for socioeconomic and other health variables which interact with age in the third model. Among the sociodemographic and cultural characteristics, ethnicity and religion show significant associations with a major depressive episode: OR 2.2 for non-Mossi compared with Mossi, OR 0.5 for Christians compared with Muslims. Notably, migrants were less likely to suffer from a major depressive episode, though this association of type of residence at birth disappears once all other health characteristics are controlled for. Experiencing a major depressive episode was significantly associated with standard of living. Both the poorest and the wealthiest were more likely to be experiencing a major depressive episode at the time of the survey compared to the poorer middle category. The latter has the lowest rate of major depressive episodes and no significant difference from the wealthier middle category in the full model. However, only the poorest have a significant higher risk in the model 3 which include health characteristics. Individuals who live in households with frequent food shortages in the previous month also have a higher odds of being depressed than those who did not report food shortages. Physical health is strongly associated with reporting a major depressive episode: people who reported one chronic disease have much higher risks of being identified as depressed , and the risk is even much higher among those who reported having at least two chronic diseases . We also found a strong association of functional limitations and experiencing a major depressive episode . People who reported having being victim of a physical violence in the past year had higher odds of being depressed than those who had not . Last, people who drink regularly had higher odds of being depressed than those who don't drink . Previous studies on depression in sub-Saharan Africa found substantial variation in prevalence levels, depending on the study location, and most cited measure the prevalence of episodes in the last 12 months or over the entire life time [4,20,24,27]. Our study found that 4.3 % of individuals in an urban setting over the age of fifteen were experiencing a major depressive episode. This estimate is not negligible, particularly considering our analysis limited the measurement of major depressive episodes to those diagnosed at the time of the survey, suggesting an even greater prevalence of the population who experience a major depressive episode over their lifetime. Our results confirm a number of broad findings on the associations with depression from previous studies, including those carried out in sub-Saharan Africa. In our study population, major depressive episodes were strongly associated with self-reported chronic diseases and disability, underscoring the well-established link between depression and other health problems . Several mechanisms explain this significant co-morbidity: on one hand, being sick is a risk factor of a major depressive disorder, on the other hand, having a major depressive disorder may lead to the onset of diseases, hasten their development or even interfere with their treatment [36]. In addition, access to medical care remains limited in sub-Saharan Africa, even in urban areas. Treatment is primarily geared towards communicable diseases and maternal and child health, despite the growing adult population that is more susceptible to chronic conditions and functional problems related to ageing and illness. In Burkina Faso, not only is prevention practically nonexistent for such health problems, but available care for chronic diseases is also extremely rare or simply financially inaccessible because people do not have health insurance. In such conditions, when people learn they have a chronic disease or disability and their health state deteriorates, they may lose sight of the possibility of an improvement in their condition, which could explain a greater propensity for depression. This association may be reinforced by such kind of study based on selfreported information. Women are generally more likely to experience depression than men, although the degree of the gender gap varies depending on the context and the content of gendered social roles [36][37][38]. In our study population, this gender gap is not maintained when holding all other covariates constant. This is not the only study where such result has been found [39], although in both studies the prevalence was higher among women than men and the absence of significance may be simply due to a methodological issue linked to small sample sizes. Regarding marital status, prevalence of major depressive disorder is generally lower among married people than other single people, especially those who are widowed or divorced [40,41]. In this survey, only a few persons were widowed or divorced which probably explain the absence of significance. Having been a victim of physical violence, another known correlate of depression [4,12,29], was also found to be significantly associated with a major depressive episode in Ouagadougou, as was living in a household that experienced frequent food shortages. Regarding unhealthy behaviors, smokers and alcohol drinkers were more likely to report symptoms of a major depressive episode , in line with another general finding from the broader literature on major depressive disorder, but an association is only found for drinkers, and not for smokers. Last, our results showed that the middle-aged population is the most likely to experience a major depressive episode, this confirms findings from most previous studies [36] but in Ouagadougou, socioeconomic and health characteristics explain this association. The findings here related to the association of sociocultural differences and major depressive disorders are more likely to be dependent on the Ouagadougou context and may not be generalizable. For example, the relationship of ethnicity with a major depressive episode is difficult to interpret, particularly as the non-Mossi group in Ouagadougou is small and heterogeneous. While religion has been generally found to generally provide social support and prevents depression, there is little research specifically on Christian/Muslim differentials with the exception of studies focusing on religion as a minority status carried out in some developed country settings. We found that in Ouagadougou, after controlling for alcohol consumption , Muslims are more likely to experience a major depressive episode than Christians, although this could be a reflection of influences associated with religion in this specific context . Finally, while a link between migration and mental health is often cited in the global literature [12], empirical results do not always show a systematic relation, particularly in the case of internal migratory flows from the countryside to the cities [42]. In Ouagadougou, having been born in a rural area was associated with fewer episodes of a major depressive disorder, but this association disappeared after controlling for other measures of health status. Although education is generally associated with a lower likelihood of developing depression [17], we found no significant relationship between education and depression in our survey sample. Previous empirical studies in other settings and much of the theory on depression emphasize the association between depressive disorder and poverty. Our results confirm that the poorest are more likely to experience a major depressive episode [4,17], and this is supported by findings from a previous qualitative research project on poverty and psychological vulnerability conducted in the Ouaga HDSS setting [43]. Though, we found an unexpected non-linear association between major depressive episodes and standard of living among individuals in Ouagadougou. The association between reporting symptoms of major depressive episode and standard of living in our study population follows a U-shaped pattern: it is highest among the poorest and the wealthiest, and lowest among those in the middle, though the association is stronger among the poorest than the wealthiest for which the significance disappears once controlling the health characteristics. The standard of living is here measured based on household's material goods, and does not necessarily reflect the household's financial situation at the time of the study, but rather its capacity to save or invest in past years. This cross-sectional indicator cannot measure an individual's satisfaction with his or her life trajectory, known to be associated with depression [44], as an individual's satisfaction with his or her current degree of wealth often depends on one's previous level of wealth. It is possible that some of those considered wealthy here have in fact experienced a recent decline in their standard of living, living in the periphery of Ouagadougou. Another hypothesis could be that the richest individuals might be leading more modern and individualistic lifestyles, which are believed to be more conducive to the development of major depressive disorder than traditional ones [10]. Alternatively, living in an informal neighborhood could be assumed to be positively associated with major depressive episodes. We did not find, however, that neighborhood type was significantly correlated with the likelihood of experiencing a major depressive episode in this Ouagadougou setting. This study has several limitations that are worth noting. One is related to the self-reporting of information, which could be biased as social factors impact how one person experiences and expresses one's health state. This potential bias holds true for self-rated health in general, but self-reported health has general been shown to be a good indicator of objective health [45]. Moreover, the identification of a major depressive episode is not a direct answer given by the respondent, but rather a condition identified based on a conditioned series of simultaneous symptoms. However, some potential biases remain: first, though the MINI tool is an international standard [15,16], it has been developed in Western countries and may not be sensitive to all presentations of major depressive episodes in Ouagadougou; second, some of the questions asked may seem unusual or intrusive to respondents which may encourage them to answer in the negative; third, it is always possible that interviewers may ask prompt questions about overall mood deterioration and loss of interest in usual activities in such a way that the respondent will answer negatively allowing them to avoid the filtered questions and finish the interview more quickly. Taking these factors into account, the estimated proportion of current major depressive episode found here could be underestimated and is perhaps best considered as a minimum prevalence estimate. Additionally, results from multivariate analysis are based on regression models, the associations highlighted by our analysis depend on the specific set of variables introduced in the model and we have potential unmeasured confounders. Due to small frequencies, we couldn't assess the existence of a clustering effect of the household level, which could be present as in some cases multiple members of the same household were interviewed. For the same reason, we did not run separate models for each sex, as certain associations are probably stronger for males , while some are for females . Finally, the significant associations found in this study cannot necessarily be interpreted as causal factors for a major depressive disorder. Further research, more specifically devoted to mental health with a larger sample size, is required to confirm these first results. --- Conclusions Our study found that 4.3 % of residents in Ouagadougou were currently suffering from a major depressive episode at the time of the survey, with an overall risk of a lifetime major depressive disorder among this population likely higher. We found that experiencing a major depressive episode in Ouagadougou was associated with a reported chronic disease, disability, household food shortages, having experienced physical violence, and drinking regularly alcohol. Migrant status, neighborhood type , age group, marital status, and gender were not significantly associated with major depressive episodes. A surprising finding from this study showed that, even if the poorest were most susceptible to major depressive disorder, the wealthiest segments of society also tend to be more vulnerable to depression in this setting. Although urban populations in developing countries, like those living in Ouagadougou, still face a heavy burden of disease in the form of infectious diseases, chronic diseases and accidents [34], many individuals suffering from these health problems often also experience from major depressive disorder. This may be particularly true for chronic diseases, which showed a strong association with experiencing a major depressive disorder. As urban populations throughout sub-Saharan Africa continue to grow and experience an upward shift in their age structures, it is likely that an increasing proportion of urban dwellers will experience a major depressive episode in their lifetime, with detrimental effects for their well-being. It is thus critical that, following recommendation from the WHO [46], efforts to improve overall health in urban populations strive to incorporate mental health diagnosis and treatment in a more systematic way within established health systems. --- Endnotes 1 This study included both literate and illiterate respondents. Illiterate respondents were asked to provide a literate witness and the informed consent form was read to both the respondent and the witness. If the respondent agreed to participate, he/she was asked for a finger print on the consent form, and the witness was asked to provide his/her signature on the form as well. The proportion of illiterate respondents here is believed to be similar to that of Ouagadougou as a whole based on a data comparison with the 2010 Burkina Faso Demographic and Health Survey. 2 The Burkinabe Institute for Literacy was in charge of the translation of the questionnaire. To ensure accurate translation, a backward translation from Moore to French was also done. contributed to the theoretical background and helped draft the manuscript. All authors read, contributed to revising the manuscript, and approved the final manuscript. --- Competing interests The authors declare that they have no competing interests. ---

Background: In sub-Saharan African cities, the epidemiological transition has shifted a greater proportion of the burden of non-communicable diseases, including mental and behavioral disorder, to the adult population. The burden of major depressive disorder and its social risk factors in the urban sub-Saharan African population are not well understood and estimates vary widely. We conducted a study in Ouagadougou, Burkina Faso, in order to estimate the prevalence of major depressive episodes among adults in this urban setting. Methods: The Ouagadougou Health and Demographic System Site (HDSS) has followed the inhabitants of five outlying neighborhoods of the city since 2008. In 2010, a representative sample of 2,187 adults (aged 15 and over) from the Ouaga HDSS was interviewed in depth regarding their physical and mental health. Using criteria from the Mini International Neuropsychiatric Interview (MINI), we identified the prevalence of a major depressive episode at the time of the interview among respondents and analyzed its association with demographic, socioeconomic, and health characteristics through a multivariate analysis. Results: Major depressive episode prevalence was 4.3 % (95 % CI: 3.1-5.5 %) among the survey respondents. We found a strong association between major depressive episode and reported chronic health problems, functional limitations, ethnicity and religion, household food shortages, having been recently a victim of physical violence and regularly drinking alcohol. Results show a U-shaped association of the relationship between major depressive episode and standard of living, with individuals in both the poorest and richest groups most likely to suffer from major depressive disorder than those in the middle. Though, the poorest group remains the most vulnerable one, even when controlling by health characteristics. Conclusions: Major depressive disorder is a reality for many urban residents in Burkina Faso and likely urbanites throughout sub-Saharan Africa. Countries in the region should incorporate aspects of mental health prevention and treatment as part of overall approaches to improving health among the region's growing urban populations.

Introduction The associations between race [1], socioeconomic status [2][3][4], and health behaviors are very complex [5][6][7]. First, racial minorities and low SES children show high-risk behaviors and poor health statuses [8,9]. Second, race and SES closely overlap [10], with most racial and ethnic minorities having lower SESs [10]. As a result, some of the black-white differences in high-risk behaviors are not an effect of race, but the existing SES gap across racial groups. That is, a low SES may partially explain why race or ethnicity is linked to behavior and health [11]. Fourth, some evidence suggests that SES effects may depend on race or ethnicity, with the effects having a tendency to be weaker for black than white families [12]. This pattern is commonly known as Minorities' Diminished Returns [13,14]. According to the MDRs framework [13,14], SES indicators, such as family income, show weaker effects on the behavior and health of black [15], Hispanic [12,16], Asian American [17], Native American [18], and even marginalized white [19] families. In this view, racial and ethnic minority groups face additional difficulties; mistreatment and injustice will stay a risk despite a high SES. As a result of MDRs, we observe higher levels of anxiety [20], depression [21], and suicide [22] in high-SES black children than what is seen in high-SES white children. While there is an extensive body of research documenting MDRs of parental education [23], family income [24,25], and marital status [26] on a wide range of developmental, health, emotional, and behavioral outcomes [21,[23][24][25], we are not aware of any previous studies that test MDRs of a family's SES on children's screen time. We know that a high SES shows weaker effects on depression [21], anxiety [20], impulsivity [24], grade point average [23], attention deficit hyperactivity disorder [27], diet [28], exercise [29], sleep [30], substance use [23], obesity [31], and chronic disease [23] for black children than white children. However, there is more to be learned about the relevance of MDRs of the SES indicators to screen time for black children. --- Aims Given the existing gap in the contribution of MDRs in explaining the black-white gap in screen time, we conducted the current study with two aims. The first was to estimate the overall link between income and screen time. The second aim was to compare racial differences in terms of income and screen time, especially between black and white families. We hypothesized a protective effect of family income on children's screen time . We also expected this association to be weaker for black children than white children . --- Methods --- Design and Settings This secondary analysis borrowed data from the Adolescent Brain Cognitive Development study [32][33][34][35][36]. ABCD baseline data collection was conducted in 2016 at 21 sites across various states in the United States. For more information on the ABCD study, consult [32,37]. --- Participants and Sampling The ABCD participants were 9-11-year-old children who were selected from multiple cities across various states. ABCD recruitment primarily relied on the US school system. For a detailed description of the sampling and recruitment for the ABCD study, consult [38]. The eligibility of our analysis was having valid data on all our study variables, including race, family income, and screen time. The analytical sample of this paper was 15,022. --- Study Variables The study variables included race, ethnicity, sex, family income, household income, parental education, marital status, and screen time. Race was self-identified as black, Asian, mixed or other, and white . The mixed or other race reflects Native American and individuals with mixed or other races. The sex of the child was 1 for male and 0 for female. Parental marital status was reported by the parents and was 1 for married and 0 for other. Household income, reported by the parent, was a three-level categorical measure of less than 50 K , 50-100 K, and 100+ K. Screen time was reported by the parents. This measure was for average weekdays and measured in hours. This variable was treated as a continuous measure, with a higher value indicating more screen time. Parents' reports of children's screen time were believed to be reliable and valid [39]. --- Data Analysis We used the Data Exploration and Analysis Portal for data analysis. The DEAP uses the R package for statistical calculations. We reported the mean ) and frequency of our variables both overall and by race. We also performed the chi-square and analysis of variance tests for our bivariate analysis. We used two mixed-effects regression models for multivariable modeling that allowed us to adjust to our data's nested nature. Observations were nested to participants from families who were selected across the sites and states. We ruled out collinearity between our independent variables in our models. We also confirmed the near-to-normal distribution of error terms of our regression model. Both models were performed in the overall sample. Model 1 did not have any interaction terms between race and family income. Model 2 included interaction terms between race and family income. In both models, children's screen time was the outcome . The regression coefficient , SE, 95% CI, and p-value were reported. Statistical power was calculated based on the number of variables in the regression model, alpha, and percentage of variance explained. According to Cohen, a sample size of 15,000 will provide a power of 99% for a regression model with eight terms, which explains 0.005 of the variance and a p-value of 0.05 [40,41]. --- Ethical Aspect The ABCD study has institutional review board approval, and all participants have provided assent or consent, depending on their family income [37]. Given that our analysis was performed on fully de-identified data, our analysis was exempt from a full IRB review. --- Results --- Descriptives Overall, 15,022 observations from 9-11-year-old children were analyzed. Most participants were whites, followed by blacks. From all observations, 10,249 were for white children, 1961 were for black children, 339 were for Asian children, and 2473 were for other or mixed-race children. On average, children had 3.55 hours on average of weekday screen time. Table 2 presents descriptive data by race. This table also compares racial groups by the study variables. As this table shows, black and mixed or other race participants had the lowest parental educations and incomes, and white and Asian children had the highest parental educations and household incomes. The average weekday screen time was highest for black children and lowest for white and Asian American children . --- Multivariate Models Table 3 presents the results of three mixed-effects regression models in the overall sample. Model 1 shows the association between family income and race with children's screen time. Compared with an income less than 50 K, an income between 50 and 100 K was associated with a significant decline in children's average weekdays' screen times . Compared with an income less than 50 K, an income of more than 100 K was associated with a larger decline in children's average weekdays' screen times . The Asian American race was also associated with lower average weekdays' screen times. At the same time, black and mixed or other races were associated with a higher average screen time for children. Model 2 showed an interaction between family income and race with children's screen time. This interaction indicated that the boosting effect of family income and children's screen time was weaker for black children than white children. Figure 2 shows the interaction between family income and race with children's screen time. This interaction indicated that the boosting effect of family income on children's screen time was weaker in black families than white families. --- Discussion This study showed an inverse association between family income and children's screen time overall. However, this association was stronger for white children than black children. That is to say, while a high family income accompanied less screen time for American children, this association was weaker in black families than white families. As a result, high-income black children had high screen times, a pattern absent for high-income white children. Our findings reflected the impact of the MDRs of family income on children's screen time in black families. This is in line with what is already established for the effects of MDRs of many economic resources, such as family income, on impulsivity [24], reward responsiveness [42], inhibitory control [43], attention [44], and ADHD [27] for black children compared with white children. Similar MDRs were also reported for the effects of family SES indicators, such as parental Figure 2 shows the interaction between family income and race with children's screen time. This interaction indicated that the boosting effect of family income on children's screen time was weaker in black families than white families. Figure 2 shows the interaction between family income and race with children's screen time. This interaction indicated that the boosting effect of family income on children's screen time was weaker in black families than white families. --- Discussion This study showed an inverse association between family income and children's screen time overall. However, this association was stronger for white children than black children. That is to say, while a high family income accompanied less screen time for American children, this association was weaker in black families than white families. As a result, high-income black children had high screen times, a pattern absent for high-income white children. Our findings reflected the impact of the MDRs of family income on children's screen time in --- Discussion This study showed an inverse association between family income and children's screen time overall. However, this association was stronger for white children than black children. That is to say, while a high family income accompanied less screen time for American children, this association was weaker in black families than white families. As a result, high-income black children had high screen times, a pattern absent for high-income white children. Our findings reflected the impact of the MDRs of family income on children's screen time in black families. This is in line with what is already established for the effects of MDRs of many economic resources, such as family income, on impulsivity [24], reward responsiveness [42], inhibitory control [43], attention [44], and ADHD [27] for black children compared with white children. Similar MDRs were also reported for the effects of family SES indicators, such as parental education, household income, and marital status, on behavioral risks such as substance use [23], diet [28], exercise [29], anxiety [20], depression [21], and suicide [22]. These are all diminishing returns of economic resources for black youths when compared with white youths [16,[45][46][47]. These MDRs are not specific to any particular domain or outcome, suggesting that they are due to society, but not culture, behavior, or biology. Thus, high-income black children do not have high screen times because they are innately different or lazy. Similarly, the diminished slope of income with screen time is not because blacks have less interest in or undervalue physical activity, since similar MDRs are shown for all marginalized groups across marginalized identities [13,14]. MDRs are not specific to blacks [25], but are also evident in Hispanics [12,16,48,49], Asian Americans [17], Native Americans [18], and even marginalized whites [19]. They are also not specific to a particular SES resource, behavioral outcome, or age group. These patterns are present for children [24,25,50], adults [46], and older adults [51]. This paper extends the previous work on the MDRs of family income and screen time in black children. Due to sociological and structural factors such as segregation, labor market discrimination, and racism, black families live under tremendous stress regardless of their SES [52][53][54]. High-SES black families generate less income and wealth and live in worse neighborhoods than whites with similar SESs [55][56][57]. Black workers with the very same educational attainment work in worse conditions and receive worse income than whites with similar educations [55]. A wide range of economic and non-economic mechanisms can explain the MDRs of family income and economic resources and their greater effect on children's screen time in black children than white families. Black families experience high levels of stress across all SES levels [58]. Social mobility has been more taxing for high-SES black families than white families [59]. Exposure [60,61] and vulnerability [62] to discrimination are high for high-SES black families. While low-SES black families struggle with food insecurity, poverty, and neighborhood disorders, high-SES black families experience discrimination due to their proximity to whites [60]. As discrimination reduces the chance of healthy brain development [61,62], high-SES black children may remain at risk of impulsivity [24]. As a result, family income, one of the main drivers of children's brain development, shows weaker effects for black families than white families. While low SESs and poor outcomes are one type of disadvantage in black communities, MDRs reflect the second type of disadvantage [13,14]. These two sources of disadvantages are qualitatively different. While the former reflects unequal outcomes and opportunities, the latter reflects the low response to the presence of individual-level resources such as family income and SES. It is due to the latter that policymakers may observe sustained inequality despite investments. To address the latter, there is a need to address the systemic causes of inequalities. As a result of the combination of these two, black families experience double jeopardy. Not only are resources such as SES scarce, their influences are also hindered and dampened due to the multitude of constraints in their environment [13,63]. Multi-level economic and environmental mechanisms reduce the marginal returns of family SES, such as income, for black families [13,63]. MDRs are attributed to social stratification, segregation, discrimination, racism, and marginalization. These processes function across multiple societal institutions [13,63]. Racial injustice, prejudice, and discrimination have historically interfered with the gain of resources and assets in black communities [64][65][66]. One of the many causes of MDRs may be childhood poverty [57]. As a result of historical and structural injustice, MDRs hinder the effects of existing resources and assets across settings and social groups. --- Implications High screen time has a negative influence on child brain development. Using ABCD data, we found that high-SES black children have high screen times regardless of their SES. Our finding is in line with other studies that have documented diminished effects of high SES on the brain development of black children, when compared to white children. In a study using ABCD data, high-SES black children remained at risk of depression and suicide, while high-SES white children were at low risk of depression and suicide [22]. In other ABCD studies, high-SES black children remained at risk of reward dependence. However, white high-SES children had low reward dependencies [42,67]. In other studies, high-SES black children remained at risk of social, emotional, and behavioral problems. However, this was not the case for white children from high-SES families [68]. In another analysis of the ABCD study, a high subjective SES was independently associated with the amygdala size in larger children. However, the boosting effect of a high SES on total amygdala volume was weaker for black children than white children. As such, high-SES black children remain at risk regarding amygdala size, regardless of their SES, which is due to the diminished return of SES on all aspects of the development of children due to racism [69]. --- Limitations The current study has some methodological limitations. This study does not suggest causal inferences. However, it is more likely that poverty impacts sedentary behavior than the children's screen time shapes the household SES. Still, the findings reported here are correlations, not causes. To establish causal evidence, we need to have repeated measurements of SES and screen time and test if changes in family income precede a change in children's screen times. We only tested the MDRs of family income. Previous work had established MDRs of various family SES indicators, such as parental education [42,67]. Future research should test if similar MDRs also apply for other behaviors and time spent on other activities. Additionally, we only controlled for family SES, and all our confounders were family-and individual-level SES indicators. It is imperative to control for contextual and neighborhood-level indicators, as well as health. Finally, we did not study how these MDRs changed over time, how they emerged, and how family dynamics and parenting contributed to them. We also do not know how parents' screen times and family norms, expectations, and rules shaped such MDRs. More research is needed on whether these black-white gaps narrow, maintain, or widen over time and whether any factor can undo them. Finally, it is unknown whether MDRs related to inequalities in children's screen time contribute to the higher-than-expected risk of chronic disease [70][71][72][73] in high-SES black people. --- Conclusions Relative to white children, black children show weaker family income effects on screen time, meaning that screen time stays high in black children across all family income groups. We did not observe the same pattern for high-income white families. The diminishing returns of family income on lifestyle are important because screen time is a driver of obesity, depression, and several other health problems. To minimize the black-white health and behavioral gap, it is imperative to address societal and structural barriers beyond SES. These include racism, stratification, injustice, and hopelessness in black communities. Such factors may reduce the health return of family SES for black populations. There is a need to find public and social policies that minimize MDRs. Such policies may need to address the multi-level mechanism by which racism impacts the lives of black American families. Systemic, structural, and societal causes of inequalities should be discussed rather than ignored. Funding: Shervin Assari was supported by the National Institutes of Health grants 5S21MD000103, D084526-03, CA201415 02, DA035811-05, U54MD008149, U54MD007598, and U54CA229974. The ABCD study was supported by the National Institutes of Health and additional federal partners under award numbers U01DA041022, U01DA041028, U01DA041048, U01DA041089, U01DA041106, U01DA041117, U01DA041120, U01DA041134, U01DA041148, U01DA041156, U01DA041174, U24DA041123, U24DA041147, U01DA041093, and U01DA041025. A full list of supporters is available at https://abcdstudy.org/federal-partners.html. A listing of participating sites and a complete listing of the study investigators can be found at https://abcdstudy.org/Consortium_Members.pdf. ABCD consortium investigators designed and implemented the study or provided data, but did not necessarily participate in the analysis or writing of this report. This manuscript reflects the views of the authors and may not reflect the opinions or views of the NIH or ABCD consortium investigators. The ABCD data repository grows and changes over time. The current paper used Curated Annual Release 2.0, also defined in NDA Study 634 . DEAP is software provided by the Data Analysis and Informatics Center of ABCD located at UC San Diego, with generous support from the National Institutes of Health and the Centers for Disease Control and Prevention under award number U24DA041123. The DEAP project information and links to its source code are available under the resource identifier RRID: SCR_016158.

While increased household income is associated with overall decreased screen time for children, less is known about the effect of racial variation on this association. According to Minorities' Diminished Returns (MDRs) theory, family income and other economic resources show weaker association with children's developmental, behavioral, and health outcomes for racialized groups such as black families, due to the effect of racism and social stratification. In this study, we investigated the association, by race, between family income and children's screen time, as a proxy of screen time. This longitudinal study followed 15,022 American children aged 9-11 over a 1-year period. The data came from the baseline of the Adolescent Brain Cognitive Development (ABCD) study. The independent variable was family income, and it was categorized as a three-level nominal variable. The dependent variable, screen time, was a continuous variable. Ethnicity, gender, parental education, and marital status were the covariates. The results showed that family income was inversely associated with children's screen time. However, there was a weaker inverse association seen in black families when compared with white families. This was documented by a significant statistical interaction between race and family income on children's screen time. Diminished association between family income and children's screen time for black families, compared with white families, is similar to MDRs and reflects a health risk to high-income black children. In a society where race and skin color determine opportunities and treatment by society, children from middle class black families remain at risk across multiple domains. We should not assume that income similarly promotes the health of all racial and ethnic groups. Addressing health and behavioral inequalities requires interventions that go beyond equalizing socioeconomic resources for black families. Marginalization, racism, and poverty interfere with the normal family income-related development of American children.

INTRODUCTION Indonesia has one of the largest Muslim populations in the world, but its Islamic financial assets are only USD99 billion compared to Malaysia whose assets have already grown to USD570 billion as at 2019 . According to Yumna , Islamic banking assets have been increasing constantly in the last few years. However, one of the challenges of developing the Indonesian Sharīʿah economy is the low level of Islamic financial literacy in the community . Low IFL may affect financial behaviour, such as savings, credit management, investments and insurance. A survey by Otoritas Jasa Keuangan revealed that the percentage of Indonesia's population who are Islamically financially literate was only 8.93 per cent in 2019 . This percentage is far below conventional financial literacy which stood at 37.72 per cent . This factor influences the index of public access to Islamic financial products and services . According to the data, the percentage of public access to products and services of Islamic financial institutions was only 9 per cent . Hence, it is a plausible reason why the Islamic finance industry's market share in 2019 was only 8.69 per cent of the total financial industry in Indonesia . One means of improving the Islamic economy market share in Indonesia is by educating people in positive financial behaviour from an Islamic finance perspective. It starts with increasing the financial literacy of the community, especially the younger generation . The younger generation is generally technology literate. On the one hand, they are potential Islamic economic actors in the retail banking sector and are the key factor for the Islamic economy's growth in Indonesia . On the other hand, the younger generation has insufficient financial knowledge to improve their future financial prosperity; hence, they require an increase in IFL . Generation Z is a young generation who was born between 1995 and 2001 . Based on available data, Gen Z's population in Indonesia is 75.49 million, representing 27.94 per cent of the total population . In addition, in the research context of Indonesia they have a strong commitment to religion . Therefore, to achieve the objective of increasing the Sharīʿah economy, it is important that Gen Z partakes in Islamic financial activities as they are generally good in technology and strongly committed to religion. The financial literacy of Gen Z is influenced by several factors . Previous studies highlighted the factors that influence financial literacy and financial behaviour, such as communication patterns in the family , the role of financial literacy on the customer's tolerance of risk , the influence of financial literacy on financial behaviour and investment behaviour . For example, Zhu argues that direct parental teaching influences financial behaviour, and Grohmann asserts that financial literacy leads to financial behaviour. However, the studies have not addressed these variables in a more complex research model. In addition, these studies focused on conventional financial literacy rather than IFL. Hence, studies in the IFL context were neglected. Besides, financial literacy for Muslim customers is an emerging topic, especially in Muslim-majority countries such as Indonesia . Therefore, this study examines the effect of family communication patterns on IFL and risk tolerance, the role of IFL on risk tolerance, and the effect of IFL and risk tolerance on financial behaviour. It also examines the moderating role of IFL on the link between risk tolerance and financial behaviour, which is an aspect that has been neglected in previous research . According to Alshater et al. , one of the future research streams is the attitude and perception of customers towards Islamic finance. They argue that the role of IFL in this regard needs to be further explored in future research. This paper is accordingly organised as follows: after this introduction, the second section discusses the literature review on key themes related to the topic. The third section delineates the methodology that has been applied in this research. It is followed by the presentation and discussion of the results. The last section concludes the paper, laying out its theoretical contribution, practical implications, research limitations and recommendations for future research. --- LITERATURE REVIEW Financial Behaviour Financial behaviour refers to an individual's behaviour in carrying out financial activities such as investing in stocks and securities . Rey-Ares et al. recognise the following financial behaviours of the individual: investment behaviour and indebtedness . Dew and Xiao similarly classify financial behaviour into savings and investment, but also added other aspects of individuals' behaviour including management of cash, credit and insurance. Financial behaviour is influenced by factors such as individual's financial literacy, demographics and financial risk tolerance . --- Islamic Financial Literacy In the study of financial literacy, some researchers have examined IFL, which focuses on the financial literacy of Muslim communities . IFL is still a new concept. Therefore, only a few studies have focused on a scale or measurement tools for IFL . According to Nawi et al. , IFL can be measured by using four dimensions, including money basics, Islamic banking, Islamic investment and takāful. Money in Islam serves as a medium of transaction. It performs a social role, benefitting the society at large. Earning interest on the lending of money is not allowed in Islam, and Islamic banking is instead founded on interest-free practices. Moreover, Islamic banks only finance Sharīʿah-compliant businesses. Islamic investments are based on the concepts of risk sharing and no guaranteed profit. Gharar and maysir are also prohibited elements in Islamic investments. Takāful is an insurance concept based on the mutual intention to protect one another against financial mishaps. It is based on tabarruʿ among participants. Previous studies have highlighted the effects of financial literacy on financial behaviour . However, there is no investigation into the IFL-financial behaviour relationship yet. Furthermore, financial literacy is influenced by several factors such as family communication patterns . Financial literacy affects peoples' risk tolerance and their financial behaviour in the future . According to Abdullah et al. , the factors influencing IFL are gender, education level, and attitude towards financial management. Ali et al. considered two dimensions of Islamic financial inclusion determinants:  On the demand side, there are four influencing factors, including financial literacy, religious commitment, and social and economic influence.  On the supply side, there are four influencing factors, including human capital, products and services, infrastructure readiness, and policies and regulation. IFL in this study is taken from the research by Nawi et al. . They argue that IFL is 'the knowledge gained through education and/or specific experience relating to important concepts and products of Islamic finance' . This study has eight hypotheses based on the relationship between variables which are discussed below. --- Family Communication Patterns and Risk Tolerance Previous studies have stated that people's risk tolerance is influenced by various factors such as demographics, socio-demography and individual and environmental factors . Rabbani et al. also found that personality factors influence risk tolerance. Furthermore, Hanson and Olson highlighted that communication patterns will affect attitudes towards risk and trust. One of the attitudes towards risk is the individual's tolerance for possible risk. Better communication patterns in the family are expected to provide better selfconfidence for individuals. Hence, it is argued that they can be more tolerant towards risks. Although research on this relationship is still limited, it is believed that communication patterns within the family might significantly influence an individual's tolerance for risk. Hanson and Olson stated that family communication patterns can be divided into two types: conformity orientation and conversation/discussion orientation. Conformity orientation refers to one-way communication methods where the communication flow is topdown from parents to children. Conversation/discussion orientation, on the other hand, is a type of communication where parents and children play their roles equally and they use two-way communication methods in their conversations. Presumably, children prefer conversation orientation and it leads to their tolerance in taking risks, whilst the conformity patterns will affect children's risk tolerance negatively . Based on this discussion, therefore, this study proposes the following hypotheses: H1: Conformity communication patterns have a negative and significant impact on risk tolerance. H2: Conversation communication patterns have a positive and significant impact on risk tolerance. --- Family Communication Patterns and Islamic Financial Literacy Communication patterns in the family might greatly influence socialisation in the family and thus affect financial literacy, including IFL . Discussion-oriented family communication patterns, as described by Hanson and Olson , will be more suitable for families to socialise, allowing the younger generation to learn through discussions than through one-way or conformity communication patterns. Hence, discussion-oriented family communication patterns might increase the younger generation's level of financial literacy . Previous studies have only focused on financial literacy in general and only limited studies have addressed IFL . Meanwhile, the number of Muslim members of Gen Z in Indonesia is high. They might potentially benefit from a higher level of IFL. Therefore, this study examines the impact of family communication patterns on IFL and proposes the following hypotheses: H3: Conformity family communication patterns have a significant impact on IFL. H4: Conversation family communication patterns also have a significant link to IFL. --- Islamic Financial Literacy and Risk Tolerance According to Zhu , a person's financial knowledge might increase their tolerance to financial risks. This level of financial knowledge can be real or perceptual . People who understand or feel that they understand the risks they are facing become more courageous in taking risks. Risk tolerance is an important factor of IFL, which may have a significant impact on people's behaviour in the future. Hermansson and Jonsson highlighted the significant impact of financial literacy on risk tolerance. Similarly, Hsiao and Tsai found that individuals with high financial knowledge have a higher preference for investing in the capital market than those with low levels of financial knowledge. They will invest in more risky products because they have calculated the risk due to their sufficient knowledge. This is applicable not only in the case of general financial literacy but also for IFL. Therefore, this study hypothesises that: H5: IFL is an antecedent of risk tolerance. --- IFL, Risk Tolerance and Financial Behaviour According to Cameron et al. and Kannadhasan , a person's tolerance of risk affects their financial behaviour such as cash management, savings, insurance, investment allocation and loans. People with high risk tolerance will presumably have more complex investments with higher risk . People with a higher level of risk tolerance and who have been supported by good financial literacy will understand various alternative investment instruments in terms of risk and return. Hence, they can choose from the different financial instruments available to build a more diverse and complex investment portfolio. Hermansson and Jonsson have also asserted that financial literacy is an antecedent of financial risk tolerance. Accordingly, financial literacy might strengthen the financial risk | 26 ISRA International Journal of Islamic Finance • Volume 15 • Number 2 • 2023 tolerance-financial behaviour link. This research hypothesises the following about IFL, which is a specific type of financial literacy: H6: IFL significantly moderates the risk tolerance-financial behaviour relationship. --- Islamic Financial Literacy and Financial Behaviour According to Antara et al. , the level of IFL can be measured using four indicators, namely basic financial knowledge , loans/credit, investment/savings, and insurance/protection of resources. Previous studies have addressed the financial literacy impact on financial behaviour . However, they have not addressed the IFL-financial behaviour link yet. Knowledge about Islamic finance, such as the prohibition of interest, will affect people's investment choices which are related to financial behaviour. The above literature review supports the formulation of the following hypothesis: H7: IFL is an antecedent of financial behaviour. --- Risk Tolerance and Financial Behaviour Several previous studies stated that financial behaviour is also determined by the level of risk tolerance . When people have a higher risk tolerance level, they will have more courageous investment decision-making. Accordingly, this study argues that: H8: Risk tolerance is a significant antecedent of financial behaviour. --- METHODOLOGY Population and Sample This study used Muslim members of Gen Z in West Sumatra, Indonesia, who were at least 17 years old as the research population. The Generation Z in this study is a person who was born from 1995 to 2001 . This study used G* Power in calculating the minimum sample. The minimum sample using G* Power with effect size = 0.2, alpha = 0.05 and power 0.99 was 443 respondents. This study analysed 571 respondents; therefore, it has met the minimum criterion. --- Data Collection Procedures The sampling method of this study is the convenience sampling method. The respondents of this study are members of Gen Z who have social media accounts and agreed to fill in the questionnaire via Google Forms. The Likert scale with five levels from 'Strongly Agree' to 'Strongly Disagree' was used as a measurement tool in the questionnaire. This study employed two stages of data collection. The first stage was pilot or preliminary data collection to 30 respondents. The pilot survey data were analysed to test the items' validity and reliability. The second stage was primary data collection. The data collection was conducted for two months over June-July 2020. The data have been collected using an online survey through Google Forms. The questionnaires were distributed to several social media platforms such as Facebook and Instagram and by email. To increase the response rate, the questionnaires were sent four times and gentle reminders were also sent to the respondents. Finally, after two months of data | 27 ISRA International Journal of Islamic Finance • Volume 15 • Number 2 • 2023 collection, this study administered 668 responses and after the preliminary checks, including missing value and non-response biases, only 571 responses were used for the analysis. --- Variables and Measurements This study consisted of four variables, including financial behaviour, IFL, risk tolerance and family communication patterns. The measurement of financial behaviour was adapted from Dew and Xiao . The 12 items of financial behaviour have been measured using scales 1 = never, 2 = seldom, 3 = sometimes, 4 = often, 5 = always. IFL measurements were adopted from Nawi et al. . This variable used 22 items, including money basics , Islamic banking , Islamic investment and takāful . These items have been measured using scales 'True, False and Not Sure'. The value of this variable was measured using the number of correct answers from the respondents. Furthermore, risk tolerance measurement item was taken from Bannier and Neubert . The respondents were asked about their willingness to take risks. It has been measured using scales from '0' to '10', where 10 is the highest willingness to take risks. Finally, the last variable was family communication patterns. The measurement items of this variable were adopted from Hanson and Olson . They classified family communication patterns into two dimensions, including conformity family communication patterns and conversation family communication patterns . These items have been measured using a 5-point Likert scale from 'Strongly disagree = 1' to 'Strongly agree = 5'. Table 1 and Table 2 show the details of the variable measurements. The Sharīʿah prohibits the element of maysir in investments. X 4 Otoritas Jasa Keuangan and Dewan Pengawas Syariah are responsible to monitor and approve Sharīʿah-compliant investment products. X 5 Muḍārabah is an example of a Sharīʿah-compliant investment contract. X Takāful 1 The takāful contract is based on the principle of tabarruʿ. X 2 In takāful, the participants guarantee each other. X 3 Takāful consists of the element of gharar just as in the case of conventional insurance. --- X 4 The contract of takāful insurance is based on the sale contract. X 5 There is no insurer-insured relationship between the takāful operator and participants. --- Data Analysis This study employed two data analysis methods, including descriptive and Structural Equation Modeling using Partial Least Square . The descriptive analysis aims to provide general information regarding the minimum, maximum, and average and standard deviation values. Before the main analysis, this study conducted several preliminary tests such as normality with kurtosis ±3, heteroscedasticity by using the Glejser test, and multicollinearity test with cut-off point Variance Inflation Factor value < 10 . This study found the kurtosis values for all variables to be between 0.77 and 2.20. Based on the Glejser test, the significant value of all variables is more than 0.05. VIF values of all variables are between 1.04 and 1.10. Therefore, all variables have met the criteria of preliminary analysis. Furthermore, it was a cross-sectional study; hence, it will face a common method bias problem. To anticipate this problem, Harman's single-factor test using Exploratory Factor Analysis was employed. It found the total variance explained for single-factor is 21.77 per cent, which was less than the cut-off point of 50 per cent . Moreover, this study employed SEM-PLS for the analysis of data. This analysis has two models, notably the measurement model and the structural model . For the measurement model, validity and reliability tests have been conducted. This study used Convergent Validity by using loading factor and Average Variance Extracted value . The loading factor should be ≥ 0.6 and AVE should be greater than 0.5. It has also tested Fornell and Larcker's criterion for the discriminant validity, where the value of the square root of AVE in the diagonal should be greater than the correlation value between variables. For the reliability test, this study has calculated the Composite Reliability and Cronbach's Alpha. The value should be ≥ 0.7. Table 3 andTable 4 show that all constructs have good validity and reliability. Bootstrapping with 5000 re-sampling was also used to solve the normality problem . After the validity and reliability tests, several items were deleted and Table 3 shows that only 48 items have been included in the analysis. My parents make me enjoy when we discuss about something, despite the fact that we have different ideas. --- 0.830 My parents like to discuss about my ideas, even when they have different opinions. --- 0.838 My parents let me discuss my ideas with them. 0.842 We often discuss about our family problems together. 0.799 In our family, we often discuss about our plans and hopes freely. --- 0.789 My parents often say every family member should be involved in family decision-making. --- 0.650 My parents often discuss with me about our financial conditions. --- 0.652 My parents let me give my point of view on their ideas. 0.688 I usually discuss about my financial situation with my parents. 0.644 I can discuss anything with my parents. 0.730 In our family we are free to discuss our feelings and opinions. --- 0.771 My parents and I often have time to discuss everything. 0.749 IFL Correct answers from 22 items 1.000 1.000 1.000 Risk Tolerance I do not mind taking risks with respect to financial matters. 1.000 1.000 1.000 --- RESULTS AND DISCUSSION --- Results --- Respondents Based on the results, most of the respondents were women , and they have graduated from senior high school . Most of them were born in 1999 . Their personal income was less than Rupiah 2 million . The largest category of parents' occupation was entrepreneurship , and 44 per cent of parents are senior high school graduates. --- Hypothesis Testing This study tested the hypothesis by using SEM-PLS with SMART-PLS 3.2.6. Table 5 shows the results of the hypothesis testing. This study highlighted that parent communication patterns have a significant impact on risk tolerance, including conformity and conversation . Parent communication patterns also have a significant impact on IFL, including conformity and conversation . This study found that IFL has a significant impact on risk tolerance . IFL is a moderating variable of the risk tolerance-financial behaviour relationship . IFL has a significant impact on financial behaviour . Finally, this study also found that financial behaviour will be significantly influenced by risk tolerance . --- Discussion Communication patterns in the family have a significant impact on risk tolerance as follows:  Conformity patterns have a significant and negative impact on risk tolerance.  Conversation patterns have a positive relationship with risk tolerance. These findings support previous research . According to León and Pfeifer , socio-demographic factors such as communication in the family will lead to the risk tolerance of family members. Therefore, families who have good communication patterns-such as two-way communication and discussion-will give confidence to children, which will lead to higher levels of risk tolerance. On the other hand, families with conformity patterns-where parents give top-down orders to their children-will affect their children's risk tolerance in a negative way. Communication patterns within the family are also a significant antecedent of IFL . These findings also support previous studies . Grohmann et al. highlighted that communication in a family relates to child financial literacy. Conversation or discussion-oriented communication patterns will encourage children to think critically and help to establish a learning process so that children will have sufficient knowledge of IFL. Members of Generation Z will be more inclined to discuss than to communicate with a doctrinal pattern. They will accept opinions or concepts which are logical and understandable. IFL also significantly affects risk tolerance . This finding is similar with some previous studies, such as Zhu who assert that people who are more knowledgeable in finance will have a greater tolerance of financial risk. They can calculate the financial risk in more detail, and this will affect their tolerance level of the financial risk because they have more confidence in making a decision. This study also found that IFL is a significant moderating variable on the link between risk tolerance and financial behaviour . Even though previous studies have not addressed this link yet, Cameron et al. have highlighted that higher risk tolerance will affect people's financial behaviour. Accordingly, when people have a good level of financial literacy, it will strengthen the relationship between financial risk and financial behaviour because they can calculate potential risks and have a higher confidence in making a decision. IFL has a moderating role in strengthening the link between risk tolerance and financial behaviour. This finding is congruent with several previous studies . Consumers who have good financial knowledge will have a better financial behaviour such as making the right consumption patterns and being able to make the right investment choices. IFL is a significant influencing factor of financial behaviour . This finding is in line with some previous studies . For example, García and Vila found that financial literacy has a significant impact on financial behaviour. Even though previous studies were not in the Indonesian context, financial behaviour was also found to be significantly affected by risk tolerance . This finding supports the findings of previous studies. León and Pfeifer argue that tolerance the Islamic context and especially for Generation Z . 3. It also extends prior studies by addressing the moderating role of Islamic finance literacy on the risk tolerance-financial behaviour relationship. Financial literacy is already known as the antecedent of financial behaviour . Financial literacy also relates to risk tolerance . To the best of the authors' knowledge, there was no previous study that examined the moderating effect of IFL on the risk tolerance-financial behaviour relationship. This study is also expected to help decision makers such as OJK, the government, and management of Islamic banks and IFIs to improve Gen Z financial behaviour by considering their IFL, risk tolerance and family communication patterns. Hence, the decision makers might educate not only the Muslim Gen Z about IFL but also their parents. Socialisation, promotion and training to Gen Z might be considered by the decision makers to increase the level of IFL. Increasing the understanding of IFL will in turn improve their risk tolerance. They will be aware of financial risks which will impact on their future financial behaviour. Moreover, the government might also give some incentives to Islamic banks or the central bank to educate Gen Z about IFL. Finally, these findings are useful to societies that attribute importance to IFL development. They can use these findings to build parents' awareness on how to communicate with their children about Islamic financial knowledge, which may influence their children's financial behaviour. Regarding limitations and future research, first, this study faces a common method bias problem because it is a cross-sectional study. Even though it has conducted Harman's singlefactor analysis, for future research it is suggested that a longitudinal study be carried out. Secondly, this study focused on members of Gen Z who were born from 1995 to 2001, which means that most of the respondents were students and their financial behaviour is affected by their parents. Future research may include older respondents who have an occupation. Presumably, they will decide their own financial behaviour. Finally, future research might include other factors which might influence IFL and financial behaviour such as demographic factors and religiosity.

Purpose -This research aims to examine the effect of family communication patterns on Islamic financial literacy (IFL) and risk tolerance of Generation Z (Gen Z) in Indonesia. It also examines the link between IFL, risk tolerance and financial behaviour. Finally, it highlights the moderating impact of IFL on the link between risk tolerance and financial behaviour. Design/Methodology/Approach -This study used Muslim members of Gen Z in West Sumatra Indonesia who were born between 1995 and 2001 as the population. This study collected 668 responses in two months. After some preliminary tests, such as missing value and normality test, it employed 571 usable responses as the sample. Purposive sampling was conducted and variancebased structural equation model-partial least square (SEM-PLS) was employed as the method for data analysis. Findings -The results show that family communication patterns, both conformity and conversation orientation, positively and significantly influence IFL and risk tolerance. IFL has a positive and significant impact on risk tolerance and financial behaviour. Surprisingly, IFL also moderates the link between risk tolerance and financial behaviour positively and significantly. Finally, risk tolerance significantly affects financial behaviour.

Introduction Compared to other described processes, which boil down to the exclusion of specific social groups, such as sexism or racism, ageism has been noticed and described relatively recently [1]. The current definition, developed by Iversen et al. in 2009, can be characterized as "negative or positive stereotypes, prejudices and/or discrimination against people based on their chronological age. Ageism can be hidden or explicit, and it can be expressed at the micro, meso or macro level" [2]. The latest research on the phenomenon of ageism in individual age groups has shown that the phenomenon of discrimination is more often noticed and reported by young people, but this process is common in all groups at risk of exclusion on the basis of age [3]. The influence of mood and self-feeling on the phenomenon of stereotyping, which is the main factor that may lead to the development of ageism, has already been described many times. In a study conducted almost 30 years ago, in which the respondents were evoked positive, neutral, and negative moods, a different perception of minority social groups in the context of their stereotyping was noticed [4]. These observations were corroborated in another study, which also induced a certain mood in participants, thanks to which it was noted that in people with a negative mood and malaise, the stereotyping process may be more radical [5]. When trying to determine the influence of both mood and well-being on the tendency to age discrimination, one should not forget about the influence of time spent in the company of elderly people, which may also be important in shaping attitudes towards older people [6]. Based on the research conducted so far on the phenomenon of exclusion of the elderly, we have put forward a hypothesis that the above-mentioned factors may be potential predictors of attitudes towards the elderly in Polish society. --- Material and Methods 2.1. Objectives. The aim of our research was to determine whether the type of personality, manifested mood, and well-being can influence attitudes towards the elderly. We also asked detailed research questions: Can attitudes towards older people be associated with our positive affect? Can attitudes towards older people be associated with our well-being? Can attitudes towards older people be associated with contact with the elderly? When completing the database, we wanted the distribution in individual age groups and gender distribution to be proportional. Only fully completed adult questionnaires were included in the analysis. --- Methods. During the research, we used the following tools: demographic questionnaire, Kogan's Attitudes Toward Old People Scale, Euthymia Scale, and WHO-5 Well-Being Index. Kogan's Attitudes Toward Old People Scale contains 34 statements which the respondent chooses from 6-point Likert answers [7]. The respondent could receive from 34 to 204 points. Higher scores indicate a more positive attitude towards older people, and lower scores indicate a more negative attitude towards the elderly [8]. We used the 10-item Euthymia Scale [9] to assess the manifested mood. Each question of the questionnaire could be marked either true or false. The presented tool was based on Jahoda's conceptualization of euthymia [10,11]. Another tool used was the WHO-5 index, which we used to assess well-being, applying a 6-point Likert-like scale . The maximum number of points was 100, and scores < 50 suggesting impaired well-being were assumed, while ≤28 indicate likely depression [12]. The last used tool was the proprietary questionnaire with questions about one's own experiences in the field of contacts with the elderly . --- Ethical Approval. The study was approved by the Bioethics Committee of the Nicolaus Copernicus University Collegium Medicum in Bydgoszcz, Poland . The research was conducted in accordance with the Helsinki Declaration. All participants provided informed consent for the research. 2.5. Data Analysis. We used R 4.0.2 to analyse the data [13]. To model the KOAP score as a function of the predictors, we used Bayesian robust linear regression with t distribution [14]. In each of the analyses, the "no opinion" response In Bayesian statistics, the inference is based on analysing the posterior probability distributions of a model parameters , obtained by integrating likelihood with prior probability distributions. Regression weight is said to be statistically credible when 95% credible intervals of the posterior distribution exclude zero [14]. A point estimate of the effect means the posterior distributions are presented. Default improper flat priors were used for the regression weights. To investigate the relationship between dependent variables, credible predicted marginal means with corresponding 95% CI are presented in figures. These values represent the median of a posterior distribution of the predicted KOAP values. To approximate posterior distributions of the models, Markov chain Monte Carlo sampling procedure was used via brms package [15]. Six parallel chains were used, each consisting of 8,000 samples, with 4,000 samples used as warm-up period and every 10th sample recorded, resulting in 2,400 recorded samples in total. Sampling procedure was efficient and resulted with well-mixed and not autocorrelated chains and unimodal posteriors. Model accuracy was assessed with posterior predictive checks [16]. --- Results The characteristics of the participants are presented in Table 2, and descriptive statistics for KOAP, WHO-5, and ES are presented in Table 3. Model coefficients of the Bayesian robust linear regression with KOAP as dependent variable, WHO-5 and ES scores as predictors, and frequency of contacts with the elderly with professional status as additional control variables are summarized 4, and the model predictions are presented in Figure 1. We did not observe statistically credible relationships between WHO-5 or ES scores and the KOAP score in the tested model. --- Discussion The most important findings of our study were that the levels of well-being and euthymia in the participants of our study were low. Recognizing low levels of well-being is clinically important, because it affects the overall functioning of the patient and may affect the manifestation of symptoms, their severity, or the patient's concentration on negative aspects. It should be noted that the tools we used contain positively formulated statements, in contrast to other scales focusing on negative symptoms or problems reported by the evaluated persons [18,19]. Considering the low result obtained in our own research, it is worth screening the level of wellbeing in order to struggle with people at risk of mental difficulties in the future, or those already struggling with them and not covered by appropriate medical care, and if appropriate, proposing professional support. It is possible that the low result was influenced by the period of time in which the research was conducted, namely, the period of the COVID-19 pandemic. It is worth emphasizing the relatively low average point value in the WHO-5 Well-Being Index obtained by the participants of the study. It amounted to 54.87 points, where the value < 50 suggests poor emotional well-being and is a sign for further testing [12]. The impact of the duration of the study during the SARS-CoV-2 pandemic on the results obtained is corroborated by the results of studies conducted by White and Van Der Boor among 600 British, where the mean WHO-5 point value was 41.72. The authors of the study emphasize that both self-isolation before official closure and the feeling of isolation during a pandemic were associated with poorer mental health, general well-being, and quality of life [20]. Our participants' results are lower than the average reported in another study that was also conducted online also during the pandemic period [21]. In our study, we did not observe statistically significant relationships between the sense of well-being assessed in the WHO-5 Well-Being Index scale and ageism, nor 3). Vertical lines and shaded area are 95% credible intervals. Note: KOAP, WHO-5, and ES are on a standardized scale. Grey transparent points show data. 4 BioMed Research International statistically credible relationships between Euthymia Scale results and ageism. However, there are some critical points to be made with regard to our research methodology itself; for example, some people with a specific personality are involved in filling out questionnaires at all [22]. Despite the possible limitations, the study has many strengths-a large research sample and multifaceted approach to the phenomena of ageism contribute to deepening the knowledge in this field. Research on personality and manifested attitudes towards the elderly was also undertaken by Allan et al., who observed a relationship with attitudes towards old age, where personality traits had a direct or indirect impact on attitudes towards the elderly [23]. In other studies, the authors also emphasize the relationship between the fear of death and dying and ageism, where it was corroborated that agreeableness, openness, and fear of death of others were negatively related to the phenomenon of ageism [24]. --- Limitation We did not observe statistically significant relationships between the sense of well-being assessed in the WHO-5 Well-Being Index scale and ageism, nor statistically credible relationships between Euthymia Scale results and ageism. Both self-isolation before official closure and the feeling of isolation as well as concerns about the continuing pandemic were associated with poorer mental health, general wellbeing, and quality of life [20]. Unfortunately, the study itself showed many limitations in the research tool used to assess personality. It can be assumed that active Internet users or people who voluntarily and willingly participate in the survey may also represent certain personality types [25][26][27]. In the future, it would be useful to take advantage of the marked expansion of the literature and to consider the choice of this instrument and also to develop methodologies for individual meetings with respondents. --- Conclusion The conducted study did not show any correlation between mood, sense of well-being, or contact with the elderly and attitude towards the elderly. The results of our study may become the basis for further research to find the relationship between the mood and well-being shown in attitudes towards the elderly. Perhaps, however, the assumption should also be made that there is no such relationship. Regardless of the position taken, research in this area should be continued, as it will allow the development of measures to reduce ageism. Future studies may consider the potential role of a pandemic or include postpandemic data collection. --- Data Availability Data are available on request from the corresponding author [email protected]). --- Conflicts of Interest The authors declare that they have no conflicts of interest.

Objectives. The age discrimination, i.e., ageism, is still a current social problem. Therefore, it is justified to conduct studies increasing the knowledge about this phenomenon. The aim of our research was to determine whether the type of personality, manifested mood, and well-being can influence the attitudes towards the elderly. Material and Methods. During the research, we collected information from 923 participants and we used the following tools: demographic questionnaire, Kogan's Attitudes Toward Old People Scale, Euthymia Scale, WHO-5 Well-Being Index, and the author's questionnaire on contacts with an elderly person. Results. We used R 4.0.2 to analyse the data. In our current model study, we did not observe statistically significant relationships between the WHO-5 or ES scores and the KOAP score. Conclusions. The conducted study did not show any correlation between mood, sense of well-being, or contact with the elderly and attitude towards the elderly. The results of our study may become the basis for further research to find the relationship between the mood and well-being shown in attitudes towards the elderly. Perhaps, however, the assumption should also be made that there is no such relationship.

There is a substantial body of research on intergroup relationships between children and early adolescents. The predominant focus in this research is either on gender differences or on racial and ethnic groups. In general, children have a more positive attitude toward their own gender and toward their racial or ethnic in-group compared to out-groups . Gender, race, and ethnicity are important social markers that define group identities and that can lead to intergroup tensions and conflicts. However, children's social world is structured by many other characteristics, including religion. Although there is research on children's understanding of religious beliefs and practices , very little is known about religious intergroup relations. This is unfortunate for at least two reasons. One is that religious differences can be of great importance to children's lives and religious groups are among the more salient buttresses of identity . The other is that questions of diversity are increasingly questions of religious diversity. In particular, Islam has emerged as the focus of immigration and diversity debates in Europe and is at the heart of what is perceived as a "crisis of multiculturalism" . This study is concerned with Christian, Muslim, and nonreligious early adolescents living in the Netherlands. Early adolescence is an important period for the development of group identities and intergroup attitudes . During this period children begin to identify more closely with, for example, their ethnic or racial group and they start to recognize the broader social implications of ethnic and racial group differences . The present study focuses on early adolescents' feelings toward Muslims, Christians, Jews, and nonbelievers. Furthermore, the degree of religious group identification was examined, as was the association of identification with the children's feelings toward the various groups. In addition, it was examined whether the religious composition of the school class affects children's group evaluations. The theoretical basis for examining these issues is derived from social developmental work on intergroup relations. --- Religious In-Group Evaluation and Identification Social psychology has a long tradition of examining individual's attitudes and behaviors toward their in-group and out-groups. These attitudes are typically examined in relation to group identification and the intergroup context. One of the most influential social psychological perspectives is Social Identity Theory . This theory is increasingly being used to examine intergroup relations among children and early adolescents , and Social Identity Development Theory has been proposed by Nesdale . Nesdale argued that group evaluations emerge depending on, first, the extent to which children identify with their social group, second, the norms and beliefs held by the members of the child's social group, and, third, the extent to which the out-group is perceived as threatening one's group in some way. According to SIDT, individuals seek to belong to groups that provide them with a secure and positive social identity and are motivated to maintain positive distinctiveness through intergroup comparisons. This tendency and motivation implies that people will tend to evaluate their in-group more positively than out-groups. Hence we expected that the different groups of participants will evaluate their religious in-group more positively than the different out-groups. SIDT interprets group identification in terms of individual differences in the degree to which psychologically central and valued group memberships develop. Some children are more inclined than others to see themselves as a religious group member and to value their religious group membership. Children who feel highly committed to their group are inclined to act in terms of their group membership. Thus the tendency to evaluate the in-group positively is a function of the intensity of group identification. That means that we can expect that for the Christian and Muslim early adolescents, stronger religious group identification is associated with more positive feelings toward their religious in-group. We did not consider group identification of the nonreligious early adolescents because in the context of the Netherlands there is no single community of nonreligious people. Thus individuals can define themselves as being nonreligious but there is no clear community to identify with. The category of nonreligious people is significant for "outsiders" but it is not an organized community similar to Christians or Muslims. --- Religious Out-Groups and Threat According to SIDT , not only group identification but also in-group norms and beliefs as well as the perception of out-group threat instigate children's prejudice. We did not examine these latter two conditions directly but used the Dutch intergroup context to derive specific hypotheses. Previous studies have shown that children tend to reproduce the social representations or societal discourses and shared beliefs about the nature of group differences. In a multinational study among 6-year-old children, for example, Bennett and colleagues found that out-group attitudes were influenced by the own nation's particular widespread beliefs about other nationalities. . Furthermore, Bar-Tal and Teichman have shown that in a societal context of group conflict, children as young as 4 years absorb negative messages about out-groups and report fear when viewing drawings of out-group members. In the last 7 to 8 years in the Netherlands, Islam has increasingly become the "negative other." In the Dutch media, Islam has become symbolic for problems related to ethnic minorities and immigration and influential politicians have defined Islam as a backward religion and Muslims as a "fifth column," and have argued that "a cold war against Islam is unavoidable" . Hence according to some commentators there is an ongoing "Dutch-Muslim" cultural war . Research has shown that half of the Dutch general population and of Dutch middle adolescents have explicit negative attitudes toward Muslims. These attitudes are strongly related to perceived threat to Dutch identity and culture . Muslims are typically considered to undermine traditional Dutch values and norms. The Netherlands is one of the most secular countries in the world but it also has a Christian history and Christian values and beliefs remain present in many spheres of live. Thus it can be expected that Christian and nonreligious early adolescents will have more negative feelings toward Muslims compared to other religious out-groups. The societal emphasis on the need to compel Islamic minority groups to assimilate implies that Islamic youth in the Netherlands face high levels of threat to their religious identity and the ability to maintain a valuable and distinctive identity. For example, early adolescents of Turkish background, who are predominantly Muslim, have been found to experience the highest level of peer discrimination . According to SIT, under identity threatening circumstances, people will try to maintain or restore a positive and distinct collective identity. Therefore, we expected that Muslim identity will be very important to most of the Muslim early adolescents. In addition, identity threat can lead to negative feelings toward out-groups . To enhance the value and distinctiveness of one's own religious group, group members can derogate other religious groups . Hence the feelings of the Muslim early adolescents toward Christians can be expected to be rather negative. Two contrasting hypotheses about the role of religious group identification for the evaluation of the religious out-group can be formulated. Early adolescents with a strong group identification derive the most meaning from their religious group affiliations. This means that they can be expected to be sensitive to out-group threats. This leads to the expectation that for the Christian and Muslim early adolescents higher group identification is associated with more negative feelings toward the religious out-group . On the other hand, it can be argued that Christianity and Islam have many similarities and that religious group identification increases the respect and appreciation for fellow-believers. The shared religiosity might act as a kind of superordinate category that reduces negative out-group feelings, particularly for high group identifiers . This process of shared religiosity does not apply to nonreligious people. --- Nonreligious Out-Group Religious belief is not so much about personal preferences or social conventions, but rather about convictions. It is concerned with the moral good and divine truth that is difficult to reconcile with moral and epistemic diversity. The observant believer believes that he or she is right and will find it difficult to have positive feelings toward nonbelievers who implicitly challenge his or her religious life. In a study in Rotterdam, around 45% of Muslim adolescents indicated that they had "no sympathy" for nonbelievers . Similarly, 31% of Christian youth were found to have unfavorable views about nonbelievers. Hence it is likely that the feelings of Christian and Muslim early adolescents toward nonbelievers will be rather negative, especially for high religious in-group identifiers. Thus for the Christian and Muslim participants we expected religious group identification to be negatively associated with the feelings toward nonbelievers. --- Jewish Out-Group SIDT argued that not only out-group threat can underlie children's group evaluations but also in-group norms and beliefs that promote out-group rejection or acceptance . In real-life situations perceived out-group threat and in-group norms and beliefs are often confounded. Hence it is not easy to draw conclusions about the role of these norms and beliefs in addition to threat. A possible solution is to examine the attitude toward a religious out-group that does not present a threat to the position and identity of the in-group, like the Jews. The Jewish community in the Netherlands is very small, geographically dispersed, very well integrated, and not clearly visible in public life. Hence it is very likely that the early adolescents' feelings toward Jews are based on in-group norms and beliefs rather than on perceived threat. We expected the feelings toward Jews to be quite negative among the Muslim children. In many European countries, including the Netherlands, there is a growing concern about increased anti-Semitism, which certainly is not only limited to Islamic groups but which has also been manifest in, for example, Mosques across Europe and among Muslim youth . This concern is further illustrated by the many books, Web sites, and school initiatives that find it necessary to argue that Islam actually denounces anti-Semitism. One source for this anti-Semitism is the ongoing conflict between Palestinians and Israelis, and between observant Muslims and observant Jews in particular . For some Muslims, the Jews form a "negative other" and there are indications that Dutch Muslim children reproduce the negative beliefs about Jews that exist in parts of the Muslim community . Hence we expected Muslim early adolescents to reproduce these unfavorable beliefs and to have rather negative feelings toward Jews. These feelings were expected to be strongest among the high group identifiers. For the Christian early adolescents the situation is different. Their religious identity is not challenged by the Jews and anti-Jewish sentiments are not common in mainstream Dutch society. This implies that the religious norms of acceptance and tolerance are more likely to guide children's out-group reactions. Thus Christian early adolescents were expected to have more positive feelings toward Jews than the Muslim children. --- School Composition and Contact In their meta-analytical examination of the role of intergroup contact in predicting children's intergroup attitudes, Tropp and Prenevost concluded that contact promotes positive intergroup attitudes. In general, contact appears to improve group relations even when Allport's optimal conditions are not fulfilled . Having frequent contact with out-group members may increase liking and positive affect via mere exposure . Through frequent contact children can acquire out-group knowledge, correct negative stereotypes, and develop sympathies. Among Dutch adolescents, for example, more contact with Muslims turns out to be related to less anti-Muslim feelings . A school class is a social context in which contact with classmates is inevitable. However, school classes differ in the opportunities for intergroup contact. Various studies have examined the effect of classroom composition on intergroup attitudes and friendships. In the context of the United States, Quillian and Campbell , for example, found that cross-race friendships increase with school racial diversity and Moody showed that friendship segregation declines at highest levels of diversity. In the Netherlands, Verkuyten and Thijs found a similar result for intergroup attitudes and ethnic diversity in schools. Thus more opportunities for interreligious contact in school classes can be expected to be related to more positive feelings toward religious and nonreligious out-groups. We tested this expectation by examining whether the proportion of classmates of a particular out-group has a positive effect on the evaluation of that out-group. --- In Summary The aim of this study is to examine religious group relations among Christian, Muslim, and nonreligious early adolescents living in the Net herlands. We expected, first, that the three groups of participants will favor their in-group over the out-groups and that for the Christian and Muslim participants higher group identification is positively related to in-group eva luation. Second, and considering the intergroup context in the Netherlands, we expected Muslim children to be rather negative toward Christians and nonbelievers and the latter two groups of participants to be rather negative toward Muslims. Furthermore, we examined whether in-group identification was associated with the evaluation of these out-groups. Third, Jews were expected to be evaluated more negatively by the Muslim participants than by the Christians and higher Muslim identification was expected to be associated with a stronger negative attitude. Fourth, in line with the contact hypothesis, we expected that a higher proportion of classmates of a particular religious out-group is positively related to the evaluation of that out-group. --- Method --- Participants and Procedure Participants were 1,064 Grade 5 and 6 students from 24 regular primary schools in a medium-sized town in the west of the Netherlands. They completed a short questionnaire in their classrooms. Participation was voluntary and anonymity was guaranteed. All children were willing to participate in the study. In the present analyses, we included all participants who self-identified as either Christian , Muslim 1 , or nonreligious . These 1,011 students visited 48 self-contained classes with an average size of 22.17 students. Their mean age was 11.01 years , and 45.2% were female. 2 --- Measures Religious group membership and religious classroom composition. Students' religious group membership was assessed by means of self-definition. To examine the differences between Christian, Muslim, and nonreligious students, three dummy variables were calculated for each respondent group . As described below, multilevel analyses were used. These revealed that a significant portion of the variation in the group dummies was at the class level . This indicates that classrooms differed in religious composition. Thus for each classroom we calculated the proportion of Christian students , Muslim students , and nonreligious students . The first two measures were not significantly related . However, the proportion of nonreligious students showed strong and significant negative relations with the proportions of Christian and Muslim students . Religious identification. Religious identification was assessed with three items that are typically used in social psychological measures of in-group identification : "How important is your religion to you?" "Are you proud of your religion?" and "Do you sometimes get angry when people do not respect your religion?" Answers were rated on a scale ranging from 1 to 5 . For both Christians and Muslims the items loaded on one component explaining, respectively, 70.6% and 69.2% of the variance. Cronbach's alpha was .79 for the Christian participants and .75 for the Muslim participants. Religious group evaluations. Students' evaluations of, respectively, Christians, Jews, Muslims, and nonbelievers were assessed with so-called "feeling thermometers." The feeling thermometer has been used as a global measure of in-group and out-group feelings among members of different ethnic and religious groups . For each target group, the thermometer ran from 0 degrees to 100 degrees. Students were instructed that 100 degrees indicates very positive or warm feelings, 0 degrees indicates very cold or negative feelings, and 50 degrees means neutral feelings. Following this instruction, four religious groups were listed in the following order: Christians, Muslims, Jews, and nonbelievers. Under each target group a scale was presented running from 0 to 100. --- Data Analytic Strategy In the Netherlands, primary school children spend their days in the same class with the same classmates. As respondents were nested within school classes, their individual responses were probably not statistically independent. For this reason, we used multilevel analyses to properly estimate the effects of the classroom composition measures and religious identification. To simultaneously examine students' evaluations of the different religious groups multivariate multilevel models were tested using MLwiN version 2.0 . In these models, three levels were specified: Level 1 was included to define the multivariate structure and pertained to the different evaluations by individual students, Level 2 concerns differences within classes, and Level 3 concerns differences between classes. All models were estimated using the Iterative Generalized Least Squares algorithm, and relative model improvement was assessed by comparing the fit of nested models. Differences between these statistics follow a chi-square distribution with degrees of freedom given by the difference in parameters . Prior to examining the multilevel models, we conducted repeated measures analyses to test our hypothesis that each group of participants favored the in-group over the out-groups. To account for the multilevel structure of our data, the evaluation measures were corrected for mean differences between classes. That is to say, the repeated measures analyses were performed on the pooled within-classes covariance matrix. --- Results --- Group Evaluations Table 1 shows statistics for the four group evaluations reported by the three groups of early adolescents. It can be seen that the large major ity of participants reported positive feelings toward their own group , and this held for Muslims in particular. Furthermore, Muslim participants reported negative feelings toward the out-groups . Almost half of these students was explicitly negative about Jews and a third indicated negative feelings toward Christians. In turn, almost a third of the Christians and of the nonbelievers were negative toward Muslims. To examine whether participants preferred their in-group over the outgroups, repeated measures analyses were performed separately for each group of participants on the pooled within-classes covariance matrix. For each group, we tested the contrasts between the in-group and each of the three out-groups. To reduce the chance of Type 1 errors, significance levels were multiplied by four. Results clearly demonstrated a preference for the in-group among all groups of participants: Christian early adolescents evaluated Christians more positive than Jews, F = 10.17 --- Variance Distributions and Group Differences Before we examined the effects of classroom composition and religious identification, two preliminary multilevel analyses were conducted. First, we examined the variance distributions of the four group evaluations across Level 2 and Level 3 . 3 We specified a multivariate model with intercepts only . For each dependent variable, this model provided an estimate of the intraclass correlation coefficient , which represents the proportion of variance at Level 3 . For all variables this coefficient was significant at p < .001: ρ was .086 for evaluations of Christians, .091 for evaluations of Jews, .264 for evaluations of Muslims, and .078 for evaluations of nonbelievers. Thus classes differed with respect to the average group evaluations reported by their students, particularly for the evaluations of Muslims. Second, we tested mean differences in evaluations between the three groups of early adolescents. To this aim, we estimated a multivariate regression model with dummy variables representing the three different groups. This model fitted the data significantly better than the intercept-only model, χ 2 = 558.86, p < .001. Results are shown in the left part of Table 2. In this model , dummy variables are included for Christian versus other , and Muslim versus other early adolescents. This means that nonreligious children are the referent group. Mean levels for this group are indicated by the constants in Model 1. 4 It appeared that Christian early adolescents rated Christians and Jews, respectively, 15.75 and 6.44 points more positively, and nonbelieving people 9.70 points less positively than nonreligious participants. The evaluation of Muslims was similar for both groups. In addition, Muslim early adolescents evaluated Christians, Jews, and nonbelievers more negatively but Muslims more positively than nonreligious participants . Differences between Christian and Muslim early adolescents cannot be directly inferred from Table 2. However, they can be obtained by subtracting the effects of Muslims from those of Christians . Compared to Muslim participants, Christian early adolescents rated Christians, Jews, and nonbelievers more positively and Muslims less positively . --- Classroom Proportions Next, we examined whether students' evaluations of Christians, Muslims, and nonreligious people depended on the actual presence of these groups in their classroom, that is, the proportion of the target group. 5 To test our hypothesis concerning the role of contact opportunity, three interaction terms were calculated. For each group that was evaluated we multiplied the corresponding classroom proportion measure with the Level 2 dummy representing the matching participant group . As a consequence, when the proportion measures and the dummy interactions were simultaneously added to the model, the "main" effects of target proportion represented the effect of out-group presence . Model 2 is the model in which these three interactions and the three proportion measures are included. Overall, the model with classroom proportions was a significant improvement over Model 1 , χ 2 = 37.5, p < .001. The proportion of Christian classmates did not affect Muslim and nonreligious early adolescents' evaluations of Christians. However, with respect to the evaluations of both Muslims and nonreligious people, there were significant effects of out-group proportions. Thus in classes with more Muslim students, both Christians and nonreligious early adolescents were more positive about Muslims . Likewise, Christian and Muslim early adolescents evaluated nonbelieving people more positively when they had more nonreligious classmates . As can be inferred from Table 2, target group proportion and its interaction with the group dummy explained 11% of additional between-classroom variance in the evaluation of Muslims and 26% of additional classroom variance in the evaluation of nonreligious people. Further inspection of Model 2 revealed that the Muslim early adolescents no longer evaluated nonreligious people more negatively than the Christian participants when the influence of target group proportion was partialled out. Furthermore, a t test revealed that the Muslim participants had less nonreligious classmates compared to the Christian children, t = -3.903, p < .001. Thus the different evaluation of the nonreligious target group by Muslim and Christian early adolescents was accounted for by the representation of that group in the classroom. --- Religious Identification In a last set of analyses, we examined the role of religious identification. Because identification was not measured among the nonreligious children, these analyses focused on the Christian and Muslim participants . To allow the testing of more parsimonious models, data were transformed in three manners. First, the four group evaluations were converted into one measure for in-group evaluation and three measures for out-group evaluation. The latter involved two religious out-groups and one nonbelieving out-group. Second, two measures were computed representing the proportion of in-group classmates and the proportion of classmates from the religious out-group . Note that these measures were Level 2 variables that varied for different students within classes. Third, both religious identification and the four evaluations were transformed into z-scores to make meaningful comparisons of the effects of the former on the latter. To examine whether the effects of religious identification differed for the Christian and Muslim early adolescents, we computed an interaction term. Prior to this calculation, the dummies were centered at their means. Thus the effects of identification could be interpreted as proper main effects. Two models were tested . In Model 3a, group evaluations were regressed on participant group , religious identification, the interaction between religious identification and participant group, and in Model 3b we added the proportion of target group students in the classroom . Note that for the ingroup and religious out-group evaluations, the proportion of the target group was a Level 2 rather than a Level 3 variable. Separate regression coefficients were estimated for each evaluation. First, in a model without the proportion of target group there was a significant interaction effect between religious group and religious identification. It turned out that for the Muslim participants, higher identification was related to a less positive evaluation of the Christian out-group , whereas for the Christian participants, higher religious identification was associated with a more positive evaluation of the Muslims . This interaction effect was no longer significant, however, in Model 3b that included the proportion of the target group. Separate regression coefficients were estimated for each evaluation. Model 3b is shown in Table 3. For all evaluations, except that of the Jewish out-group, the effects of identification were similar across participant group. Compared to low-identifiers, students who strongly identified with their religious in-group reported more positive feelings toward their in-group, similar feelings toward their Muslim or Christian out-group, and less positive feelings toward nonbelievers. In addition, with respect to the evaluation of Jews, religious identification had no effect for Christian participants but, as expected, a negative effect for the Muslim children . This latter effect was also found in model 3a that did not correct for the proportion of the target group. --- Discussion Using social identity theory , we have examined religious group identification and feelings toward multiple religious groups and nonreligious people among a large sample of early adolescents in the Netherlands. The results suggests that there is a Dutch-Muslim divide, similar to what is observed in Dutch society in general . Around a third of the Muslim early adolescents explicitly indicated to have negative feelings toward Christians and one in four had negative feelings toward nonbelievers. In contrast, the Christian and nonreligious participants had the least positive feelings toward Muslims with around a third indicating to have negative feelings. Thus many children reported explicit negative feelings and these reactions differ from the great majority of research on children's intergroup attitudes that typically find in-group preference and not out-group dislike . These findings are probably related to the intergroup situation in the Netherlands in which Muslims are portrayed as undermining Dutch culture and identity. Dutch adolescents have been found to perceive Muslims as a threat and this perception underlies prejudice toward Muslims . In addition, the public condemnation of Islam and pleas for assimilation forces Dutch Muslims in a position of having to defend and stress the importance of their religion and religious group. Islamic immigrant groups face high levels of threat to the value of their religious identity that can lead to increased in-group orientation and identification among members of these groups . As expected, all groups of early adolescents had more positive feelings toward their ingroup than the out-groups but the in-group feelings of the Muslims were the most positive ones. Furthermore, Muslim early adolescents had higher religious in-group identification than Christian contemporaries. The intergroup situation can also lead to less positive attitudes toward out-groups . To enhance the value and distinctiveness of one's Muslim group, group members can keep a distance toward or derogate other religious groups . Hence the findings suggest that in an identity threatening context social identity needs rather than religious norms of tolerance and acceptance drive early adolescents' evaluations of religious out-groups. Social identity development theory argues that out-group evaluations depend not only on perceived out-group threat but also on in-group norms and beliefs that promote out-group rejection or acceptance. A central teaching of religions is that one should accept and love others, including those that think and act differently. This normative orientation is an important guideline for perception and behavior when one identifies with one's religious group. However, religious belief is also about convictions and is difficult to reconcile with moral and epistemic diversity. Nonbelievers implicitly challenge one's way of life particularly for high religious group identifiers because they derive the most meaning from their religious group affiliations. It turned out that Christian as well as Muslim early adolescents who identified strongly with their religious in-group were more negative toward nonbelievers than early adolescents with a weak religious group identification. Hence the religious norm of acceptance and tolerance did not involve people who are not religious. The role of in-group norms and beliefs is also important for understanding the Muslim early adolescents' feelings toward Jews. In the Netherlands, the Jewish community is very small and does not form a group threat. However, almost half of the Muslim participants was explicitly negative toward Jews. In addition, Muslim early adolescents who identified more strongly with their religious in-group were more negative toward Jews. In the Netherlands, as in other West European countries, there is a concern about increased anti-Semitism among Muslim youth . And there are some indications that Muslim children reproduce the negative attitudes toward Jews that exist in parts of the Muslim community . The current findings point in the same direction. Religious group feelings did not only depend on individual differences in, for example, in-group identification. Independently and for all target groups, there were also systematic differences between school classes. This means that in addition to individual differences children in the same classes are more similar in their group feelings than children in different classes. The class level accounted for around 9% of the total variance in feelings toward nonbelievers, Christians, and Jews and explained no less than 26% of the variance in feelings toward Muslims. This indicates that school classes matter for children's religious group feelings and they matter particularly strong for the feelings toward Muslims. The current findings show that increased opportunity for contact between early adolescents from different religious and nonreligious groups contributes to positive group relations. Christian and nonreligious early adolescents had more positive feelings toward Muslims when the proportion of that out-group in class was higher. In addition, Christian and Muslim participants with more nonreligious classmates were found to evaluate nonbelieving people more positively. Furthermore, the Christian and Muslim early adolescents evaluated the nonreligious group similarly when the proportion of nonbelieving classmates was taken into account. These results show that the positive effects of contact opportunities are not restricted to ethnic or racial groups who are the focus of most of the contact research and also not to children who belong to the dominant majority in society. Thus the results contribute to the generality of contact theory. However, it also turned out that the proportion of Christian children in the class did not have an effect. This may be due to the fact that Christian identity is less uniform because of different denominations and less visible for early adolescence than, for example, Muslim identity. Furthermore, we focused on the opportunities for contact and did not consider actual contacts. Religious heterogeneous schools may still be substantially segregated along religious lines. We were not able to examine the role that the proportion of Jewish classmates plays in the attitude of the Muslim early adolescents. However, it is likely that increased exposure of Muslim adolescents to Jewish classmates will increase positive feelings toward Jews, depending on the way that the contact is presented and organized. In the context of widespread negative beliefs, mere exposure might confirm existing stereotypes and negative attitudes rather than increase out-group knowledge and positive affect . To summarize, we have examined attitudes toward religious groups among Christian, Muslim, and nonreligious early adolescents. The findings show that there is a clear group divide between Muslims early adolescents, on the one hand, and Christians and nonbelievers, on the other hand. This suggests that children tend to reproduce societal discourses and widespread beliefs like the ongoing "Dutch-Muslim cultural war" . In addition, however, the findings show that the classroom context has a substantial influence on children's attitudes. Thus it is not only the broader social context that is important but also the concrete situation in which children find themselves in. --- Limitations In evaluating the present results some qualifications should be considered. For example, we used an intergroup perspective and this assumes that students recognized religious orientation as a meaningful way to group their classmates and that they were aware of each other's religious background. This assumption is likely considering the current debates on Islam and religion more generally and the fact that most schools teach about religions and religious differences. Furthermore, results showed that there were relatively strong effects for classroom composition. For the Christian and Muslim early adolescents, religious identification had an effect on out-group evaluation when the proportion of out-group classmates was not included as a predictor. The presence of Christian or Muslim out-group classmates did matter for the evaluation of the Christian or Muslim out-group and the findings for the classroom level suggests that the early adolescents did recognize religious orientations. However, it should be noted that early adolescents' attitudes will not only depend on the intergroup setting but also on, for example, parental attitudes, both directly and indirectly. Children can internalize the beliefs and attitudes that parents express in verbal and nonverbal ways, including widespread negative beliefs such as about Jews in some Muslim communities and about Muslims in the Netherlands . Parents can also have a more indirect influence on their children's attitude by, for example, their choice to send their child to a religious homogenous or rather diverse school. That would mean that other factors that vary along with classroom composition might explain the classroom-level differences found. This alternative explanation is not very likely in the current study because classrooms within schools varied in religious composition. 6 However, future studies should examine parental influences more systematically and in relation to intergroup factors. There are some other limitations that should be mentioned. For example, no measures of perceived out-group threat, of in-group norms and beliefs, and of actual intergroup contacts or friendships were available. Furthermore, we focused on general group feelings, but these feelings can differ from attitudes toward individual classmates from different religious groups. In addition, only 48 classrooms were considered and there was no information available on religious teaching in the different schools. Religious identification was measured with items that are commonly used in social psy chological research. However, it seems important to examine different dimensions of religious identity in future studies, such as religious beliefs, behaviors, and practices. For example, the degree to which adolescents subscribe to the religious concepts of acceptance and tolerance could be assessed. It is possible that for other dimensions, religious group identification plays a different role in intergroup relations. Furthermore, future studies could examine religious group attitudes among other religious minority groups as well as different groups of Christians and Muslims living in the Netherlands. In addition, it would be interesting to examine these issues in other European and non-European countries. The Netherlands is one of the most secular countries in the world and European nations differ in their historical and current approaches for dealing with religious and cultural diversity. Furthermore, in most European countries there is a long history of an established majority group, and issues stemming from immigration, migrant minorities, and diversity are relatively novel. In contrast, countries such as Canada and the United States are largely composed of immigrants and cultural diversity is a defining characteristic of these nations. Furthermore, countries such as the United States are much more religious than the Netherlands. --- Practical Implications The main practical implication of our findings relates to the school context. The classroom context appears to have a substantial influence on children's attitudes, explaining, for example, no less than 26% of the feelings toward Muslims. This means that the educational context offers important possibilities for improving religious group relations. Schools in the Netherlands are trying to make a contribution to more positive group relations in many different ways. They consider the classroom composition, they try to teach --- Journal of Early Adolescence 30 children basic aspects of democracy, citizenship and tolerance, and they use forms of multicultural education. Schools differ considerably in their approach and it is unclear which approach is most effective. What is clear, however, is that schools can make a real difference. Future studies should examine more closely why and which aspects of the school context, the curriculum, and teaching practices actually make a difference for children's religious group attitudes. --- Notes 1. There live around 850,000 Muslims in The Netherlands . The number of Muslim pupils at schools differs greatly, ranging from schools that have no Muslim children to Islamic schools that have only Muslim pupils. 2. Gender was equally distributed across the three religious groups. The Muslim participants were somewhat older than the Christian and nonreligious early adolescents . However, there were no unique relations between age and the four group evaluations. Therefore, age was not included in the analyses. 3. Variance calculations and proportion measures were based upon the religious self-definition of all 1,064 students who participated in the research. 4. Constants in Model 2 do not directly indicate mean levels for nonreligious participants because target group proportions are included. 5. We could not investigate this with respect to Jewish people. There were only five Jewish students in the sample making it meaningless to compute a proportion measure for this group. 6. The classrooms within schools varied in religious composition. The within-school variation for the proportion Christians was 23%. For the nonbelievers this percentage was 17%, and for the Muslims it was 11%.

This study conducted among Christian, Muslim, and nonreligious early adolescents living in the Netherlands used intergroup theory for examining religious group evaluations. There was evidence for a religious group divide with a third of the Christian and nonreligious participants explicitly indicating negative feelings toward Muslims, and Muslim children having negative feelings toward Christians, nonbelievers, and Jews. Furthermore, the Muslim early adolescents had high religious in-group identification and higher identification was associated with more negative feelings toward nonbelievers and Jews. In addition, the results show that increased opportunities in school for contact between early adolescents from different religious and nonreligious groups contributes to more positive group relations. It is argued that the pattern of results is in agreement with an intergroup perspective in which the role of the broader social context and the concrete situation is taken into account.

Introduction Family is the basic unit in society that can cause, prevent, improve, and influence the health of its members. 1 Families play an essential role in meeting the care needs of individuals who require assistance due to illness and/or disability. 2,3 However, without adequate support, family health and wellbeing can be compromised. As the number of individuals facing chronic illnesses continues to rise worldwide, including the ongoing COVID-19 pandemic, there is a pressing need to recognize the significant caregiving contributions made by family members. 4,5 According to the United Nations Educational, Scientific, and Cultural Organization's COVID-19 global monitoring of school closures, as of June 2021, around 50% of students worldwide were affected by partial or full school closures, impacting nearly 200 million pupils. 6 As of December 31, 2021, the number of household heads in Indonesia was 87.83 million people, according to data from the Ministry of Home Affairs. The total population of East Java in 2021 was approximately 40.16 million people. 7 Until June 2022, East Java Province ranked second among Java's provinces with the highest number of household N o n -c o m m e r c i a l u s e o n l y heads, reaching 14.09 million families with a population density of 857 people per square kilometer. Data from the Malang Regency Social Service, collected in 2019-2020, revealed that there were more than 250,000 families categorized into underprivileged, prosperous, and well-off groups. 8,9 To combat COVID-19, the Indonesian government enforced Large-Scale Social Restrictions starting in June 2020. The need to enhance the quality of life and family health became critical during these restrictions. 10,11 In terms of healthcare provision and services, patients, families, and communities are active partners in scientific research, collaborators in shared decision-making for care and related matters, and advocates for the health priorities pertinent to their communities. They are not mere passive recipients. 12 Family healthcare involves identifying health issues in the family, taking appropriate actions, providing care, promoting a healthy environment, using healthcare services, eating well and getting enough sleep, engaging in recreational activities, taking prescribed medications, addressing alcohol and tobacco use, and practicing self-care together as a family. 11,13 During pandemics, over 80% of participants found it somewhat helpful to care for their family, including children. COVID-19 disruptions affected child and family health and access to healthcare. 14 In the case of school-aged children, a study by Putri revealed that parents were required to assume the role of educators, serving as role models and motivating children to excel academically. 15 Parents had to take on the role of educators and motivators for school-aged children, as education shifted to homes. 16 However, the connection between family healthcare functions and school-aged children's development during pandemics remains underexplored. The enforced social isolation, along with associated lifestyle changes, presented significant challenges for families, particularly those with school-age children. 17,18 With school closures, working parents were compelled to care for their children throughout the day without regular support systems, such as day-care facilities or grandparents. Parents who attempted to work from home found it exceedingly challenging to balance their professional responsibilities with childcare and domestic work, resulting in decreased productivity. 19,20 These challenges led to poor dietary choices, limited physical activity, and increased obesity risk for children. Parents working outside the home as essential workers faced a difficult balance between their professional responsibilities and the needs of their isolated children. Indonesian parents, particularly dualearner families with young children, faced added strain due to their dual roles as workers and caregivers. 21,22 Children between the ages of 6 and 12 years typically begin to value friendships and become more involved in activities like sports and art. However, the health and social inequalities observed during the pandemic have the potential to restrict their social development. 23 School-age children have various independent social developmental tasks to fulfill. Additionally, as per the Vineland Social Maturity Scale, developmental tasks encompass self-help skills in general, dressing, eating, self-direction, socialization, occupational skills, communication, and locomotion. 24,25 During the pandemic, these developmental tasks were jeopardized due to increased sedentary behavior. A 2020 study with 656 primary caregivers in Canada found that poor parenting quality during the pandemic was linked to various household and pandemic-related factors, with caregiver depression consistently affecting the parent-child relationship. 26 Given this context, the objective of this study was to explore the relationship between family roles in healthcare provision and the social development of children. We hypothesize that there is a significant relationship between these variables, especially in the unique context of the COVID-19 pandemic, which has brought forth unprecedented challenges to family dynamics and child development. --- Materials and Methods --- Research design This study employed a correlational research design with a cross-sectional approach, aimed at analyzing the correlation between the family's role in healthcare provision and the social development of children. The research investigates how family roles, such as recognizing health problems, making appropriate decisions, providing care, utilizing healthcare facilities, modifying the environment, and engaging in recreational activities, influence the development of various skills in children, including dressing and eating independently, self-direction, self-socialization, selfhelp in general, communication, and locomotion. --- Study participants The study involved a total of 37 parents of grade 4-5 elementary school students. Participants were selected through convenience sampling based on specific inclusion criteria: they had to be parents of children aged 10-11 years in grades 4-5, complete the entire questionnaire, and be proficient in using a mobile phone. Ethical considerations were observed, and the necessary permissions were obtained from the school and homeroom teacher. Due to social restrictions, the study's explanation, informed consent, and the Google Form questionnaire were distributed exclusively within a social media group of parents . To minimize potential biases related to sampling, non-responses, or response bias, the researcher collaborated with the homeroom teacher and remained active in the parents' online group throughout the study. The researcher ensured the use of neutral and non-leading questions, allowed for participant anonymity, and refrained from using any brand identification. The study was conducted at Landungsari 01 Elementary School, Landungsari Village, Malang Regency in December 2021. --- Variable, instrument and data collection Demographic factors, including age, gender, education, occupation, and income, were considered as confounding variables. The independent variables included nine sub-criteria, focusing on various aspects of the family's role in healthcare provision. These criteria encompass recognizing family health problems, making informed decisions, utilizing healthcare facilities, modifying the environment, and managing family practices related to diet, sleep hygiene, recreation, therapeutic approaches, and substance use . The questionnaire used in this study was developed by Ayuningtyas . 27 The instrument employed a 4point Likert scale, where strongly agree = 4, agree = 3, disagree = 2, and strongly disagree = 1. The dependent variable was the social development of schoolage children, which included seven sub-criteria encompassing selfhelp in dressing and eating, self-direction, socialization, self-help in general, communication, and locomotion skills. The questionnaire for assessing school-age child social development was adapted from the Vineland Social Maturity Scale developed by Doll , 24 a commonly used tool for assessing the social development of school-aged children. 25 A Likert scale was used, with response options ranging from never performed = 1, rarely = 2, to always = 3. The reliability of this questionnaire was verified [ --- Data analysis The analysis conducted in this study includes descriptive analysis and a chi-square test. The chi-square test was utilized to assess the association between variables, although it does not provide information regarding the strength of the association or whether the relationship is causal. 29 Both variables in this study are categorical, mutually exclusive, and meet the expected values. --- Ethical clearance This research received ethical approval from the Health Research Ethics Commission under ethical certificate No. 335/EC/KEPK-S2/09/2021. Throughout the research process, the researcher adhered to ethical principles, including informed consent, respect for human rights, beneficence, and non-maleficence. --- Results Based on the demographic information presented in Table 1, the average age of fathers and mothers fell within the range of 37 to 38 years old. The majority of fathers had completed senior high school, undergraduate degrees, or held a Magister/Ph.D. level of education . Similarly, most mothers had educational backgrounds ranging from senior high school to undergraduate degrees and Magister/Ph.D. levels . In terms of parental occupation, 45.9% of fathers worked as private employees , regarding the occupation of mothers, approximately 40.5% engaged in domestic work at home . Table 2, which provides information on the distribution of family roles in healthcare provision, reveals that the majority of participants fulfilled their tasks at a rate of 83.8% to 100%. Notably, all parents successfully performed environmental modification or protection tasks for their family members, including school-age children. No participant failed in this aspect. Only 6 participants did not perform sleep hygiene practices for their family members. In Table 3, which outlines the distribution of schoolaged children's social development tasks, we observe that about 89.2% of the participants were able to successfully perform self-help in dressing. For self-help in eating, 70.3% of participants were proficient, 81.1% of children performed self-direction successfully, and 83.8% effectively developed self-socialization, even when adhering to social restriction rules. However, 75.6% exhibited less development in self-help in general, and 78.4% showed commendable development in communication and locomotion. The results of the correlational analysis, as displayed in Table 4, reveal significant associations between various variables. Notably, the family's ability to recognize health problems showed a negative and significant correlation with the self-direction development task among school-aged children . This suggests that when families perform poorly in recognizing health problems, school-aged children tend to exhibit higher levels of self-direction. These findings suggest that children have the potential to become self-directed learners, engaging deeply in their own learning processes to fulfill their needs. 30,31 However, by performing study at home or homeschooling due to pandemics the parents should developed more ideas in equipping, encouraging, enabling, and empowering especially to educate their own children also not merely in recognizing the family health problem. The parents were still obligated to give a direction to the school-age children in a properly and not intimidating. --- Discussion The research results have indicated significant associations between independent and dependent variables. Notably, the role of making the right decisions exhibited a significant negative association with the development of communication among school-aged children. In other words, when families perform poorly in making the right decisions, their children tend to develop stronger communication skills. In situations where families fail to make the right This finding underscores the significance of making the right decisions within families, as it serves as a protective factor against juvenile delinquency. 32 Communication skills can be enhanced through play and exploration in young children. These methods aim to foster affective and experiential connections with nature in young children, helping them establish bonds with the natural world. This foundation may encourage them to explore issues and take appropriate action as they mature. 33 Every interaction between parents and children constitutes a form of communication that not only teaches children how to interact with others but also shapes their emotional development and how they build relationships later in life. Effective communication skills can be developed by actively and reflectively listening, speaking clearly with children, avoiding the use of bribes to encourage better behavior, assisting them in reflecting on or explaining their feelings, and setting an example of kindness. These strategies can help develop children's communication skills, even when parents fail to make the right decisions. 34 Regarding the utilization of healthcare facilities as a family role, it exhibited a negative correlation with self-help dressing and self-help socialization. In contrast, with regard to self-help eating, this task displayed a positive correlation. The utilization of healthcare facilities represents one of the family's roles in maintaining the health and high productivity of its members. This function goes beyond the basic responsibility of maintaining the health status of family members. 27 A study conducted by Kaka, Putri, and Devi explained that family health care functions influence the nutritional status of children. 35 The results of this study suggest that families with poor health care functions have a 3.727 times higher risk of having children with poor and deficient nutritional status compared to families with good health care functions. Parental modeling is also key to making healthy food choices within families. 36 The role of healthcare is crucial in family assessment as the family serves as the fundamental unit in society, regulating behavior, implementing healthcare, and ensuring its protection. Families engage in preventive healthcare and jointly care for sick family members. Integrating the family into the healthcare team offers multiple benefits and can reduce barriers to healthcare utilization. 37 Encouraging families to establish and maintain connections with healthcare facilities is recommended for post-pandemic recovery planning. Schools should also reopen access to various family assistance and health programs that were hindered during social restrictions to promote social development and independence among children. [ Family dietary practices involve dietary arrangements aimed at preventing overweight and underweight, which can lead to various diseases. 38 Poor family dietary habits can lead to obesity, which is linked to health issues like high blood pressure, heart disease, cancer, and more. Assessing food choices should involve collaboration between families and caregivers. 39 Fewer healthy dietary practices can negatively affect children's self-help skills and socialization. 40,41 Sharing meals as a family promotes relational closeness and has various benefits, including emotional regulation, positive parenting, and conflict prevention according to Glazen's review study. 42 Cross-sectional studies linking overall family functioning and the frequency of family mealtimes have generally found positive connections. 38 In a study on engaging in active leisure, families with an adequate level of physical activity exhibited higher scores in happiness, contentment, well-being, and quality of life. 43,44 Family exercise and recreational practices represent the family's ability to engage in physical activities to enhance general health through increased range of motion, which can help reduce body fat and the risk of disease. 37 Adults and family members are encouraged to engage in mostly moderate physical activity for at least 30 minutes a day to maintain their health status. Regarding self-direction, a study involving 1,425 students enrolled in Hong Kong Primary levels 1 to 6 showed that self-directed leisure activities during homework-free holidays increase students' confidence, intention, and ability to think freely. This approach is known to improve academic competence and create positive perceptions of fulfilling school tasks. 45 The role of family healthcare provision should be supported by nursing professionals to help families understand parental role construction, provide guidance and support for implementing physical healthcare, promote safe environments, apply theories, principles, and parenting and child programs, develop therapeutic relationships, manage parenting and childcare, promote access to support networks, guide parental figures through the course of life, and use scientific evidence to guide healthcare practices. 46 In Indonesia, such programs exist, such as Bina Keluarga Remaja or Adolescent Family Development. BKR is an organization comprising families with teenagers aged 10 to 24 years, established with the aim of enhancing the knowledge and skills of parents and other family members to nurture and guide adolescent development. Through the empowerment of BKR cadre and family Transforming Healthcare in Low-Resource Settings: a Multidisciplinary Approach Towards Sustainable Solutions health nurses, BKR helps parents and family members fulfill their roles in nurturing and guiding adolescents. 47 However, during social restrictions, programs like these were curtailed, limiting the ability of nurses to reach schools and implement support programs. Reactivating such collaborative programs is recommended to strengthen the family role in the new normal era. A limitation of this study is the relatively small number of participants, which may reduce the generalizability of the results. It is advisable to include more participants and conduct research at multiple centers to increase the effect size. --- Conclusions The role of the family in healthcare provision significantly impacts various social development tasks among school-age children, particularly during a pandemic when children have more interactions with their immediate family members. Strengthening the family's function can be achieved through collaborations between nursing professionals and family development programs, such as "Bina Keluarga Remaja" or Adolescent Family Development in schools. Collaboration and a holistic approach are key to fostering a healthier and more socially developed generation. Additionally, it is advisable to conduct further analyses to explore differences in the family's role in healthcare and its influence on children's social development in the new normal era. --- Transforming --- the research has received ethical approval from the Health Research Ethics Commission based on ethical certificate No. 335/EC/KEPK-S2/09/2021. During the research, the researcher pays attention to the ethical principles of information to consent, respect for human rights, beneficence and non-maleficence. Participant consent for publication: online informed consent was obtained for anonymized participant information to be published in this article. Funding: this research did not receive external funding. Availability of data and materials: all data generated or analyzed during this study are included in this published article. Acknowledgement: we would like to thankful to Headmaster and teachers of SDN Landungsari 01 Kabupaten Malang for their valuable insights and contributions to this study.

The COVID-19 pandemic has compelled children to embrace remote learning and interact more with their immediate families. However, this increased family interaction may pose a potential risk to children's social development. This study aimed to investigates the relationship between family involvement in healthcare and the social development of children. A cross-sectional correlational design was used, involving reliable questionnaires on family healthcare roles and a school-age child social development questionnaire adapted from the V-SMS (Vineland -Social Maturity Scale). Data were collected from 37 parents of 4th and 5th-grade of elementary school students. The data were analyzed using chi-square tests. The study revealed significant associations between family healthcare roles and children's self-direction ability, decision-making, communication skills, and their abilities to dress, eat, and socialize independently. Negative correlation coefficients for specific variables indicated that family healthcare involvement could potentially hinder school-age children's social development. Further research should consider controlling for various factors affecting family healthcare roles and children's social development. Strengthening the family's role, particularly for those facing challenges in fulfilling various tasks, is crucial. Collaboration programs between nursing and family development, such as "Bina Keluarga Remaja" or Adolescent Family Development in schools, can be beneficial. Additionally, further analysis is needed to understand the impact of family healthcare roles on children's social development during the new-normal era.

Introduction Women of reproductive age who have autoimmune rheumatic diseases [ARDs] such as rheumatoid arthritis [RA] and systemic lupus erythematosus [SLE] have expressed a need to be better supported with making decisions about pregnancy [1][2][3]. Many women with ARDs will have positive pregnancy and parenting outcomes, but there are risks involved that need to be considered, including the potential effects of the disease itself and some disease-modifying treatments on fertility, pregnancy and breastfeeding [1,[4][5][6]. Women with ARDs want their motherhood identities and associated preferences to be taken into account in decisions about their healthcare [2]. In this manuscript, we investigate the development and interplay between illness and motherhood identities of women with ARDs as they navigate the healthcare system. The diagnosis of chronic illness, such as autoimmune rheumatic disease, can lead to changes in an individual's sense of 'self' as the previous identity is replaced by an 'illness identity' once a diagnosis is made and treatment instigated. [7]. An illness identity includes perceptions in relation to physical changes, emotional reactions to the physical changes, cognitive constructions of the illness and the impact this has on the self-perceived future [8]. When women who have an ARD become mothers, they can be challenged by the symptoms associated with their illness such as fatigue, pain and impaired psychological well-being in addition to the usual challenges presented to women without a long-term condition [1,9]. The social category of being a mother is not neutral and holds a particular meaning and significance across time and place. There is an ideology of the expectant mother and the identity she must hold [10]. The identity of a 'moral and good' mother is one that is taken on before birth or even pregnancy and embodies the guiding ethic of 'proper' care of children [11]. In contrast to the 'moral and good mother' is that of the 'immoral and bad mother' [10]. The identity of an 'immoral and bad mother' mother would be that of someone who fails to live up to the social ideal of a mother that has been imposed by societal and cultural expectations [10]. The terms 'moral/immoral', 'good/bad' throughout this manuscript are going to be used to reflect the societal judgements placed on individuals and reflected within referenced theories. They are not as a means to stigmatise. Developing the identity of a mother involves internalising perceived ideals of how women ought to function as a mother, whilst understanding the sometimes contradictory lived reality of mothering [12,13]. Women who do not reach, or struggle to maintain, this ideal identity have been found to feel increasing tension over their shortcomings and have reported negative psychological effects such as feelings of guilt [14,15]. The Self-Discrepancy Theory [16] suggests that the degree of discrepancy between the actual self [the attributes a person possesses], ought self [who a person feels they should or must be], and ideal self [who a person hopes or wishes to be] can lead to negative affect, guilt and shame [17]. The values that people attach to their actual, ought, and ideal identities act as 'self-guides' in influencing behaviour and motivation. Analysis of conversations of a UK online parenting forum suggested that women can feel pressure to be an unrealistic 'selfless' and 'ideal' mother [18]. Women's perceived shortcomings in fulfilling their own and others' expectations of their role as a mother can leave them feeling dissatisfied and questioning their identity [19,20]. In the context of women with ARDs, Self-Discrepency Theory would suggest that if mothers experience a discrepancy between their actual self [a woman managing a painful and potentially physically disabling condition], their ideal self [the mother they wish to be] and their ought self [the mother they feel they should be], this would lead to negative self-evaluations [7,17]. There is a lack of research focusing on exploring the influence of healthcare interactions on women's motherhood and illness identities. Understanding how women's illness and motherhood identities influence, and are influenced by, interactions with healthcare professionals and services is important in designing patient-centred approaches to healthcare. In this study, we use participatory qualitative methods to explore the self-identity of women with ARDs as they consider pregnancy, are pregnant, and/or have young children. --- Methods --- Design Timeline-facilitated qualitative interviews with women with ARDs who took part in the STAR Family Survey [1]. --- Ethical approval Ethical approval for the study, including the consent process, was granted by the Cardiff University School of Medicine Research Ethics Committee on 20/10/16 REF16/56. --- Participants and recruitment Women who had expressed an interest in taking part in an interview were sampled from the STAR Family Survey [1]. The STAR Family Survey was a United Kingdom-based cross-sectional on-line survey [n = 128] focusing on women with ARDs' experiences of pre-conception, pregnancy and early parenting. Interview participants were purposively sampled to broadly achieve equal representation of those who were: thinking about getting pregnant, currently pregnant, or had young children. The aim was to explore the identity in women who were at different stages of starting a family, whilst also managing an ARD. Participants were contacted through e-mail or telephone, based on the contact details provided in the survey. Those who expressed interest in being interviewed were sent a study information pack containing participant information sheet, consent form and stamped return envelope. Participants were provided with a £20 shopping voucher as a thank you for taking part. --- Inclusion criteria Women aged 18-49 years, who had been diagnosed with an ARD [i.e. inflammatory arthritis or auto-immune connective tissue disease for which people would normally be under the care of a rheumatologist]. Women also needed to be planning on becoming pregnant in the next 5 years, currently pregnant or have a child/children who were under the age of 5 years. Participants had completed the STAR Family Study survey [1] and consented to interview. --- Exclusion criteria Women who were under 18 or over 50 years of age and those who had a disease not classified as an ARD [e.g. joint hypermobility, fibromyalgia]. --- Interview procedure A flexible narrative approach using visual methods was used to encourage participants to talk about their lived experiences regarding their experiences of identity, motherhood and healthcare in their own words, focusing on things that were important to them, rather than being guided by researcher-generated topics [21]. Before the interviews, women were sent a resource pack containing an example timeline outline, stationary, paper, stickers, and an information sheet outlining examples of the topics we were interested in covering during the interview. The timelines provided a visual tool which enabled women to map out their journey so far. The timelines were used in interviews to provide cues, prompt discussion, provide an opportunity for self-reflection, and help with building a rapport between participants and the interviewer [22,23]. Preparing a timeline was voluntary and women were encouraged to use their own formats [e.g. notes, diagrams they had drawn] as an alternative if they wished to do so. Interviews were conducted face-to-face or over the telephone. Face-to-face interviews were conducted in the participant's home or at Cardiff University. For more information about the interview materials and procedure please refer to Pell et al [2020] [23]. Interviews were conducted by Denitza Williams, PhD [DW] and Bethan Pell, BSc [BP]. At the time of the interviews, DW was a post-doctoral researcher and BP was a research assistant, both identified as female. Neither interviewer identified as having a long-term condition; DW was a mother to two children, BP did not have any children. The interviewers had no prior relationships with the participants. Informed consent was obtained from participants before the interview [written and verbal]. --- Analysis All interviews were audio-recorded and transcribed verbatim by a professional transcription company. DW conducted the analysis. Data were analysed thematically and analysis was based primarily on social phenomenology [24], focussing on exploring how women made sense of their identity and motherhood choices. The data were analysed using Braun and Clarke's Phases of Thematic Analysis [25] which involved familiarisation, generating initial codes and identifying themes. NViVo10 was used to support the analysis. The protocol did not include double coding of data. Instead, the qualitative research team [DW, BP, AG, RP] held regular data analysis meetings to discuss the themes arising from the data, and the development of the coding framework. Each member of the qualitative research group [RP, AG, DW, BP] contributed their own perspective, relating to disciplinary backgrounds including psychology, sociology, social policy and public health. This approach has been identified as appropriate for enhancing the validity and quality of qualitative research [26]. Analysis was guided by the concept of information power [27] instead of data saturation. --- Results Twenty-two women were interviewed face-to-face [N = 6] or over the telephone [N = 16]. Participant demographic details are reported in Phillips et al [2018] [28]. Interview length ranged from 20 minutes to 70 minutes, with a mean length of 48 minutes. Through thematic analysis, three main themes were identified: prioritisation, discrepancy, and trade-off [see Fig 1]. Each of the main themes and sub-themes will be discussed below. --- Prioritisation of motherhood versus illness identities Women spoke about prioritisation of the disease versus their desire for motherhood. The prioritisation sub-themes of diagnosis and transition were identified. Women were often 'juggling meaningful identities' [29]; their identity as a woman, a mother, and a person with a long term condition. --- Diagnosis: The disease comes first Women often reflected on the diagnosis of their ARD and their respective treatment options, contextualising their experiences in terms of their identity as an individual who was at crisis point and in desperate need of effective disease management. As disease management was the priority at this point, when the decision to use disease modifying drugs that might have an impact on fertility was made, their future motherhood plans were often not discussed. A woman's own self-identity and the prioritised identify by the clinician at that point was of a sick patient in need of treatment, and not a woman who might in the future like to have a child. Clinical practice guidelines can also contribute to this, as they focus firstly on ensuring optional disease control before the possibility of a pregnancy. This generates a direct impact on the prioritised identity at the time of a consultation. --- Yeah, I think we did talk about kind of because of the chemotherapy treatments it was likely that possibly I wouldn't be able to conceive. So it was kind of not brushed under the carpet but just one of those ok this has happened and we're talking about it but she's not in the place where she's thinking about having children and actually it's better to be healthy [P1: 37 years old, 1 child, Systemic Lupus] Transition: More than the disease For women who were diagnosed before becoming a mother, shifts in priority were evident following diagnosis. Women reported that their disease needed to be well-managed to enable them to become a mother. Some women reported concern about the potential effects of their ARD when parenting. Concerns about being able to incorporate children into the 'boundaries of themselves' [30] in terms of the expansion of the self to incorporate children and the associated demands were raised. if we're kind of successful in that [becoming a mother], the impact of having a child to bring up and everything that you need to be able to do with that, am I going to be well enough to do --- that in terms of sitting and you know being able to provide that child with everything it needs in terms of energy and everything else. [P2: 26 years old, no children undisclosed ARD] Women discussed how the decision to have children needed to be planned well ahead of time, because many of the drugs used to treat their ARD have teratogenic effects: --- Yeah I guess with like children it's always been ingrained that it has to be planned and stuff like [P2: 26 years old, no children undisclosed ARD] Women needed to transition by modifying the drugs that they were taking to manage their disease before trying for a family. It was felt that this need to plan ahead was forced upon them due to their ARD and felt that most people do not need to actively transition so far in advance, or even to consult a health professional at all when planning to become a mother. The transition signified a motivational and behavioural change between effective management of disease and actively trying to conceive. The women in our study described this transition as a difficult time, because they had made the decision that they would like to have a child but often felt uncertainty about the effect of a pregnancy on their ARD. --- I've got my own kind of own concerns about you know being at a point where I'm well enough to become pregnant, being well during the pregnancy, can I stay well enough for 9 months that I don't need extra medication that would impact on the pregnancy [P9: 39years old, twins, Inflammatory arthritis] Self-guides are specific representations of the self or others and can include mental representations of valued attributes [31]. Self-guides can originate from diverse sources such as general social norms, particularly if the guides have relevance in the terms of the life experiences of the women [31]. Here, the values attached to women's illness identity and motherhood identity are guiding them in different directions, creating tension as they try to decide which values are most important at a given time. --- Self-discrepancy Some women identified discrepancies between their perceptions of motherhood ideals and their actual ability to meet these. Sub-themes included discrepancies in fertility, physical ability and opportunity. Fertility. Women reflected on the impact of their ARD, their perceived fertility and the chances of a pregnancy ending in miscarriage. One woman experienced a miscarriage during a disease flare. She talked about a discrepancy of feelings because she was informed that her poorly managed disease contributed to the miscarriage, she was also aware that women with ARDs are more likely to suffer a miscarriage, but also that miscarriage is common among healthy women. She reported frustration that she had been advised to restart teratogenic medication and to postpone planning for a baby until her disease was fully under control. --- So yeah following the miscarriage I was all for planning to just get pregnant again, but on advice from the rheumatologist and the renal specialist, they you know because I have been so ill that is probably why I miscarried and they have managed to convince me that I'm better off waiting, I'm back on the Micophenolate effectively and I'm in a treatment plan that's not pregnancy friendly to try and get my lupus under control [P15: 32 years of age, no children, Lupus] This signals a loss of a valued mothering identity, an important part of the self [32]. This signifies a discrepancy between the self-guides held by the health professional and the woman in terms of pursuing a course of action that is consistent with her wish to become a mother [31]. Physical ability. The potential impact of physical disability was identified as a factor in motherhood identity. One woman reflected on her social identity in terms of the practical social help she received from a volunteer. The availability of this support helped the mother feel that her children were not missing out on 'normal' childhood due to her long-term condition. This seemed to ease some of the guilt she felt in relation to limited mobility: the only thing I have to say that's helped my kids through it all is in [city] there is the young carer's project they are a charitable organisation [. . .] they take kids out every other week to do things like to the cinema or the park, or they go bowling. Once a year they go on a residential trip as well so it's kind of to give them a break from caring [P14: 34years of age, 3 --- children, Lupus] Women reported that their physical disabilities affected their ability to participate in social group activities relating to motherhood. For example, most women had attended a local mother and baby group, but found it difficult to participate, mainly because activities were often floor based. Women reported feeling embarrassed that they were unable to participate like other mothers, questioning their motherhood identity and highlighting the discrepancy between their perceived ought and actual self. For example, one woman took her child to a baby massage group. The activity was floor based and due to the effect of the ARD on her physical abilities she struggled to get back up from the floor and to pick her baby up: --- If I was on the floor I couldn't get up so the first couple of weeks I went I was sat on a chair leaning over [daughter] but then I just it started to make my back ache and everything else and I think for like I wasn't quite bonding with her because it felt like she was so far away so I thought sod it I'll get on the floor with her but that would mean that I would have to crawl to the chair-hand on to something to get myself up and I would literally have to haul myself up off the floor, but yeah I needed somebody to pick <<daughter1>> up off the floor for me and either carry her to the car, or put her in the sling that I had [. . .] you feel like the odd [one] [P4: 40 years old, 1 child, Dermatamyositis] This event led to a sense of shame and isolation. Motherhood can have an effect on women; their positive and negative qualities can be heightened when reflecting on the mother-child relationship [28]. --- the biggest thing for me is the kids they miss out on so much because of this stupid thing with me. I feel rotten because they have to be no holidays forever and all their pals are going to Spain, or Italy or Portugal or States or holidays and my kids are saying maybe we should go to Norway or Finland because it's so cold and it's not that sunny [it wouldn't cause disease flare] [[crying]] [P14: 34years old, 3 children, Lupus] Women felt that they received discrepancies in recommendation from health professionals, especially for breastfeeding. They were often encouraged to initiate and maintain breastfeeding by midwives and health visitors, even though felt it was not the best thing for themselves or their child at that time due to their condition: --- there is such a drive to push you to breastfeed as it is best for the baby and I accepted it, but I also accept that sometimes the best thing in a research report isn't the best thing for the individual and in our case me and my two little ones it certainly wasn't and I think that was why in terms of the health visitor and the midwife and the breastfeeding advocates they're so hell bent on getting a twin mum to breastfeed, because it's that extra pressure again that they focus so much on that they forget the rest of the situation [P8, 37years old, 1 child, Idiopathic Juvenile Arthritis] Although participants reported that there was a general pressure to continue breastfeeding, some mothers felt that they were pressured to stop breastfeeding by family members or health professionals, so that they could restart required certain disease modifying medications. Many women reported feeling guilty about their inability to breastfeed, or if they had to stop breastfeeding to re-start medication. Reports of feeling guilty seemed to stem in part from healthcare professional and peer attitudes towards formula feeding. Women reported feeling guilty and ashamed of their desire not to breastfeed, inability to breastfeed or their decision to stop breastfeeding: "The breastfeeding advocates had been trying to get me to consider alternative medication and stuff which in an ideal world when we have lots of time maybe that would be a nice thing to do, but when things go downhill drastically you know and I was in a lot of pain and I was struggling to hold 2 babies to breastfeed as well because my joints were sore. . . I marched on and then at 6 weeks I dropped a child. . --- . it took me actually dropping the child I think that made me think she was worth listening to" [P8, 37years old, 1 child, Idiopathic Juvenile Arthritis] Opportunity. One woman reflected on the impact of her physical disability on her actual self as a mother. Her baby had an accident at home which needed hospital treatment. Due to her lack of mobility, the woman could not drive to take her child to the hospital independently and had to rely on a neighbour for help. The woman felt that she did not have the opportunity to behave as a 'normal' mother who could react fast and take immediate action to help her child. This intensified her feelings of self-discrepancy and guilt. --- "[The] gap underneath the bathroom door was quite big, he [son] must've put his toes underneath it and either pulled the door, or pushed it so he chopped off the end of his big toe it was kind of hanging you can see the bone it was so bad. And I was in my jammies, I was so tired. I asked my neighbour to get my other 2 boys ready for school and get them to school in time and I just took a taxi [. . .] because I can't drive [. . .] I was again feeling totally useless because I should have been able to drive him, I should've been able to just get dressed get him in the car and to the hospital and I couldn't put one foot in front of the other I was so tired [. . .] it was horrible you know that feeling that, that you should be able to do more for your kids" [P14: 34years old, 3 children, Lupus] A few women talked about the possibility of adopting a child but felt that other people's perceptions of them due to their ARD limited their opportunities. One woman and her husband felt that managing her ARD, as well as having a successful pregnancy, would be too difficult as altering her medication was likely to have a detrimental impact on her health. She felt that her ability to be a mother was pre-determined by adoption agencies who made judgements in relation to her long-term condition. --- I would have to prove my ability to be a parent so they wouldn't take [stepchildren] at face value, my step children and I would have to go to parenting classes, they'd have to see evidence of me walking, carrying, being able to, they wanted me to have a-you know a pretend baby at home that they can download and things to prove that I was able to meet its needs all throughout the day and night endlessly and it was really full on [husband] and I got furious and I, I was really pissed off at them I remember saying to them you know in this day and age where you can't discriminate against colour, or sexuality you feel you can discriminate against disability you know [P8, 37years old, 1 child, Idiopathic Juvenile Arthritis] Here, there appears to be a discrepancy between women's actual and ideal self-identity and agencies' perceptions of what a mother ought to be. --- Trade-offs between competing values The complex identity changes women go through when planning to become a mother can require multiple types of trade-offs, including disease management, mothering aspirations and mothering ideals. If women's identities have developed in ways in which they are committed to becoming a parent and being part of a family, the threat their illness poses to their ability to do this means they have to revisit that commitment, which can lead to uncertainty and psychological distress. At this stage, women needed to reflect on their goals and values and what the future holds for them as well as their ability to fulfil their perceived identity of a 'good' mother [10]. This can sometimes be influenced by perceived or experienced ableist societal judgments placed upon women who at that time do not possess the socially described 'essential' criteria for motherhood [33,34]. Women reflected on their pre-conception journey and the transitions that needed to be made in terms of their identity as a person with a disease that needed to be managed and as a woman who wanted to become a mother. It was often felt that women needed to balance a trade-off between the self as a patient in need of medication to control their ARD [the actual self] and the self as an individual who is well enough to be a mother to meet their own expectations and that of society about mothering [the ought self]. However, the 'ideal' self is also present here in the sense that women's own goals and aspirations, which might be at odds with the 'ought self', that is what they think others expect of them. Planning a family involved a balance of clinician advice about the trade-offs associated with the management of an ARD and maternal desire. They had to balance multiple identities and expected roles, that of a woman who has a long-term condition and that of a woman who wants to have a child or to expand her family. Some women felt that their motherhood identity was being threatened by the management of their ARD and healthcare communication: --- So they then decided they wanted to start me on floudamide I think it's called which was a disease modifying thing but they told me I couldn't pregnant on it and that the likelihood of me taking it would affect my fertility. So at that point we said well hang on a minute you know I'm only sort of 25, 26, don't really want to start taking stuff that going to affect my fertility, maybe we should just put up with stuff the way it is, try and get pregnant, have another baby because we always knew we wanted a second child then go on the drugs at that point if its needs it. [P3, 34years old, 2 children, Vasculitis] Others spoke of specific trade-offs in relation to the management of their disease and their desire to have children. One woman reflected on how she had traded optimum disease management with the chance of conceiving a baby despite an active disease flare. For her, the priority at that time was to try to ensure that she had a chance of having a child, and her clinical team seemed to be supportive by putting in place necessary referrals: things were flaring back up and they said well look, what we can do is we can do a referral to the fertility clinic, they can take a look and see if there's anything specifically wrong that's stopping it you know you may be eligible for IVF or you know with a discussion with them it might be that they're able to take your eggs and freeze them so that we can start the medication and do it that way. [P3, 34 years old, 2 children, Vasculitis] Women with ARDs who had successful pregnancies felt that they were "lucky" because they identified themselves at a higher risk of complications on their journey to motherhood compared to women who do not have a long-term condition. They spoke of trade-offs in relation to miscarriage risk and the chance of a successful pregnancy: I am very very grateful for my two children because there's a lot of people who really suffer and suffer many miscarriages and stuff whereas I have only suffered one miscarriage in my life, a lot of people have suffered a lot worse and haven't been able to have children or have had their fertility robbed because of the medications they've been on and stuff. [P3, 34 years old, 2 children, Vasculitis] --- Discussion In this study, we set out to investigate the development and interplay of women's self-identity as a person with an ARD and as a mother as they navigated the healthcare system during preconception planning, pregnancy, and early parenting. Prioritisation of illness versus motherhood identities, self-discrepancy, and trade-offs between competing values were identified as major themes in the discussion surrounding women's experiences. The difficulties in balancing multiple identities in healthcare encounters was evident, both for the women and for the healthcare professionals. Women seemed to use 'self-guides' [31] as a reference for priority setting in a dynamic process that shifted as their level of disease activity altered and as their motherhood identity became more or less of a focus at a given point in time. Women often felt that they needed to balance their 'actual' self [individual with a long-term condition] with their perceptions of the 'ought' self [individual well enough to meet their own as well as society's expectations of being a mother] as well as perceptions of the 'ideal self' [women's own goals and aspirations]. Self-guides became at odds with each other when the action that was needed to manage disease well conflicted with the action required to become a mother or to be a 'good mother' as perceived by the individual. Motherhood identities were also sometimes in conflict with medical recommendations in relation to achieving optimum disease management. When these discrepancies in identity and associated self-guides came into conflict [16], women reported feelings of anxiety, shame, guilt, and frustration. Our findings highlighted the need to consider women's multiple and sometimes conflicting identities, with their associated goals and values, in providing patient-centred healthcare. It is possible that clinical guidelines and recommendations for optimum disease can have a direct impact on identities felt my women and it is therefore important that women's goals and identities as an individual are explored. A person-centred care approach can help facilitate this. It has been argued that a mother's pregnancy and long-term condition must be viewed as separate but co-existing entities [35]. This suggests that women view their identities separately, with the identity as a mother working separately to the identity as someone who has a long-term illness. Women's illness and motherhood identities did not present in isolation in our study, but rather were intertwined, highlighting the need for holistic patient-centred care that supports women with the complex and emotive decisions that they need to make. The need for holistic healthcare is reflected in the ethos of the socio-ecological model of health [36] which understands health to be defined more broadly by physical as well as mental and social wellbeing. The socio-ecological model defines health as a broader concept than the active disease status of a person. It can be argued that women's individual and social goals including motherhood identity and reproductive decision making are all part of the overall 'health' of the individual and should be viewed holistically. Previous research has found that the accommodation and recognition of identity shifts in healthcare communication are imperative for supportive patient decision-making [37]. A patient-centred communication style can facilitate such discussions in consultations, placing 'what matters most to the patient' first. Person-centred care encompasses 'Shared decisionmaking' a process where health professionals provide understandable information, discuss the pros and cons associated with different treatment options, and fully involve patients in treatment decisions, considering their personal circumstances and preferences [38,39]. This enables patients to make more informed decisions that align with their preferences and thus fit better with their identity, become more active and empowered in their own healthcare, to have better relationships with their health care professionals, and to feel more satisfied with the choices that they make [40]. There is strong policy support for the routine use of shared decision-making [SDM] in clinical practice [41,42]. The National Institute for Health and Care Excellence recently published guidance for the incorporation of SDM within healthcare [43]. To facilitate a person-centred approach to pre-conception decision making for women with an ARD, changes at individual and organisational level are needed [29] to create supportive and holistic health care policies. Recommendations from this study include: investment in shared decision-making training for clinical staff, incorporation of a broader focus on reproductive and pre-conception health options in consultations and raised awareness of how conversations can impact on women's identities and reproductive goals. Adopting SDM in healthcare teams would enable both illness and motherhood identities and associated goals and values to be considered by women and their healthcare teams so that women's preferences and priorities can be better incorporated in healthcare decisions. --- Strengths and limitations This study explored women's perceived illness and motherhood identities using in-depth participant centred interviews. This enabled us to understand the way in which women perceived their motherhood and illness identities and the way this influenced motivation and behaviour in relation to the management of disease and becoming [or being] a mother. The women who took part in our interviews were more highly educated and more likely to be employed than the overall STAR Family survey participants, which is indicative of a sampling bias that should be taken into consideration when generalising from the findings. Furthermore, this study was conducted before the global COVID-19 pandemic, which saw significant shifts in the delivery of healthcare, with modifications to services including remote consultations for non-urgent healthcare appointments, and significant changes in the provision of maternity services [44]. Therefore, findings might not be applicable to all women with ARDs or other long-term conditions. Further research should explore experiences in those under-represented in this study and the changes during the COVID pandemic which might have impacted on healthcare communication. --- Conclusion This study explored self-identity during pre-conception, pregnancy and early parenting in women with an autoimmune rheumatic disease related to healthcare communication and decision-making. Women's illness and motherhood aspirations were key parts of the formation of their self-identity. Women were balancing and shifting through intertwined illness and mothering identities, which sometimes came into direct conflict with each other. It is imperative that both illness and motherhood identities are acknowledged and explored during healthcare encounters to facilitate preference-sensitive healthcare decisions. A person-centred approach to healthcare communication would facilitate this by encouraging dialogue between healthcare professionals and women about preferences associated with treatment decisionmaking that incorporate both their illness and mothering identities. Additional research will be required to develop and test the feasibility and acceptability of a person-centred intervention to be used in healthcare encounters. --- The data are all contained within the paper. ---

Women of reproductive age who have autoimmune rheumatic diseases [ARDs] have expressed a need to be better supported with making decisions about pregnancy. Women with ARDs want their motherhood identities and associated preferences to be taken into account in decisions about their healthcare. The aim of this study was to explore the interplay between illness and motherhood identities of women with ARDs during preconception decision making.Timeline-facilitated qualitative interviews with women diagnosed with an ARD [18-49 years old]. Participants were purposively sampled based on the following three criteria: thinking about getting pregnant, currently pregnant, or had young children. Interviews were thematically analysed.Twenty-two women were interviewed face-to-face [N = 6] or over the telephone [N = 16]. Interview length ranged from 20 minutes to 70 minutes, with a mean length of 48 minutes. Three main themes were identified: prioritisation, discrepancy, and trade-off. Difficulties in balancing multiple identities in healthcare encounters were reported. Women used 'selfguides' as a reference for priority setting in a dynamic process that shifted as their level of disease activity altered and as their motherhood identity became more or less of a focus at a

can "expose" poorer quality care and lead to deleterious or disparate outcomes. These findings, however, do not imply that addressing this discordance necessarily entails slotting any and all Latinx patients into the Spanish language administrative rubric. Not all Latinx speak Spanish as their first or primary language or even speak it at all . 6 Even those who do speak Spanish may not speak the specific variety of the language assumed by the institution. This state of affairs further compounds the extent of linguistic disadvantage experienced by Latinx individuals seeking care. To take another example, the perceived ubiquity of Spanish-a spoken language-can also marginalize Deaf and hard-of-hearing Latinx, who face, alongside the numerous socioeconomic and psychosocial factors involved in acculturation, "additional factors unique to their situation, such as community attitudes about disability, multilingualism , multicultural values, exposure to multiracial communities, limited availability of mediators , and family issues." 7 Their linguistic repertoires will necessarily be reflective of their individual linguistic histories and experiences with regard to, for example, access to sign language in country of origin, the specific sign language of that country, opportunities to acquire multimodality and multilingualism in signed and spoken languages, and so on. 7 These complex linguistic identities-and the divergent communicative preferences and needs that arise from them-highlight the factual error of equating Latinx with Spanish speaking. --- Monolingual Ideology as a Source of Harm The underlying multidimensional sociolinguistic diversity of US Latinx can serve as entrée to a linguistic state of affairs often overlooked by busy clinicians, overwhelmed patients, and harried administrators, many of whom are rarely trained specifically in navigating multilingual and multimodal realities arising from linguistic difference. US hospitals and health care systems are commonly constructed on a presupposition of a monolingual linguistic culture, dominated by the unrivaled power of English. Such administrative logic, particularly when coupled with monetary incentives geared towards serving well-off patients, means that the complex linguistic heterogeneity of US Latinx is very rarely recognized in the clinical setting. Such administrative logic cannot be understood as purely instrumental and value neutral, however. Rather, it is rooted in a distinct ideology in which language is viewed as an autonomous entity used for conveying information. 8 Linguistic difference is thus conceived as a problem, reflected in its being labeled as a barrier, because such difference introduces potential errors in information transmission and resulting inefficiencies. 8 These beliefs constitute a type of linguistic ideology according to which the perceived problem of language needs to be solved through intervention, a view that is often linked to the social and cultural subordination of the less linguistically powerful. 8 This orientation affects the care of linguistically heterogeneous Latinx in the United States in envisaging language difference primarily as a problem that requires a solution, but it is far less concerned with the power relations that define the parameters of the problem or the terms of its proposed solutions. --- Linguistic Difference in Health Communications Linguistic discordance is not merely affected by linguistic difference in the narrow sense of the difference between language A and language B. Even when both languages are understood in more inclusive terms , the existence of a language barrier is further compounded by additional layers of critical difference. For example, there is a gap in epistemic authority and credibility between the globally powerful languages of science and technology that dominate contemporary transnational knowledge development and production and less powerful languages in which conceptions of health, illness, and wellbeing are grounded in local and traditional knowledge systems. There are also other layers of linguistic difference. Linguistic differences might arise in the care of individuals with complex communication needs . Still another vitally important layer pertains to the difference between habitual language and disrupted linguistic agency . In another example, the complex linguistic response to trauma, in terms of both experience and narration, may drastically affect a person's habitual interaction with linguistic interlocutors, as manifested in "his or her inclination to learn languages, to use, retain, or abandon a particular language, or to take refuge in silence." 10 Health care practitioners-and institutions more broadly-should not lose sight of these additional layers in planning and evaluating the care they provide to linguistic minorities. The inadequacy of slotting Latinx patients into neat language boxes, or the tendency to see language only as a barrier rather than as intrinsically linked to individual and societal complexity, is even more relevant in the current Covid-associated syndemic. 11 In the context of the clinical experience of the first author , a common assumption of administrative logic is that merely replacing English with Spanish is enough to fully meet the communicative needs of Latinx during the Covid-19 pandemic, an assumption that, in addition to overlooking Latinx' linguistic heterogeneity, fails to acknowledge the intersection of economic, social, political, and juridical constraints on this community. Cultivating Linguistic Humility These considerations illustrate just how complex and multifaceted health communication is even in supposedly linguistically concordant settings, let alone in their discordant counterparts. They also emphasize the need to keep in mind that interpretation-even in best-case scenarios when it is available, funded, and certified-is better envisaged not as a fix to communicative barriers but rather as a tool that needs to be properly utilized and whose benefits as well limitations need to be understood. 12 The awareness that interpreters are not merely language conduits between language A and language B and the need to reconcile and calibrate the different perceptions of all parties involved in interpreted health communication 13 constitutes another important insight into the crucial shortcoming of an uncritical conception of language that often underlies the administrative logic of health care institutions. It is pivotal to highlight that an uncritical conception of language is not merely a theoretical or intellectual preoccupation. Rather, it has a clear and immediate effect on the efficient utilization of various resources and services provided with the aim of delivering equal care to patients of all linguistic profiles and identities. Concerns about the cost of such services , given real-world resource limitations, rarely address the problem of their inefficient utilization by clinicians and administrators due to lack of sufficient language awareness. For example, an administrator might incorrectly infer on the basis of minimal input that a patient in fact speaks English and therefore dismisses the interpreter. Or a clinician might request sign language interpretation to communicate with a Deaf patient without being aware that there exists more than a single sign language and that different sign languages are not necessarily any more mutually intelligible than different spoken languages are. We thus suggest that cost-related concerns over linguistic accessibility in health care ought to be consideredand perhaps also reevaluated-in relation to the question of whether or not health care institutions are in fact sufficiently language aware to adequately utilize the existing range of resources 14 and how to assess ongoing and future needs. Relatedly, we also wish to emphasize that attempts to "fix the fix" on the part of practitioners with some competency in the patient's language, by cutting out the middleman altogether, should be regarded with caution. Such shortcuts incur the risk of false fluency and the encouragement of an institutional linguistic culture of "getting by" at the expense of certified interpretation. 15 This tendency is reinforced by the expectation on the part of some academic medical centers that in-person interpretation is a dispensable luxury, completely replaceable by video or audio equivalents. Our point is not to completely discourage practitioners from using their linguistic competency in such encounters. Indeed, we are mindful of the risk that an institutional linguistic culture centered on formal assessment and credentials might result in appropriating language from noncertified bilingual practitioners and minimizing the contribution of domestic speakers of the language. 8 Rather, we seek to highlight the importance of clinicians developing a sense of linguistic epistemic humility based on "an attitude of awareness … of their own linguistic epistemic capacities … the recognition of their limitations, and the active search for sources outside one's own linguistic epistemic capacities to help overcome them." 16 In the context of health care provision for US Latinx, linguistic epistemic humility entails not simply an honest self-assessment of one's Spanish competency, but also the capacity to assess the degree of relevance of Spanish to begin with, given the linguistic heterogeneity among Latinx patients, as described above. --- Conclusion Making health care institutions more linguistically inclusive for Latinx patients entails challenging a monolingual and Anglo-centric administrative logic. However, doing so requires more than simply equating Latinx with Spanish speaking. A better-informed understanding of the spoken and signed linguistic heterogeneity of US Latinx is fundamental for a more equitable health care delivery committed to patients' equal linguistic dignity. Power often gives rise to an unexamined assumption of sufficient knowledge, 17 whose presence can be detected in the linguistic ideologies that underlie the administrative logic of present-day health care systems and institutions. Challenging the assumption of sufficient knowledge by pursuing enhanced linguistic understanding can therefore contribute significantly to addressing present-day US Latinx health inequities.

Language is a social determinant of health, no less so in the case of Latinx persons, who make up the second largest ethnic group in the United States. In US health care, language and linguistic difference are often conceived in discrete, instrumental, and monolithic terms. This article characterizes this conception of language as administrative logic, which is in sharp contrast to language conceived as a richly complex, heterogeneous, communally lived human experience. This article emphasizes the importance of system-level language awareness and epistemic humility for promoting equity, as well as the need to avoid toonarrow focus on linguistic assessment.Many residents of the United States, including health care workers, are aware that Latinx make up the second largest ethnic/racial minority group in the United States. Less often realized is the heterogeneity that underlies this classification. 1 (We use the label Latinx throughout primarily for reasons of thematic and conceptual coherence with the rest of the contributions in the thematic issue, while recognizing the social significance and political salience of relevant competing or interrelated notions, such as Latinidad and Indigenidad.) In what follows, we outline the implications of the sociolinguistic context of US Latinx for their clinical care. More specifically, we consider the ways in which the quality and equity of care provided to US Latinx are affected by a profound tension between the Anglo-centric and often monolingual administrative logic of health care institutions, on the one hand, and the complexities of Latinx' linguistic lived experience of health and illness, on the other. More specifically, we offer a critique of this administrative logic in the context of health care, and we examine how a less regulative-centered conception of language and linguistic agency can better serve health care policies and practices committed to equality, diversity, and inclusion and likewise guide their development, implementation, and assessment.The goal of equitable health care delivery to US Latinx-and to other US demographic groups-has a critical linguistic dimension. Most studies show that linguistically concordant care improves health outcomes. 2,3 Linguistically discordant care often leads to patient dissatisfaction, lower quality of care, inappropriate follow-up, and worse health outcomes. 4,5 That is, care in a language other than that preferred by the patient

An important facet of the complexity in studying religiosity and family well-being is the extent to which adolescents agree or disagree with their families on matters of religion. As Pearce and Haynie pointed out, children may not always share their parents' religious beliefs, values, and practices, and that disconnect may impact the quality of family relationships. The past decades have seen substantial research interest in religious agreement and family well-being, with the general pattern of findings indicating that religious concordance is related to positive family relationships, while religious disagreement is related to poorer relationships . Two recent studies using data from the National Longitudinal Study of Adolescent Health have extended this initial focus on family religious concordance by exploring not just whether adolescents and parents are concordant or discordant in religiosity, but in what direction any dissimilarity occurs, that is, whether it is the parents or the adolescents who are more religious. In a study of adolescent delinquency and religiosity , Pearce and Haynie found the lowest rates of delinquency among girls who were similar to their mothers in high religiosity, slightly higher rates among girls who were similar to their mothers in low religiosity, and the highest rates among girls who were religiously dissimilar from their mothers, regardless of whether it was the mother or child who was more religious. However, in an examination of relationship quality, Stokes and Regnerus found religious discordance only predicted lower adolescent/parent relationship quality when adolescents reported lower religious salience than their parents. While valuable, the contributions of these studies are limited by the methodological treatment of the religious concordance variables. In the Pearce and Haynie study, the items tapping adolescents' and parents' salience and attendance were averaged together into religiosity variables for each party, which were then combined in an interaction term. Thus, an adolescent with extremely high salience and extremely low attendance would have a similar religiosity score to one with moderate salience and attendance. In the Stokes and Regnerus study, the items tapping adolescents' and parents' salience, attendance, and identification/denomination were considered individually. We argue here that, rather than use these variables separately or in an additive fashion, it would be more useful to explore how these variables might cluster together to form certain patterns of agreement or disagreement. Such an approach more closely aligns with how the multiple dimensions of religiosity are experienced in everyday life; that is, they don't operate individually or even additively, but they cluster together in more interactive and dynamic ways to form unique and meaningful patterns-patterns that may be linked to family relations. In other words, a given adolescent experiences how important religion is to her in conjunction with other dimensions of religiosity, such as her participation in a religious community, and she experiences these dimensions alongside her parents' own experience of those dimensions. This approach also accounts for the possibility that adolescents may not experience religious concordance or discordance with parents across the board, but rather may agree on some dimensions and disagree on others. --- The Current Study In the current study, we use data from Add Health and employ an analytic technique, latent class analysis , that allows us to examine how adolescents' and mothers' relative concordance or discordance on three dimensions of religiosity cluster together into classes or profiles. Research Question 1 examines what profiles of religious concordance or discordance exist between adolescents and mothers. Research Question 2 asks whether these identified profiles differ on family relations, as well as on certain demographic characteristics, including religious affiliation, race/ethnicity, urbanicity, geographical region, age of adolescent, gender of adolescent, and maternal education level. The work described in this research note is part of an emerging focus in religiosity research to discover profiles of religiosity experiences among individuals. Smith and Denton described three types of overall religiosity among adolescents in the National Study of Youth and Religion . They also described types of religious service attenders . Smith and Denton's work did not involve empirically derived types, but rather was based on "common understandings" in the culture of how religiosity might operate for teenagers, which were then confirmed by the NSYR data. More recently, Pearce and Denton used latent cluster analysis with NSYR data and identified five groups of teenagers . In a study of ethnic and gender variations in religiosity, Jones and colleagues identified five profiles of religious involvement among young adults . To this emerging focus on profiles of religiosity experiences, we add the first study that we know of to use latent class analysis to form dyadic profiles of religious concordance or discordance in families. --- Methods --- Sample Data for the present study come from the Wave I in-home interviews of the National Longitudinal Study of Adolescent Health , a longitudinal study of the characteristics, behavior, and environments of adolescents and their health and achievement in adulthood. The Add Health sample is nationally representative of U.S. adolescents who were in grades 7-12 during the 1994-1995 school year. Our study sample consists of adolescents in the core sample who had never been married and whose mother completed the parent survey. The sample for the latent class analysis consists of 14,202 adolescent/mother pairs. The analysis sample for Research Question 2 is smaller. For these analyses, pairs were assigned membership in a "best-fitting" latent class using Bayesian posterior probability estimates. With this technique, not all pairs can be definitively identified as belonging to a single best fitting class. As a result, approximately 2000 pairs were dropped. --- Measures Demographic variables consist of gender, age, race/ethnicity, mother's education , family type, urbanicity of residence, and geographical region. Other control variables include the following adolescent-reported variables: neighborhood resources , neighborhood safety , and peer relations. One multiple-item scale, self esteem, is also used as a control, in recognition of its established association with teen/parent relations . This scale, consisting of four items, has adequate internal consistency . We also examine three indicators of adolescents' relations with mothers. We examine adolescents' reports of the number of the following shared activities with mothers in the past four weeks: going shopping; playing a sport; attending a movie, play, museum, concert, sports event; working on a school project. Responses range from 0 to 5 with an overall mean of 1.58 . They were also asked about their overall satisfaction with communication with their mother. Responses range from 1 to 5 with an overall mean of 4.06 . Adolescents were asked about relationship quality with their mothers, including how close they feel to their mothers, how much they think their mothers care about them, how much they agree that their mothers are warm and loving towards them, and how satisfied they are overall with their relationship with their mothers. This scale is coded such that higher values represent higher relationship quality. The variable ranges from 1 to 5 with an overall mean of 4.33 . This scale has a high internal consistency coefficient . As described earlier, we use three variables for the concordance/discordance profiles : religious salience, frequency of prayer, and attendance at religious services. In Add Health, adolescents and mothers who actively claimed no religion were skipped out of the section containing these salience, prayer, and attendance questions. Thus, pairs containing these Add Health participants are omitted from this analysis. Religious salience is assessed by mothers' and adolescents' reports of how important religion is to them . Frequency of prayer is assessed by asking mothers and adolescents "How often do you pray?" . Attendance at religious services is assessed by asking mothers and adolescents how often they attended religious services in the past 12 months . We consider religious affiliation as a descriptive variable, rather than a dimension along which to base the religiosity profiles; whether or not the mother/adolescent pair claim the same religious affiliation was coded as yes/no. --- Results --- Research Question 1 The first research question examined what profiles of religious concordance exist between adolescents and mothers. Construction of Religiosity Profiles-Latent class analysis was performed on adolescent and mother reports of religious salience, frequency of prayer, and attendance at religious services. 1 Latent class analysis is conceptually equivalent to conventional cluster analysis . Rather than examine characteristics of variables, cluster analysis and latent class analysis examine the characteristics of individuals by grouping individuals according to their responses across variables. It is important to note that a mother/adolescent pair's profile may not exactly fit any of the constructed profiles. It is helpful to think of pairs as having a probability of belonging to each of the profiles, with the highest probability associated with the profile that best captures the essence of the pair's responses. In order to assure the generalizability of the findings, we conducted latent class analyses in two random halves of the sample: selecting among alternative models using data from the first half-sample and replicating the solution with data from the second half-sample. We considered profile solutions that contained 2-, 3-, 4-, 5-, and 6-latent profiles. 2 In this study, the BIC criterion pointed to the 6-profile solution as being optimal . 3 The LRT was not helpful in this application in that all of the alternative model solutions represented successively significant improvements over the previous models without reaching a point of diminishing returns. The 2-profile solution had the best classification certainty -not surprising since there is little room for misclassification in a 2-profile model. The classification certainty was not appreciably degraded with higherlevel profile solutions. The estimated proportion of the population belonging to each profile showed that profiles with fewer than 10% did not appear until the 6-profile solution. Further, the 6-profile solution showed signs of estimation difficulty . While the selection results were good for nearly all solutions, we selected the 5-profile solution. Our replication sample results show the stability of the 5-profile solution with respect to overall fit. An inspection of the parameter estimates in the estimation and replication samples for the 5-profile solution showed that the two were virtually identical. The full sample 5-profile solution was retained for all further analyses. 1 Adolescent age and gender are modeled as covariate effects on each profile indicator. The effects of these variables on each outcome can be obtained from the first author. 2 Model fit statistics for these solutions are available from the first author by request. 3 In selecting the optimal number of latent profiles, we compared the fit of 2-, 3-, 4-, 5-, and 6-profile solutions using the BIC statistic and the Lo-Mendell-Rubin likelihood ratio test , as recommended by Nylund et al. . We also considered evidence of estimation difficulties , the loglikelihood value for each solution, and the relative certainty of classification for each latent profile solution. The entropy statistic, a summary statistic reflecting the overall classification certainty, was also considered. This value is standardized; entropy values closer to 1 indicate a better fit. Description of Religiosity Profiles-Analyses uncovered five classes or profiles of mother-adolescent dyads: one highly concordant profile, two highly discordant profiles, and two profiles of mixed concordance and discordance. The expected responses of mother and adolescent for each class are given in Table 2. We quantified the overall degree of discordance in each profile by creating z-scores for each estimated mother and adolescent response to each of the three religiosity indicator. The z-scores for adolescents were then subtracted from the z-scores for mothers and the absolute value of this difference was taken. The absolute mother-adolescent differences for each indicator and across all three indicators are shown in Fig. 1. Profile 1: Fully Affiliated Believers : In the one highly concordant profile , both the mother and the adolescent reported high levels of importance, high frequency of prayer, and high frequency of attendance at religious services. For ease of reference, we refer to members of this profile as Fully Affiliated Believers. 4 There were two highly discordant profiles, both characterized by mothers with higher religiosity scores than their adolescent offspring, but with very different patterns. Profile 2: Ambivalent Mothers/Unengaged Adolescents : Profile 2 contains mothers who reported religion to be somewhere between "fairly unimportant" and "fairly important," they prayed "at least once a week," and they reported some, though infrequent, religious service attendance. Their children, however, had the lowest religiosity scores in the study. Religion was "not important at all" to them, their average prayer frequency fell between "never" and "less than once a month," and they did not attend religious services. For ease of reference, we refer to this profile as Ambivalent Mothers/Unengaged Adolescents. Profile 3: Mothers Higher : A second highly discordant profile was Profile 3 . In this profile, mothers were highly religious, reporting the highest possible responses for the importance of religion in their lives, frequency of prayer, and frequency of attendance at religious services. Their offspring, however, reported less attendance, less frequent prayer, and much lower levels of salience: average levels of "fairly unimportant" as compared to their mothers' "very important." For ease of reference, we name this profile Mothers Higher. There were two profiles of mixed concordance and discordance. Profile 4: Partially Affiliated Believers : Mothers in Profile 4 , reported religion as between fairly and very important, and prayed every day, but they reported infrequent attendance at services . Their children reported religion as "very important" and they prayed slightly more than once a week. However, they attended services more frequently than did their parents: 4 Just as mother/adolescent pairs might not perfectly fit the profiles to which they belong, but rather have a higher probability of belonging to that profile than to the other profiles, the names for these profiles may not be perfectly descriptive of each dyad in the profile . Rather, the profile names are used for illustrative purposes and to guide the reader. more than once a month, but less than once a week. Thus, we name members of this mixed profile Partially Affiliated Believers. Profile 5: Mothers and Adolescents as Seekers : In Profile 5 , mothers reported religion as close to "fairly important" and they reported praying more than once a week. However, they reported the lowest religious service attendance of study participants claiming a religion-almost never. Their children rated religion as "fairly unimportant," reported praying almost once a month, and reported attending services less than once a month. A note of clarification is in order when interpreting profile 5. In profiles 2, 3, and 5, mothers are higher than adolescents in most or all religiosity indicators. However, profile 3 is distinct from profiles 2 and 5 because the mothers are higher than adolescents in all indicators and because the overall discordance between mothers and adolescents is high. A chance for confusion, we believe, is strongest when distinguishing between the discordant profile 2 and the somewhat discordant profile 5 . The differences between the two profiles are as follows: there is less overall discordance between mothers and adolescents in profile 5 than in profile 2, the profile 5 adolescents are less disengaged on all indicators than are those in profile 2, and the profile 5 mothers report higher levels of salience and prayer than do profile 2 mothers. Thus, we name profile 5 Mothers and Adolescents as Seekers to denote that these adolescents and mothers, to some degree, share a stance characterized by some engagement with religiosity. --- Research Question 2 Research Question 2 examined whether these profiles differ with respect to family dynamics as well as demographic variables. Turning first to demographics, we found profile group differences in gender, age, race/ethnicity, mother's education, family structure, urbanicity of residence, and region, as shown in Table 3. Two profiles contained a significantly greater proportion of males and were older than the other profiles. The Ambivalent Mothers and Mothers and Adolescents as Seekers profiles contained the greatest proportion of White participants and the fewest Black participants, while the Mothers Higher profile contained the greatest proportion of Hispanic participants. The Partially Affiliated Believers profile had the lowest level of mother's education and the highest proportion of single parent-headed households. Fully Affiliated Believers were least likely to live in an urban area. Mother/Adolescent pairs in three profiles were most likely to live in the South and least likely to live in the Northeast, Ambivalent Mothers/ Unengaged Adolescents were most likely to live in the Midwest and Northeast, and Mothers and Adolescents as Seekers were most likely to live in the Midwest and South. The Fully Affiliated Believers and Mothers Higher profiles had the greatest proportion of motheradolescent pairs reporting belonging to the same general denominational category. 5 With regard to family relations, we examined associations between the profiles of religiosity and the adolescents' reports of relations with mothers . These associations were tested by a series of regression models. 6 Estimated marginal means and post hoc pairwise comparisons are given in Table 4. Shared Activities-The full regression model accounts for nearly 8% of the variability in shared activities [F = 1.150, p = .001] although the contribution of profile class is minor . For this outcome, five post hoc profile comparisons were significant at the p < .05 level: adolescents in Mothers Higher and Ambivalent Mothers/ Unengaged Adolescents report significantly fewer shared activities than those in Fully Affiliated Believers and Partially Affiliated Believers , and those in Mothers Higher report significantly fewer shared activities than those in Ambivalent Mothers . These effects are very small , ranging from .01 to .09. Satisfaction with Communication-Our model explains roughly 13% of the variability in satisfaction with parent-child communication [F = 1.345, p < .001]. Profile membership, while statistically significant, explains less than 1% of this variability. Pairwise comparisons reveal significant profile group differences across all but two pairs. Adolescents in Fully Affiliated Believers and Partially Affiliated Believers report similar levels of satisfaction , levels which are significantly higher than all other profiles. Adolescents in Ambivalent Mothers/Unengaged Adolescents and Mothers Higher are statistically indistinguishable and have significantly lower levels of satisfaction than those in all other profiles. Adolescents in Ambivalent Mothers reported satisfaction levels between the two sets of profiles . Effects range from .01 to .23 . Relationship Quality-This model explains 19% of the variance in relationship quality [F = 1.216, p < .001] with 2% of the variance attributable to profile membership. Relationship quality is highest for adolescents in Fully Affiliated Believers . Adolescents in all other profiles perceived significantly lower relationship quality . Effects range from .08 to .43 . For relationship quality, the distinction between profiles is clear, in that all pairwise comparisons are significantly different. --- Discussion Building upon recent research identifying different profiles of religiosity at the level of the individual , we used 5 Denominational categories, following Regnerus , were: Conservative Protestant, Mainline Protestant, Black Protestant, Roman Catholic, Jewish, Mormon/LDS, and Other. Information on coding Add Health responses into these categories can be obtained from the first author. 6 Space considerations prevented a full reporting of these models. Tables of regression results are available on request. data from the National Longitudinal Study of Adolescent Health to identify dyadic profiles of religious concordance or discordance between 14,202 adolescents and their mothers. Using latent class analysis, we identified five distinct profiles of religious concordance and discordance among adolescents and their mothers participating in the Add Health Study: one concordant, two discordant, and two of mixed concordance/discordance. We were interested in how the profiles might represent different patterns of family dynamics, and we found that the profiles did distinguish between various levels of adolescent/mother relations. Adolescents in the concordant Fully Affiliated Believers profile and those in the mixed Partially Affiliated Believers profile reported significantly better relations with mothers than did youth in the other profiles. They did more things with their mothers, were more satisfied with the way they communicated, and felt closer to their mothers and more cared for and loved by them. In contrast, adolescents in the two discordant profiles-Mothers Higher and Ambivalent Mothers/Unengaged Adolescentshad significantly less positive relations with mothers than did other youth. The finding of better relations in the concordant Fully Affiliated Believers profile than in the two discordant profiles is consistent with previous research linking concordance with positive family relations and discordance with poorer relations . However, our use of latent profile analysis allowed for the discovery of some important nuance in the links between religiosity and family relations. It was not concordance alone that was associated with positive family relations, given that the other profile enjoying more positive relations was one of mixed concordance/ discordance. In that profile, there is some discordance, although of a lesser degree than in the two discordant profiles, and both adolescents and mothers report moderate levels of engagement with religiosity. Future studies could examine this interplay between degree of concordance and engagement with religiosity to determine the relative impact of each factor on family relations. That these profiles distinguished between other variables in predictable ways gives us additional confidence that the profiles represent meaningful categories. For example, the Mothers and Adolescents as Seekers profile had the fewest amount of Black participants, consistent with reported higher levels of religiosity among Black Americans Adolescents in the Ambivalent Mothers/Unengaged Adolescents and in the Mother Higher profiles were significantly older than adolescents in the other profiles, perhaps reflecting an established decline in adolescent religious participation with age . Mother/Adolescent pairs in three profiles were most likely to live in the South and least likely to live in the Northeast, while Ambivalent Mothers/Unengaged Adolescents were most likely to live in the Midwest and Northeast; these findings are consistent with regional differences noted elsewhere . Some study limitations are noteworthy. First, it is important to note that membership in the various concordance/discordance profiles explained only a limited amount of variance in the family relations dimensions. However, while the effect sizes of the differences between profiles on these family relations indicators are generally somewhat small, the skewed distribution of the indicators toward more positive relationships would underestimate the standardized size of the differences than if the indicators were normally distributed, as is assumed in the calculation of Cohen's d. Second, due to our cross-sectional analysis, we were unable to test for causal pathways among variables, nor are we able to provide evidence of bidirectional influence. Third, this examination focused only on families with an identified religious affiliation, due to Add Health's practice of not asking religiosity questions of unaffiliated families. While the profiles still capture a broad range of religiosity experiences, future investigations would benefit from exploration of religiosity among nonaffiliated families. Researchers wishing to use latent class analysis are urged to attend to these limitations, and to explore associations between concordance/discordance and other outcomes of interest. They are also urged to pay careful attention to the choice of religiosity variables to include in the analyses. Here, we used three variables in an existing data set, yet other options abound, either in secondary analysis of existing data or in new data collection: beliefs in scriptural inerrancy , intrinsic vs. extrinsic beliefs , frequency of attendance, intensity of beliefs, self-report of how religious one is, comfort received from religion, religious commitment, and orthodoxy . However, it is perhaps well-advised to include a measure of religious salience, because this variable has been shown related to the quality of relationship between parents and adolescents . Discordance between mothers and adolescents within each profile, expressed as the absolute value of the standardized estimated mean scores Descriptive statistics for demographic and analysis variables Difference in profile means of demographic and religiosity characteristics . 45 .45 .56 .54 1v2,1v3,2v4,2v5,3v4,3v5 Age 15.46 15.79 15.76 15.49 15.59 1v2,1v3,1v5,2v4,2v5,3v4,3v5 White .45 .69 .48 .48 .62 1v2,1v5,2v3,2v4,2v5,3v5,4v5 Hispanic .15 .17 .22 .17 .18 1v3,1v4,1v5,2v3,3v4,3v5 Black .32 .07 .19 .27 .12 All Mother's ed 3.65 3.61 3.54 3.34 3.49 1v3,1v4,1v5,2v4,3v4,4v5 Two biological .61 .53 .64 .48 .52 1v2,1v4,1v5,2v3,3v4,3v5,4v5 Single parent .22 .28 .19 .33 .28 1v2,1v4,1v5,2v3,2v4,3v4,3v5,4v5 Stepparent . --- 08 .12 .09 .12 .12 1v2,1v4,1v5,3v4,3v5 . 21 .30 1v2,1v3,1v5,2v3,2v4,3v5,4v5 South .47 .19 .33 .47 .29 1v2,1v3,1v5,2v3,2v4,2v5,3v4,4v5 Northeast .09 .27 .17 .11 .18 1v2,1v3,1v5,2v3,2v4,2v5,3v4,4v5 Same denom a .87 .79 .86 .80 .81 1v2,1v4,1v5,2v3,3v4,3v5 a This comparison was based on a set of denomination categories

This research note describes the use of latent class analysis to examine how three dimensions of religiosity-the importance of religion (religious salience), attendance at religious services, and frequency of prayer-cluster together to form unique profiles. Building upon recent research identifying different profiles of religiosity at the level of the individual, we used data from the National Longitudinal Study of Adolescent Health to identify dyadic profiles of religious concordance or discordance between 14,202 adolescents and their mothers. We identified five profiles: one concordant (27% of sample), two discordant (25% of sample), and two of mixed concordance/discordance (49%). The profiles distinguish between various levels of adolescent/ mother relations, suggesting that they may represent distinct family dynamics. They also distinguish between several variables (race, adolescent age, geographical region) in predictable ways, providing additional demonstration of the categories' meaningfulness.

Introduction Since municipal public baths were introduced in the UK during the 19th century, there have been debates about how much the public should be charged for using these facilities, reflecting tensions between public health ambitions for such services and the need to ensure they are economically viable. 1 More recently, decisions about leisure provision have, arguably, been shaped by similar tensions. A welfare orientated model has positioned leisure as a 'right', with the duty of public organizations to deliver accessible leisure for the local population. 2,3 In contrast, a commercially driven approach is more aligned with business practices that emphasize income generation and deficit reduction. 4 Yet the implications for achieving public health priorities that stem from how local authorities set entrance charges, remains less well understood. Inactivity is of global concern and contributes to several cancers and chronic diseases. 5 In the UK context there is evidence that inactivity levels are 10% higher in most deprived areas compared to the least deprived. 6 The Active Peoples Survey also estimates a participation rate of 26% in lower socioeconomic groups, compared to 39.5% in higher groups. 7 Tackling this scale of physical inactivity in populations is recognized to require action at a range of levels that includes the affordability of participation. 8 Under local government arrangements, LAs in England are the organizations responsible for delivering a range of public services across a locality. In some areas, unitary authorities are responsible for delivering all services across a geographical area but in other areas, this function is split between two tiers. LA leisure services have the potential to improve population level activity, as they provide access to facilities that reach social groups across the life course. The amount that leisure services charge for using facilities is one potential strategy that LAs have at their disposal to tackle inactivity. Reducing or removing charges could contribute to this because cost has been found to act as a barrier to participation for lower income groups. 9,10 As a discretionary service, however, there is a risk that entrance charges could increase as a consequence of austerity because LAs can charge up to the full cost of provision. 11 Between 2009 and 2014 it is estimated that the public subsidy for LA leisure facilities was cut by 32% from £550 million to £375 million in England. 12 Following changes to the organization of the public health function in England, there is also a greater opportunity for public health professionals to work more closely with leisure services to more proactively plan provision in a way that promotes physical activity . 13 Since 2013, upper tier and unitary LAs in England have been granted greater responsibilities for promoting public health and reducing inequalities, with public health departments transferring from the NHS into local government. However, previous studies have identified barriers to partnership working linked to the complexity of intersectoral collaboration as well as political and financial constraints. 14 The research aimed to investigate the components of LA pricing policies, as well as the rationales and approaches shaping how such policies were developed and implemented locally. More specifically it considered: How do goals and approaches to leisure entrance charges vary across LA areas in terms of what they are intended to achieve? What factors influence decisions about the ways that entrance charging policies are developed and implemented? The research described here formed part of a larger mixed method evaluation assessing the health inequalities impacts of LA leisure pricing strategies in the north-west of England. --- Methods --- Recruitment and data collection The fieldwork was located in seven LA areas who formally agreed to take part in the main study. The 2015 Indices of Multiple Deprivation Rank 15 provides an average deprivation summary score for each of the 326 LAs in England. Of the seven areas, five LAs were ranked within the most deprived 25% of all LAs, one area within the 50% most deprived LAs and one area in the least 50% deprived LAs. Each LA contains some of the most deprived neighbourhoods in the country in spite of the range of scores. With the exception of one area, leisure provision in all areas was owned and managed by the LA department. In the other area, the provision was managed by a Community Leisure Trust. Leisure provision in all areas included wet and dry facilities . All services provided outreach activities but these varied in scale. Members of the research team conducted fieldwork between December 2014 and April 2015. A total of 33 informants were recruited, purposefully sampled for their professional roles. Individual or group interviews with leisure professionals included senior strategic leads and service/operational managers. Where group interviews were conducted , these involved the leisure service team from the same LA. Individual interviews with public health professionals included those with strategic or commissioning roles for PA. Two local elected politicians took part in individual interviews in two areas. Written informed consent was sought from participants. Interviews followed a semistructured topic guide and ranged in length from between 40 to 70 min. Both individual and group interviews utilized the same schedule with group participants reminded of confidentiality prior to the discussion. Policy documents were collated from LA webpages and during fieldwork. Researchers utilized documents to identify details of schemes for each LA and as prompts during interviews to explore rationales underpinning schemes. A research advisory network involved leisure and public health teams from the participating LAs. These meetings also enabled the team to track any major changes to schemes during the main study. --- Analysis All interviews were audio recorded and professionally transcribed. Researchers familiarized themselves with the transcripts then extracted data to an analysis framework in Excel structured around deductive headings related to the interview topic guide. Charting and synthesizing data, used in a framework approach to analysis supported a comparative analysis of pricing goals and approaches between sites and professional groups. 16 Framework analysis enables a systematic approach to managing and mapping qualitative data particularly where the aim is to compare within and between cases. Other team members contributed to the analysis and interpretation of data. Emerging findings were discussed with practitioners during workshop sessions. Only the research team had access to transcripts and practitioners were not involved in analysing data. All quotes and excerpts used below to illustrate the results are anonymised to protect participants' identities. Codes provide the following information: LA fieldwork site + Role . Ethics approval was gained from Lancaster University's research ethics committee in December 2014 . --- Results --- Goals of leisure pricing Welfare orientated goals Professionals in all seven areas referred to the LA's responsibility to ensure publically funded services were accessible for the local population. Perspectives on the necessity of affordable entrance charges were most prevalent in LAs where facilities were located in areas characterized by insecure employment and unemployment. Reference was made to 'not wanting to price people out of participation', as these respondents explained. I think the recognition that it's one of the more deprived boroughs in the country so a high proportion of the population on low incomes, very low incomes and at the same time activity levels historically have been very low as well. In saying that it means that we have to make sure our services are universally accessible to people who have got social or economic challenges. Offering activities at a low cost or for free in 'pockets' of deprived neighbourhoods or for particular groups was also described as a means of encouraging participation among groups defined as 'socially excluded' or 'hard to reach': People in leisure believe that leisure can contribute significantly to life to some of these youngsters. That's why there's a lot of things that we are doing that are trying to encourage kids to be involved. Commercially driven goals However, the need to achieve LA budget positions and financial targets was also highlighted as a prominent factor affecting pricing decisions in all participating LAs. The way I see it simply is it's a local authority led business and my ownership as a manager is to operate non-profit really; to break even as a business so it doesn't cost the council and then look at potential profit related to future investment; so that's the way I look at it. In many areas, participants spoke of pricing decisions being affected by funding cuts. Those with strategic decision making roles expressed concerns about the ability to maintain affordable services amid increasing financial pressures: The Council will have lost over £300 million in six years and there's more cuts to come and this service is under real pressure while we are still trying to maintain a service which focuses and supports all members of the community; so it's a tough line to walk. Participants described additional financial pressures from newly opened budget gyms offering highly discounted memberships. To generate more regular income and increase competitiveness, leisure teams had intensified marketing strategies that promoted direct debit membership to attract new and retain existing users or reducing membership pricing to increase its 'competitiveness'. In practice, the balance between a focus on welfare and commercial goals was not always dichotomous. In one area, for example, income generated from direct debit memberships enabled the service to subsidize activities for groups perceived less likely to participate in leisure. So our work on our business side and our gym and aerobics stuff is really important to bring down our net requirement but also mean we can still charge a reasonable rate for these youngsters and that to be involved. Public health professionals were also cognizant of tensions in using price to reach inactive groups amid competing financial pressures. This was highlighted with regard to the ways that leisure targeted activities to either existing/regular or non-users/inactive user groups. I guess from providing that leisure service point of view you've got that balance haven't you to get right between putting classes on that people can identify with from a range of different backgrounds including people who don't currently do any exercise. I can see that's a challenge. The complexity of financial structures for leisure pricing decisions was suggested to affect the use of public health evidence as well. I think that the solution from a Public Health perspective might be relatively easy but the solution from a local authority perspective might be so much more harder because … the finance attached to the existing structures is so complex that almost the evidence based almost might be secondary to everything else that's going on. --- Public health goals Compared to the two goals above, less emphasis appeared to be placed upon public health priorities as drivers underpinning how leisure entrance charges were set. Two further themes were identified in the findings that may help explain this. First, both public health and leisure professionals acknowledged that leisure provision had a key role to play in tackling inactivity in local populations, however, more divergent perspectives were offered about the role of price more specifically in tacking inactivity. While affordability was recognized as a financial barrier for low income groups, professionals suggested there were a myriad of factors affecting public decisions to use leisure or partake in PA more generally. I think my professional view would be that there are all sorts of barriers that prevent people from taking exercise and money may be one of those barriers but it is only one barrier. Examples cited of these barriers included the physical accessibility and location of the facilities, public attitudes to gyms, timing of when classes were put on as well as family and working constraints. Second, the research identified not insignificant variation in the levels of integration or joint working between public health and leisure teams. In one area, a historical approach to partnership working had resulted in the implementation of a community wide programme, which included a universal free leisure component . This programme received considerable investment from public health and leisure budgets over a number of years. In other areas a more typical model functioned where public health teams commissioned or worked with leisure services to deliver 'discrete' projects or interventions but public health teams had less direct involvement in pricing decisions. In three areas, partnership arrangements appeared less developed. Here, factors such as workforce restructuring and staffing changes as well as a two-tiered LA structure were reported to have affected levels of partnership working. --- Approaches to reduce or remove entrance charges Targeted concessionary schemes Schemes offering concessions for particular groups were in operation in all areas. Referred to as 'passport to leisure' or 'standard concessions' their introduction was often reported as being shaped by political ideologies and a historical legacy of affordable provision. There were members politically obviously wanted to encourage exercise in the facilities when [the] financial situation wasn't as strict. They came up with the scheme of; I don't think it came from us particularly; it was just something that the council wanted to introduce. These schemes required proof of eligibility from users although differences were evident between LAs in how criteria were applied to particular groups but also in the groups who were entitled to a concession. Most users were limited to using the facilities at 'off peak' times , although in one LA, users could attend at any time of day. Mixed viewed were expressed about the impact of these concessionary schemes. In one LA, the leisure team described the scheme as 'not an underused scheme by any means'. In other areas, leisure teams expressed concern that a 'blanket' approach to offering concessions to 'just about anyone' was likely to be unsustainable in the financial climate. However, attempts to tighten eligibility criteria could be politically and publically unpopular. We had some resistance because I made the decision to take out the qualifying 'because you're old' when we have some wealthy pensioners rocking up in Mercedes and benefitting from cheaper activities, which we felt was a little unfair. In two areas, funding cuts were also reported to have led to reductions in the level of concession offered for use of facilities under these schemes. --- Free leisure provision A smaller number of areas operated schemes that provided free leisure offers but these varied in scale and in approach. There were differences in whether such provision was universally offered or targeted at particular groups. A universal approach in one area included a large free offer aimed at anyone living or working in the LA. This was introduced alongside intensive marketing activities, a team of health trainers, dedicated instructors, community development officers and a network of volunteers. It isn't just about providing free leisure in leisure centres, a lot of things Leisure Professional 4 said around volunteering, the programme, walking and cycling, it needs to be a whole scale approach to leisure, sport, physical activity. In other LAs, free leisure offers included free swimming for under 16s during the holidays, swimming lessons for housing association tenants and free leisure access for young people under 18 or adults over 75 years. Yet for some professionals, there was uncertainty if free leisure should be available at all. Reflecting on their own local experiences, some professionals suggested that free access resulted in people devaluing schemes. Again you've got to put a value on something because don't respect. If there's no value or no cost value behind nobody respects it. --- Discussion Main finding of this study The research identified three goals underpinning pricing decisions. Welfare orientated and commercial goals featured most prominently across all areas. Fewer areas articulated public health as an explicit goal of pricing policies, with models of partnership between public health and leisure varying across LAs. The scale and targeting of concessionary or free offers was also approached in different ways. Professionals in strategic roles acknowledged the 'balancing act' in managing sometimes competing goals amid organizational pressures to make cost savings. During the study this was evidenced in some areas by reductions in concessionary offers available. --- What is already known on this topic Leisure pricing is one strategy with potential to tackle population level inactivity, however, evidence from intervention studies is mixed. For example, free leisure has been found to encourage teenage children from more deprived areas to swim more frequently 17 but that cost may be less important for older adults. 18 Yet there are few empirical studies investigating the impact on participation by different socioeconomic groups. Policies may also result in 'intervention driven' inequalities in different ways. 19 Inequalities may be widened if more affluent users take more advantage of universal free schemes compared to those who are less well-off or if cost related barriers to access are not taken into account. 17 Concessionary schemes targeting those in receipt of welfare payments could result in stigmatizing groups or constrain access if use of facilities is only offered at certain times of day. Paradoxically, access to leisure facilities located in more deprived areas could be reduced if cuts to public services result in reduced hours or retrenchment of the LA's role in the services it provides. 20 --- What this study adds The study provides evidence about the ways that LA leisure pricing decisions are made, and the approaches that are adopted in practice. These findings also offer some insights into the challenges of joint working for public health teams embedded within complex socio-political and economic environments. 21 In the context of budget cuts, public health decision makers require robust evidence to support investment in preventative approaches as well as make the case for health to be considered in wider LA decisions. 22 By embedding this qualitative investigation within a wider mixed method study, which treats LA pricing policies as a 'natural experiment', 23 the study will produce more detailed evidence needed for decision making as to how leisure pricing influences participation for whom and in which contexts. --- Limitations of this study The research took place in a small number of LAs who selfselected into the study. Limited contextual information is provided about areas to preserve participant anonymity as far as this is feasible. The accounts reported here may also not reflect experiences in other parts of the country. Fewer interviews with local politicians took place as the fieldwork clashed with a local election period and due to the departure of a researcher. It was not possible to interview public health professionals in all areas mainly due to personnel changes in some LAs.

Background Reducing or eliminating entrance charges for the public use of leisure facilities is one potential tool that local authorities (LA) have to reduce inequalities in physical activity (PA). Facility charges are likely to be a greater barrier to access for those who have lower incomes. Methods Semi-structured 1-to-1 and group interviews were conducted with 33 leisure and public health professionals in seven LAs in northwest England. We investigated how approaches to pricing varied in these settings and rationales influencing decision making. Results Welfare orientated (e.g. affordability) and commercial drivers (e.g. income generation) featured most prominently across areas. Pricing policies placed less direct focus on public health goals, although tackling inactivity was articulated as part of leisure's role more generally. Local targeting of free/concessionary offers was also defined and implemented differently. Decision makers described navigating competing pressures of providing services for the public 'good' yet remaining financially viable.Many LAs are reviewing the extent of subsidy for facilities or are considering whether to invest public health budgets in leisure. The findings offer evidence of how pricing decisions are made and the approaches adopted in practice as well as the conflicting priorities for decision makers within an austerity context.

Introduction Every year, an estimated 4.2 million youths and young adults in the United States experience some form of homelessness [1]. Black youth, unmarried parenting youth, and gay, lesbian, bisexual, transgender, and queer youth are at a higher risk for experiencing homelessness compared with other groups [1][2][3]. Youth experiencing homelessness have disproportionately higher rates of negative health outcomes, including disease, injury, substance use, and trauma [4][5][6]. They are also at increased risk for victimization and incarceration [7,8]. Consequently, mortality rates among YEH are 10 times higher than those of housed youth in the general youth population, with suicide and drug overdose being the leading causes of death [9,10]. The transient and hidden nature of youth homelessness make it challenging to conduct research with this marginalized population [11]. Unlike stably housed youth, YEH experience barriers to accessing health and social services [12,13]. Even when available, YEH are reluctant to engage with helpful services over concerns for loss of confidentiality and perceived provider stigma, prejudice, and discrimination [14,15]. In-depth interviews from the Voices of Youth Count initiative revealed that the need for YEH to protect their identity and their history of stressed, often toxic relationships with parents and other adults make them less inclined to interact with unfamiliar resources and institutions [16]. For many YEH, weighing out the risks and benefits of engaging with others is paramount because, from their perspective "nothing is for free" [16]. In recent years, the widespread availability of smartphones and advances in mobile technology have made it feasible to conduct ecological momentary assessment studies among hard-to-reach populations including YEH [17][18][19]. EMA collects real-time data as individuals go about their daily lives. As a result, some of the benefits of EMA include a reduction in recall and memory biases, an increase in ecological validity, and the ability to capture intrasubject variability among factors that contribute to behaviors [20][21][22]. EMA has been extensively used in adult addiction research [23][24][25], pediatric and adult mental health disorders [26,27], and nutrition and physical activity studies [28,29]. Research shows that EMA is acceptable among stably housed youth, yet the receptiveness of YEH to EMA is not well understood. Although there are similarities in the experience of EMA between the two groups, YEH face different challenges that require unique approaches [30]. To start, recommendations for recruitment of stably housed youth emphasize using schools as a starting point, a source that is not available for many YEH studies [31]. Clinical studies in stably housed youth also rely heavily on clinical institutions, such as primary care clinics, as touchpoints throughout the study, but YEH often face barriers to accessing these institutions [32,33]. Although stably housed youth have reported technology-related issues and recommendations similar to YEH, the lack of availability of outlets to charge a phone and disruptions in connectivity are significant barriers not commonly reported in stably housed youth studies [31,34]. Considering the higher incidence of victimization, trauma, and mental health issues among YEH, they may perceive EMA questions differently; thus, strategies that balance a thorough assessment within the context of potentially traumatic experiences need to be explored. The risks associated with mobile health technology in this vulnerable group is also not well understood. The few EMA studies conducted with YEH demonstrated high response rates and acceptability [6,18,35]. Compared to previous cross-sectional methods used to understand the daily experiences of this highly mobile group, EMA allows for a more naturalistic approach to studying YEH behavior and their environmental circumstances [36]. Using longitudinal EMA data, Suchting et al [37] expanded on our understanding of shelter use among YEH by predicting patterns of sheltering behaviors. They revealed that stress related to not having a place to stay and experiencing discrimination to be among some of the strongest predictors for YEH not staying in shelters. Santa Maria et al [6] used EMA to capture patterns of YEH rates of drug use and to identify modifiable risk factors for future interventions. In another EMA study, sexual urge and drug use predicted day-level sexual activity, and YEH primarily engaged in condomless sex, increasing their risk for HIV and other sexually transmitted infections [38]. Few studies have explored the usability and acceptability of mobile phone assessments among YEH. Using focus group interviews, Jennings et al [34] found that YEH perceived mobile phone health interventions as beneficial if they addressed their health concerns, maintained their confidentiality, and allowed them access to calling and texting features. YEH were also receptive to EMA using SMS text messaging and had several suggestions for the question format, including the ability to explain answers instead of being limited to preset responses [35]. These findings provide valuable recommendations for YEH EMA studies but more studies are needed to further explore EMA use in YEH and no studies have investigated the perceived impact of EMA on YEH behavior [17,34,39]. We conducted this study to address these gaps and to describe the perceived benefits, usability, acceptability, and barriers to the use of EMA from the YEH perspective. --- Methods --- Parent Study Participants in this study were recruited from a parent EMA study examining real-time factors, such as urge, stress, sexual behaviors, and substance use. Youths from the parent study were between 18 and 24 years of age, English speaking, experiencing homelessness or unstable housing, and recruited from shelters and drop-in centers serving YEH in 1 city in the southern United States [6]. Youths participated in up to 21 days of EMA that prompted them to complete a brief assessment that asked about stress, affect, sexual activity, substance use, and sheltering arrangements 5 times a day. More information on the parent study can be found in the original publications [6,40]. --- EMA Survey and Technology Participants received prompts to complete EMA surveys through a Samsung Galaxy Light smartphone with an Android 4.2 operating system . The surveys took fewer than 5 minutes to complete, and youths were cued visually and audibly for 30 seconds for each assessment. Daily assessments required participants to recall the past 24 hours, and random assessments occurred 4 times a day and focused on assessing real-time affect including feelings of anger, guilt, irritability, shame, and sadness [6]. Urge to use drugs or engage in sex or high risk behaviors including stealing were also assessed. Aside from the assessments, the phones could be used to send and receive phone calls and text messages and provided access to the internet and social media. At the end of the 21-day study period, participants met with the study team to return the study mobile phones and receive grocery store gift cards valued up to US $95 depending on the percentage of random and daily EMAs completed. --- Recruitment After completion of the EMA study, youths were asked if they would like to participate in an exit interview individually or in a small group of peers. Interviews were conducted using a semistructured, conversational guide that inquired about their experience with the EMA app, what they liked and did not like about the app, and how they thought it impacted them. The interview guide was designed to address predetermined themes related to app acceptability and usability with open-ended questions to allow for emerging concepts. Finally, youths were asked if they would use the app again or recommend it to a friend and what they would change about the app or the study to make it better for others. Sessions were audio-recorded and transcribed verbatim. Ultimately, 14 interviews were conducted in total, including 2 group interviews and 12 individual interviews . On average, both individual and group interviews lasted approximately 20 minutes. --- Analysis An inductive approach guided the data analysis for this study. A team of 2 doctoral-trained investigators performed the analyses. First cycle coding included descriptive, attributive, in vivo, process and values coding [41]. Codes were then clustered into categories and patterns to generate an initial codebook. Second cycle coding focused on thematic analysis to identify themes and subthemes. The final codebook was refined and included definitions, exemplars, contrast quotes, and number of discussions identified for each theme. ATLAS.ti was used to organize and code the data. Credibility and confirmability of the analysis was established through peer debriefing for each coding cycle, use of thick descriptions from the interview transcripts, and analytic memos during the analysis phase. --- Results --- Participant Characteristics A total of 18 youths participated in this study . Ages ranged from 19-24 years with a mean age of 21 years. For race/ethnicity, 11 out of 18 participants identified as male, 10 out of 18 identified as non-Hispanic Black or African American, 5 out of 18 identified as multiracial , 1 identified as American Indian or Native Alaskan, and 1 participant reported "something else". Most of the youths identified as heterosexual . --- Perceived Impacts of EMA --- Increased Awareness of High Risk Behaviors The majority of youths reported that the EMA increased their awareness of their high risk behaviors. The daily questions asking about their smoking habits, drug use, and sexual activity served as reminders of the frequency of their behaviors: "It kept me informed of how many drugs I use" . Another youth reported, "I would notice how many cigarettes I would smoke when I first started this survey, so that really made me like, okay, I really need to stop smoking those" . This youth further described how the repetition of questions led to increased behavioral and emotional awareness: --- Yeah, it made me step back and look at everything because like I said, since it's so redundant, it makes you notice your habits as well because you sit there, and you answer, and you start noticing I'm answering this question so many times this answer or this answer, and you start noticing. Then that makes you think to yourself, maybe I need to try and fix this because that'll help me change my mood or whatever. [Y10] Four youths reported changing their behavior as a result of the EMA surveys. They described decreasing their substance use and taking precautions prior to engaging in any sexual activity: "It just make you think about certain things like you about to get ready to do something or it just makes you think or take precautions about what you're doing" . One youth described decreasing his use of marijuana as a result of the survey: --- I noticed that, you know, the marijuana really wasn't helping. I still didn't stop completely at that moment, but you know, down the line, I wean myself off of it. I'm still weaning myself off of it. I haven't really smoked in like 2 weeks. So that's good. [Y17] Youths reported making changes in their sexual behaviors with some ensuring they had a condom available at all times: With the sex, it kind of helped me slow down a little. We were having sex-me and my girlfriend-and it did kind of slow us down a little bit because I noticed we were just having sex recklessly. [Y10] --- Increased Awareness of Emotional State Six youths conveyed that the EMA surveys helped them be aware of their emotional state. Youths reported that the words in the questions allowed them to recognize and name their emotions: --- It was like the way that they ask the questions, it helped me kind of put it in words because I would sometimes feel like fidgety or irritated. I'm like, I'm not really sure why. But then the way they would ask I was like "Ok, that's probably why." Maybe that's the word I'm supposed to use. [Y3] Another youth described the survey as a way of coping with their emotions: It help me be more aware of how I was feeling, because it was like you asked the right questions at the right time. So, it's more like self-coping. Therapy in a way. Like, basically you're being able to --- XSL • FO --- RenderX acknowledge how you are feeling, like what is causing that emotion that's holding you down. [Y8] Youths described the benefits of acknowledging their stress and identifying emotional triggers: "It brings to your own attention that okay, well, a lot of the stress that I'm having is sometimes due to me or it's sometimes due to things around me" . Another youth reported, "Most of my stress came from not just everyday life but mostly from how I put myself through life" . One youth described the impact of the survey on her stress and coping skills: I mean, --- Provided Social Support and Connection For many youths, participating in the survey made them feel heard and supported: Emotionally, I felt like there was actually somebody who cared or wanted to know how I felt. I like that. --- [Y18] They reported that the ability to not only express themselves but also the knowledge that someone was reading their responses was comforting. One youth described how the survey helped her cope with loneliness: …it was kind of helpful to me because sometimes I had no one to talk to and then I be feeling like I can't express, but when I take the survey and it ask like how you feeling or are you depressed. It make it seem like someone actually going to look at this and say, well, she's feeling this type of way. Somebody actually sit down and cares about your feelings. That's what I liked about the survey. [Y17] Seven youths explained that the ability to connect with friends and family through phone calls, text, and social media was most beneficial. One youth used the phone to reconnect with her family whom she had not spoken to for some time: "…haven't talked to my mom in a long time. I finally gave her a call" . --- Acceptability of EMA --- Provided a Distraction From Homelessness In general, YEH were receptive to the EMA surveys. Some reported that the surveys distracted them from homelessness: --- I be like trying to get my mind off things, like focus on something else. I'll be like "Okay, well I'll take my mind off this and focus on the surveys." [Y18] Several reported it helped them avoid risky behaviors: --- If I'm bored, I get adventurous. If I'm on the phone, I just be still stuck on the phone just watching movies or videos, anything, for hours. [Y12] YEH also expressed appreciation for being able to use the phones for other purposes including phone calls and social media: You didn't just give them a phone and oh, here are those surveys. Even though we are young kids, we have access to Facebook, we have access to things we actually do. They can entertain themselves, and they're on the streets, so they can charge the phone and be on the phone all night, something for them to do. It's not just doing surveys on this phone and still sitting outside bored. [Y15] --- Perception of Being Watched Youths described instances in which they felt that the prompts came as a response to their real-time behavior. They mentioned feeling "watched over" or being monitored especially when the survey questions applied to their real-life situation: "Sometimes I felt like it was the signs; it was kind of weird because it would pop up only when I'm about to do something" . One youth described his experience after a police encounter: --- One time I actually thought the phone knew what I was doing because 2 minutes later, I had got into it with the lawman and right after I finished talking to the lawman, the survey came up and its first question was "Did you encounter a lawman or a police officer?" I was like, "really, bro?" [Y12] --- Barriers to Acceptability and Engagement Redundancy of questions was a frequently mentioned issue. As a result, youths described being able to predict the questions and answering without much thought to their responses: "Because I would get bored because it's just like now, I know exactly what it's going to say, and it's like click, click, click" . One pregnant youth described that the lack of customized questions was a problem: "Only one time did I get something that asked if I was pregnant, and I was like yeah, and then it asked me the same exact questions it would ask me even if I wasn't pregnant" . Some youths reported that the lack of response to their answers was dissatisfying. One youth described the following: --- Let's say, by the third time, I say I'm extremely stressed, maybe that could send y'all a red alert like, "Hey this person's stressed. They might need someone to talk to." Then you could send them a message like, "Are you okay? Want someone to talk to? Need anything?" I feel that would make certain individuals feel a little bit more looked after, a little more cared for. And it's not just like, oh, it's a survey…if you really need somebody, this is for us to find out. [Y13] --- Usability of EMA --- Easy to Use but With Glitches Youths reported that the surveys were easy to navigate but that the questions were often predictable: "…it got to the point where my boyfriend, he knew every question, so he could just type" . Most youths found the redundancy of questions to be boring and not applicable to their specific situations. Although they did not mind the number of surveys administered daily, some of the youths found the timing of the surveys to be challenging: "If you're not on it [phone] or not around it, it'll go off and you hear it…and then you missed it" . A few of the youths reported "glitches" occurring while taking the surveys describing that the surveys would turn off midway leaving some questions unanswered: "…like it would pop up-sometimes a survey wouldn't pop up; it would just go off, so that's why some of them weren't answered" . --- Connection and Other Technology-Related Issues Technology-related issues were common. Some described not being able to make phone calls: "…when I try to call out, it'll say only numbers that can be dialed or numbers that are programmed" . Others reported issues with internet access: "When I first got the phone, it was going really fast; and then like two days after that, everything started slowing down…it got really, really slow" . Some stated that the slow internet access also slowed down their ability to send the completed surveys. Battery issues were commonly reported by youths who often struggle to find a place to charge their phones: "The app was good, just the phone was bull crap because it only worked for like 2 days, and then after that it just went down" . --- YEH Recommendations for Future EMA Studies The majority of youths recommended that the survey questions should be customized to each participant to avoid redundancy: "I don't think each survey should say the same thing; I think you should switch it up" . YEH also expressed that although the questions about mental states and substance use were important, they would also like to be asked everyday life questions such as "How are you feeling today?" or "Are you hungry?" or "Did you work out today?" They also reported that there needed to be more questions and that questions should "go more deep into it [topics]" . Two youths expressed that they would have liked questions about their plans for the future and where they saw themselves in 5 years. Some of the youths recommended that the ability to type in their responses instead of choosing from preset answers would be preferable: --- Yeah, type in your answer. "I feel upset because,"… I just feel like they shouldn't be "sometimes," "often," "agree," "disagree". Have distinct answer. [Y15] Youths expressed that additional supports should be provided through the EMA surveys. However, they recommended that support should be personalized so that the participants feel "more cared for". Describing this type of support, one youth stated the following: Don't read them a script. Literally sit there and ask them, why do you feel like this? Don't suggest that they go to a mental hospital. Give them encouraging words, and talk to them for a minute, ask them do they feel safe at this moment, and if they don't feel safe, then "Can you call the police for me?" "Can you call the ambulance?" And then take that step. [Y15] Three youths recommended that allowing customization of the phone ring tones may help with survey compliance. Others stated that allowing participants to go back and change their answers and to retroactively complete missed surveys would be helpful since surveys may come when they are too busy to respond, or they respond incorrectly: --- I also think you should make it to where you can be able to go back and redo your answer. I think that is a big part of it because I just-I really like it. I really liked this. It's a good thing to me. [Y15] To prevent phones from being stolen, several youths recommended that participants should be screened carefully prior to receiving their phones: "You just have to be pretty strict on who gets them…because people see that it's just a regular phone and not something you have to do a duty on" . Many of the youths expressed that allowing the youth to keep the phone after the study would help with keeping them connected to social media and provide them access to resources. Some youths found the timing of the surveys challenging and suggested set times for the survey to increase compliance. One youth indicated the following: …what would be a little bit better is if the times were consistent because the one that I missed, I was on the internet looking up the bus route for something, and it came up, and I'm like in the middle of trying to find the bus route, and then my dad started calling, and then my phone just froze. I was like, if I knew it come at 10:30 and 12:15, I would know, okay, have my phone charged. And I think-and that's why my boyfriend missed so many because he never knew when they were coming, but he never brought his charger. He was like, "Man, if I knew what time it was coming, I'd at least make sure my phone was charged for that time…" [Y13] --- Discussion --- Principal Findings The findings in this study suggest that EMA is acceptable to YEH and may have perceived behavioral benefits. Increased self-awareness as a result of receiving daily questions about their behaviors and emotions led some youths to decrease their substance use and increase condom use before sexual intercourse. Others expressed that the questions helped them recognize stressors that led to negative emotions, and as a result, youths intentionally avoided these emotional triggers in the future. Researchers have previously noted the impacts of frequent assessments on behavior and emotional awareness irrespective of interventions [42,43]. This effect, known as assessment reactivity, results from participants' constant reflection during repetitive behavior evaluations, leading to change [44]. EMA questions may serve as cues to action by providing information and reminders that trigger health-promoting behaviors [42]. However, previous EMA reactivity studies in alcohol abuse, smoking cessation, and medication adherence have found mixed results [44][45][46][47]. A randomized control trial evaluating EMA reactivity in smoking cessation discovered that although increased assessments did not promote abstinence, researchers observed small reactivity effects on anxiety, hunger, positive affect, and post-quit confidence [47]. To date, no studies have explored this concept in YEH, and it is unclear whether youths sustained their behavior change postintervention. Researchers should consider the effects of assessment reactivity in the design of future EMA studies to anticipate potential confounders in youths' subjective experiences and to assess for weaning effects over time. Youths felt high levels of perceived support through their interactions with the EMA surveys. Although the surveys were pregenerated, administered at set times during the day, and unidirectional in nature, being asked about their day made youths feel heard and looked after. Tyler and Schmitz [48] observed similar findings in their SMS study with YEH, reporting "perceptions of care and concern" as one of the perceived benefits of participation. These discoveries highlight the degree of loneliness many YEH experience; interactions, even in survey form, appear to alleviate feelings of isolation and depression common in YEH [49]. In this regard, EMA may have unintended therapeutic effects on YEH's mental well-being, particularly those with low social supports. Youths unanimously reported that the most beneficial aspect of the EMA study was the ability to connect with friends, family, and other social support through the mobile phone. Access to the internet allowed youths to communicate with their peers through social media and to access multimedia platforms that helped alleviate boredom. Some of the youths used the phones for directions to clinics and grocery stores; one youth reported a friend getting a job interview through the phone. Many of the youths recommended that participants be allowed to keep the phones after the study. This is not surprising, as previous studies observed that mobile phones serve as lifelines for youth [11,39]. Aside from communicating with peers, they used phones to access resources and other forms of support [34]. However, mobile access can have adverse effects with youths expressing concerns for loss of anonymity and misuse of their personal information [34,39]. In this study, YEH did not report these concerns; however, several described feeling watched over and believed that the surveys occurred as a response to their real-time behavior and situation. Future EMA studies should consider the potential for harmful effects of mobile connectivity on YEH privacy; clearly outlining the plans for data management and measures to ensure confidentiality may help to alleviate these fears. EMA was highly acceptable to YEH. Youths reported that the questions were easy to understand and that the frequency and length of the surveys were manageable. Some of the youths recommended that questions should "go deeper" and that participants should be allowed to provide contextual answers instead of preset responses. Youths also recommended that surveys should be customized to each individual and that redundancy of questions should be avoided. Several youths reported that redundancy led them to answer mindlessly after taking the same survey multiple times. These findings are similar to recommendations from other studies with YEH reporting that the authenticity of questions and the ability to explain their answers were most important [17,34]. YEH emphasized that authenticity may only be realized when they are directly involved in content development [34]. A collaborative approach to EMA study designs may improve participation and provide valuable insights into the YEH experience. Although YEH reported that the surveys were easy to navigate, they encountered issues with connectivity, phone battery life, and timing. The stalling of the EMA app resulted in incomplete or missed surveys. Youths reported issues with the phone battery life despite frequent charging and experienced slower connectivity as the study progressed. Youths also described instances when they did not receive notification or were unable to complete the surveys on time. Other EMA studies have reported similar issues. Mackesy-Amiti and Booram [50] observed that notification issues , stalling of the EMA app , and battery issues accounted for most of the problems encountered in their study. A systematic review of the use of EMA methods with youth found that software and hardware malfunctions and connectivity issues were common across many EMA studies [51]. As technical issues can significantly impact data collection and ongoing participation, efforts should be made to provide ongoing technical support to help troubleshoot software and hardware problems. Piloting the phone and app prior to the start of the study may help mitigate potential issues; however, lost phones were a significant problem in this study, and piloting may not always be feasible in this population. YEH had several recommendations for future EMA research. They emphasized the importance of providing real-time support for themselves and their peers during hard times. For example, youths identified during the EMA as having suicidal thoughts or who are feeling severe depressive symptoms could be connected to crisis personnel for help. Aside from additional support, youths also wished for "normal" interactions and more goal-related questions. For instance, they wanted to be asked how they were doing and about their plans for the future. These findings reflect the perceived stigma YEH experience; being seen as difficult or different from other "normal" youth are negative stereotypes [52]. Including questions unrelated to mental health or high risk behaviors may help to minimize this stigma and signal to YEH that they are valued as a whole individual. Youths also expressed the desire to receive motivational messages to help them through difficult times. Previous studies observed that youths were particularly interested in receiving motivational messages, tips on staying calm, and advice on how to navigate relationships [17,39]. Integrating these preferences into the EMA may improve acceptability and survey compliance. This study aligns with current literature that EMA has high acceptability and utility with YEH. Several of the findings are similar to those found in stably housed youth; however, the unique challenges faced by YEH and the potential of EMA to impact high risk behaviors are highlighted in this study. Although findings suggest a positive perceived behavior impact, more studies are needed to examine the direct relationship between EMA and behavior change, the impact of assessment reactivity on EMA data, and the sustainability of change postintervention. This study adds to the literature by providing recommendations for EMA research from the perspectives of YEH. Future research is needed to identify strategies for involving YEH in EMA development and understanding barriers to participation and compliance. Future EMA studies would benefit from incorporating just-in-time adaptive messages that respond in real-time to EMA data and generate youth-developed motivational, health-promoting messages. Additional trials should be conducted to assess the behavioral impacts of EMA. --- Limitations This study has several limitations. As this is a single-site study with a relatively small sample size, the generalizability of findings to other populations of YEH is limited. The experiences reported here may be unique to those who chose to participate in exit interviews; these youths may have had more positive or negative experiences, which may not be reflective of the experience of all the youths who participated in the original EMA study. There is also an underrepresentation of LGBTQ youth in the sample; estimates suggest that up to half of the general YEH population identify as LGBTQ [53]. As data were collected from both individual and group interviews, more vocal participants might have disproportionately impacted specific themes; however, all efforts to obtain exemplars from a diverse group of youth experiences were made by the researchers during the analysis and the reporting phases of the study. --- Conclusions Findings from this study describe a positive perceived behavior impact on youth high risk behaviors; however, more studies are needed to fully understand the direct effect of EMA on behavior change among a larger and more diverse YEH population. EMA methods are highly acceptable to YEH, but technology and connectivity issues serve as significant barriers to participation and survey compliance. The development of future EMA studies should involve YEH to ensure the authenticity of surveys and address their specific needs. --- Conflicts of Interest MB is the primary inventor of the Insight mHealth Platform, which was used to develop the smartphone application for this study. MB receives royalties when researchers outside of OUHSC use this platform. Other authors declare no conflicts of interest. --- Abbreviations

The use of ecological momentary assessment (EMA) to study youth experiencing homelessness (YEH) behaviors is an emerging area of research. Despite high rates of participation and potential clinical utility, few studies have investigated the acceptability and recommendations for EMA from the YEH perspective. Objective: This study aimed to describe the perceived benefits, usability, acceptability, and barriers to the use of EMA from the homeless youth perspective. Methods: YEH were recruited from a larger EMA study. Semistructured exit interviews were performed using an interview guide that focused on the YEH experience with the EMA app, and included perceived barriers and recommendations for future studies. Data analyses used an inductive approach with thematic analysis to identify major themes and subthemes. Results: A total of 18 YEH aged 19-24 years participated in individual and group exit interviews. The EMA was highly acceptable to YEH and they found the app and EMA surveys easy to navigate. Perceived benefits included increased behavioral and emotional awareness with some YEH reporting a decrease in their high-risk behaviors as a result of participation. Another significant perceived benefit was the ability to use the phones for social support and make connections to family, friends, and potential employers. Barriers were primarily survey and technology related. Survey-related barriers included the redundancy of questions, the lack of customizable responses, and the timing of survey prompts. Technology-related barriers included the "freezing" of the app, battery charge, and connectivity issues. Recommendations for future studies included the need to provide real-time mental health support for symptomatic youth, to create individually customized questions, and to test the use of personalized motivational messages that respond to the EMA data in real time. Conclusions: YEH are highly receptive to the use of EMA in studies. Further studies are warranted to understand the impact of EMA on YEH behaviors. Incorporating the YEH perspective into the design and implementation of EMA studies may help minimize barriers, increase acceptability, and improve participation rates in this hard-to-reach, disconnected population.

Introduction A nnual, anonymous HIV testing of pregnant women 1 has been central to understanding the evolving HIV epidemic in generalized epidemic settings. As epidemics mature in these settings and with increasing coverage of antiretroviral treatment and interventions to prevent vertical transmission of HIV the prevalence of HIV infection is a less reliable marker of the evolving epidemic. Thus, there is a need to expand current surveillance systems to include new adolescent populations that provide a better marker of new HIV infections. In South Africa, heterosexual transmission accounts for more than 90% of new infections. HIV prevalence in adolescents between 15 and 24 years in high disease burden communities provides a reasonable proxy for incident HIV infections, because infections are likely to be relatively recent and HIV-related mortality is likely to be minimal. [2][3][4] The burden of HIV in the South African adolescent population is unprecedentedly high, particularly in adolescent girls aged 15-19 years who acquire HIV at least 5-7 years earlier than their male peers and have a 3-to 4-fold higher incidence rate. 5 Given that the majority of adolescents in these settings attend high school, high schools provide a convenient opportunity for expanded surveillance. Furthermore, schools also provide important venues for intensifying HIV prevention efforts and surveillance in this setting, and enable the impact of interventions to be assessed. Phylogenetic methods have also been applied to investigate HIV transmission dynamics and are useful for identifying high-risk physical environments and providing insights into understanding HIV spread. 6 With modern analytical approaches, these important tools not only enable transmission clusters to be inferred, but also allow estimation of the timing of HIV transmission, thereby permitting smarter targeting of HIV prevention efforts. We assessed the potential for including at-risk adolescent populations through school-based surveillance, together with the utility of phylogenetic mapping, as measures to enhance the understanding of the evolving epidemic and transmission dynamics in a high disease burden rural district of South Africa. --- Materials and Methods --- Study setting and population The study was conducted in rural Vulindlela, a subdistrict of Umgungundlovu 150 km west of Durban in the province of KwaZulu-Natal, which is one of four highest burden HIV districts in South Africa. The HIV prevalence among pregnant women attending public sector primary health care clinics in Vulindlela in 2011 was 39.8%. 1 The area is characterized by high levels of poverty due to its limited resources and employment opportunities. The estimated population is about 150,000, of which approximately 16,000 are in high school. 7 The establishment of the school-based surveillance was preceded by extensive consultative meetings with key stakeholders including the Departments of Education and Health at the provincial, district, and school level and with school principals and educators, governing bodies, parents, and students. --- Study design and procedures Between February and May 2012, we conducted a crosssectional study in five randomly selected public sector high schools within a 15 km radius of each other. Consenting students aged 12 years and older were eligible for inclusion in this survey. Prior to study initiation, parents and/or guardians of students were notified and had an opportunity to determine whether to allow or refuse their child's participation in the study. Students were provided with general information on HIV and on concepts of HIV surveillance, study information, study procedures, and the confidentiality of their participation and the data collected. Thereafter students agreeing to participate were individually provided with details of the study and taken through the informed consent process. A brief reason was obtained from students not agreeing to participate. Students consenting to study participation had dried blood spot samples collected by trained staff who also administered a standardized structured questionnaire to obtain sociodemographic and select behavioral information. Students were assured of their anonymity both in data collection and analysis. Laboratory data were linked to sociodemographic and behavioral data through a unique code. Students had access to HIV counseling and testing , sexual reproductive health services , and medical male circumcision services that were provided through the primary health care clinics and the CA-PRISA Clinical Research site in the district; these services were also available to other family members. HIV testing was performed using the Vironostika HIV Uni-Form II plus O Assay and the Elecsys HIV Combi assay . b-Human chorionic gonadotropin was measured using the Quickvue one-step HCG combo , with all positives confirmed with HCG combo rapid test . HIV-positive DBS samples were used for phylogenetic analysis. Using the Harris UNI-CORE 3.00-mm punch, DBS spots were punched out and nucleic acid was extracted from the samples using the QIAamp DNA Mini kit according to the manufacturer's instructions, with the following modifications: a maximum of seven, 3-mm-diameter punches was processed per tube; for elution of nucleic acids, a minimum volume of 50 ll molecular biology grade water was added to the column and incubated in the column for 5 min before centrifugation. Amplicons for sequencing were generated using one-step reverse transcription polymerase chain reaction and nested PCR , 8,9 with the following differences 8 : 0.8 • reaction mix was used and 4-27 ll nucleic acids was used per RT-PCR reaction. Five microliters of prenested product was inoculated into nested PCR mixes. Prenested PCR primers used were DT1 5¢ ATG GGT GCG AGA GCG TCA GTA TT 3¢ and DT7 5¢ CCC TGA CAT GCT GTC ATC ATT TCT TCT 3¢ and nested PCR primers were DT3 5¢ CAT CTA GTA TGG GCA AGC AGG GA 3¢ and DT6 5¢ ATG CTG ACA GGG CTA TAC ATT CTT AC 3¢ . 9 PCR products were sequenced using a BigDye Terminator V3.1 sequencing kit and sequencing primers used were DT3 and DT6. Chromatograms were assembled using Sequencher software . For superimposed peaks in sequence chromatograms, the dominant base was used; if it was not clear which was dominant, the ambiguous code was used. Sequences were approximately 605-635 bp in length. --- Ethical considerations The protocol was approved by the Biomedical Research Ethics Committees of the University of KwaZulu-Natal and the KwaZulu-Natal Departments of Health and Education. The HIV sequencing protocol was approved by the University of Cape Town . --- Data management and statistical analysis Data were collected on standardized case report forms , faxed to a dedicated study database using DataFax , and linked to laboratory data. The select demographic, behavioral, and biological data were summarized using descriptive summary measures, expressed as means [standard deviation ] and/or medians [interquartile range ] for quantitative variables and percentages for categorical variables. The 95% confidence intervals and p-values are reported for the school adjusted HIV prevalence. Comparisons of adjusted summaries by gender were performed using t-tests for two independent samples. The association between HIV and demographic, behavioral, and biological measures was evaluated with the multivariable logistic regression analysis adjusted for clustering to estimate the odds ratios [adjusted odds ratio ] and 95% CI. All analyses were performed using SAS statistical package . Sequences were subtyped using the jumping profile hidden Markov model 10 and REGA HIV Subtyping Tool . 11,12 Multiple sequence alignments were generated using RAMICS, a codon-based multiple sequence alignment tool, 13 and manually checked for accuracy. Codon-sized columns containing a majority of gaps were deleted from the alignment. HIV-1 subtype C gag p17p24 sequences derived from recently infected women and students 7,14-16 from previous studies in the Durban and Vulindlela areas were used. Phylogenetic reconstruction was performed using two maximum likelihood approaches implemented in FastTree 17 and RAxML . Putative transmission clusters were characterized as groupings with support values greater than 90% in both of the resulting phylogenies. 18 --- Results A total of 3,242/3,967 students participated in the study; 365 refused to participate and 361 were absent from school at the time of the survey. --- Demographic and behavioral characteristics The median age of girls and boys was similar . The demographic and behavioral characteristics are presented overall and by gender in Table 1. The proportion of girls and boys across grades 8 to 12 was similar; attrition rates were high in both girls and boys in grades 11 and 12. Family structures tended to be unstable, with almost a quarter of students not having a mother alive, and a further 22.6% having a biological mother alive but not living with them. Fathers were generally absent in the students' lives; indeed, only 58.3% of students reported that their biological father was alive and only 51.7% lived with their fathers. Almost a third of the boys and about a fifth of the girls reported ever having had sex. The median age at sexual debut was 17 years for girls and 16 years for boys. Boys reported a higher median total number of lifetime sex partners compared to girls . Among those who reported ever having had sex, boys compared to girls were more likely to have had a sex partner either the same age or younger than themselves . However, girls were more likely to have had a partner at least 1-4 years or 5 years or older than themselves. Pregnancy prevalence was 3.3%. --- HIV infection HIV prevalence and demographic, behavioral, and biological correlates of HIV risk are reported in Table 2. Overall, HIV prevalence was 6.8% in girls and 2.7% in boys . There was substantial variability in HIV prevalence between schools, by sex and age, with an intraclass correlation of 0.005 and a design effect of 4.2. In boys, HIV prevalence ranged from 1.8% in school A to 4.1% in school C. In girls, HIV prevalence was consistently higher in all schools, ranging from 3.5% in school E to 9.0% in school C. In girls, HIV prevalence increased in a dose relationship with age from 4.6% in 12 to 15 year olds, to 5.3% in 16 to 17 year olds, 8.7% in 18 to 19 year olds, and 23.1% in 20 year olds or older. In contrast, in boys HIV prevalence was relatively constant at around 2.7% until age 19 years, but increased dramatically to 11.1% in those 20 years and older. --- Phylogenetic relationship of HIV-1 sequences Of the 120 samples sequenced , 119 were classified as subtype C and one as A1 . Figure 1 shows the reconstructed phylogenetic tree of HIV 1-subtype C groupings obtained from students in comparison to previously characterized sequences. Nine sequence clusters exclusively containing sequences from students were identified, with two samples per cluster. In four clusters , two sequences in each cluster were from the same individual but were collected at different time points. The remaining five clusters contained sequences from different students. Three of the five clusters were sequences from students of the same gender , and one male only cluster indicating a third unidentified source of infection. In comparison to the students, clustering was observed for only 7 of the 135 background data sequences , one containing a sequence from a student , suggesting that a larger number of students in schools clustered together when computed against the background community data . Of the five clusters, two sequences were from individuals from the same school and three from different schools . --- Factors associated with HIV infection The association of demographic, behavioral, and biological factors with HIV is reported in Table 3. Compared to boys, girls were three times more likely to be HIV positive, aOR = 3.0, 95% CI 2.4-3.8; p < 0.0001, and those reporting ever having had sex, aOR = 2.8, 95% CI 1.9-4.2; p < 0.001. Although not significant, girls having a sex partner with an age difference at least 5 years and older were more likely to be HIV positive. HIV prevalence was higher in those 20 years and older not having a living biological mother and ever having had sex . There was no association between pregnancy and HIV . Of note, 21 girls who reported not ever having had sex tested positive for bHCG. In contrast, HIV infection in boys was higher in those without a living biological mother or father . f Calculated for those reporting ever had sex. --- Discussion In this survey, we found that one out of every 20 high school students was HIV infected, underscoring the magnitude of the epidemic in this rural setting. The gender disparity in HIV prevalence among girls and boys is striking, with our data highlighting an almost three times greater HIV risk in adolescent girls compared to their male peers. Though similar patterns of HIV prevalence have been reported in populationbased surveys and from a representative sample of 15-24 year olds, 5 these have focused on out-of-school students. The high HIV prevalence in these adolescents also suggests that as yet there has been little or no impact of HIV prevention and treatment programs on HIV acquisition rates. A major concern is that HIV testing rates among adolescents remains low and the majority of HIV-infected individuals are unaware of their HIV status, posing a barrier to HIV prevention, care, and treatment efforts. 19 Moreover, as many of these recently infected adolescents are likely to be in the acute or early phases of HIV infection with potentially high viral loads, they also serve as reservoirs for HIV transmission sustaining the survival of the epidemic. Over the past 20 years antenatal 1 and populationbased 5,20,21 HIV surveillance has reasonably described the evolving epidemic in South Africa. However, as individuals benefit from increasing access to ART and the implementation of expanded options for HIV prevention, prevalence measures from current surveillance programs over time become less reliable. In adolescents, most HIV infections are relatively recent and HIV-related mortality is relatively low, [2][3][4] such that HIV prevalence in this age group will be useful to infer HIV incidence. Given that most young South Africans attend high school, these institutions may represent important and convenient venues for surveillance expansion to better assess trends in HIV infection. We found that undertaking such high school-based surveillance was feasible, with a student participation rate exceeding 80%. In our preliminary phylogenetic analyses designed to yield clues as to where HIV transmission is occurring and/or being sustained, our analysis identified five clusters within schools, which suggests that students are observed as groupings in a highly supported putative transmission cluster when compared to background data. Although HIV infection was not associated with intraschool transmission in this rural setting, the source of infection and the direction of HIV transmission, although difficult to ascertain, require greater coverage and further analysis of schools and community sequences. To better understand HIV transmission dynamics in adolescents, larger sample sizes with robust study designs, which include more diverse populations, are needed to understand viral diversity for targeting of interventions. These studies will be critical, as any reliable data on infection dynamics from southern African settings are limited. 22 As biomedical interventions are designed and rolled-out, it will be vital to understand the proportion of HIV infections that may be attributed to any major source of transmission over time. Such understanding requires improved geographically representative surveillance with phylogenetic and epidemiological analyses. 23 We report that at least one in four students was sexually active; sexual debut occurred at a young age of around 15 years with a mean of about two lifetime sex partners. Many adolescents in comparably high HIV burden settings also have a high burden of sexually transmitted infections, 24 pregnancies, 7,25,26 alcohol use, 27 and substance use. 28 Although we found no association between pregnancy and HIV, HIV prevalence was higher in pregnant girls, and our data underscore the importance of linking HIV prevention efforts to a broader SRH and healthy lifestyle education. Certainly, the observed prevalence of HIV infection in those female students < 20 years of age who engage with sex partners at least 5 years their senior should be incorporated as a key message within school HIV prevention programs, and in the longer term age mixing within grades may need to be addressed within South African schools. Previous literature has indicated that cohesive family structures are important determinants of risk, as parents could help instill and reinforce messages of protection from risky behavior. 29 Our data confirm that students who have lost parents are at increased risk of HIV, and highlight that identification of these young at-risk populations is critical as we identify priority groups for HIV prevention interventions. A key strength in our study is the large sample size with a response rate of over 80%; however, there are several limitations. Being a cross-sectional study, we were unable to establish the temporality of risk behaviors and HIV infection; we hope that future surveillance will enable such information to be inferred. Furthermore, we report some social desirability bias, with 1.5% of girls who reported never having had sex testing positive for bHCG. Although the phylogenetic analyses suggest that infection among students is not associated with intraschool transmission in this rural South African setting, larger sexual transmission network studies are needed to better understand HIV transmission dynamics. However, in spite of these limitations this study provides valuable information on the burden of HIV infection and prevalent risk behaviors, and provides evidence that the robust and adolescent-focused surveillance systems required to measure new infections in a mature epidemic setting are feasible. In conclusion, the epidemiology of HIV infection in this community is dynamic and complex. The high burden of infection among students, the first phylogenetic study among high school students showing limited intraschool transmission, and the absence of a living parent underscore the importance of encouraging schools to intensify HIV surveillance programs in order to monitor the epidemic, identify transmission patterns, and provide comprehensive HIV prevention packages to reduce adolescent's vulnerability to HIV infection. --- Author Disclosure Statement No competing financial interests exist. --- 964 KHARSANY ET AL.

In South Africa, adolescents constitute a key population at high risk of HIV acquisition. However, little is known about HIV transmission among students within schools. This study was undertaken to assess the risk factors for HIV infection and the extent of transmission among rural high school students. Between February and May 2012, consenting students from five randomly selected public sector high schools in rural KwaZulu-Natal participated in an anonymous cross-sectional survey. Dried blood spot samples were collected and tested for HIV. b-Human chorionic gonadotropin (bHCG) levels were measured in females for pregnancy. Family circumstances as well as sociodemographic and behavioral factors were assessed as potential risk factors. A subset (106/148, 72%) of HIV-positive samples underwent gag p17p24 sequencing for phylogenetic analysis. A total of 3,242 students (81.7% of enrolled students) participated. HIV prevalence was 6.8% [95% confidence interval (CI) 3.9-9.8%] in girls and 2.7% (CI 1.6-3.8%) in boys [adjusted odds ratio (aOR) = 3.0, CI 2.4-3.8; p < 0.001]. HIV prevalence increased from 4.6% (95% CI 1.9-7.3) in the 12-to 15-year-old girls to 23.1% (95% CI 7.7-38.5) in girls over 20 years, while in boys HIV prevalence increased from 2.7% (95% CI 0.6-4.9) in the 12-to15-year-old boys to 11.1% (95% CI 2.7-19.4) in those over 20 years. Sequencing of samples obtained from students revealed only two clusters, suggesting within-school transmission and three interschool clusters, while the remainder was most likely acquired from sources other than those currently found in students attending the school concerned. HIV prevalence in both girls (aOR = 3.6, CI 2.9-4.5; p < 0.001) and boys (aOR = 2.8, CI 1.2-6.2; p = 0.01) was higher in those without a living biological mother. The high burden of HIV infection among students was not associated with intraschool transmission in this rural setting. Lack of a living parent is an important factor defining high risk in this group of adolescents.

Introduction Social assistance policy must identify with values of freedom, equality, and justice, and this must be a commitment that its workers fall in love with from head to toe. . The work of occupational therapists in social assistance services dates back to the institutionalization of the profession in Brazil. Initially, the category was part of technical teams in institutions linked to this sector, mostly civil society organizations that provided services aimed at assisting specific population groups, still marked by disability1 . Later, in the 1970s, occupational therapists were included in services aimed at coping with social problems unrelated to the field of health, such as the situation of children, older people and, at that time, the so-called juvenile offenders under institutional care . Although the liaison between occupational therapy and social assistance was present in the historical development of the profession in the country, only in 2011 were occupational therapists formally recognized as a professional category capable of composing social assistance teams. As a result, the discussion about work in this sector and the social question has gained relevance in the profession both from the interventional and theoretical-academic points of view. There is recognition that this work requires the incorporation of theoretical and methodological frameworks suitable for the understanding of the needs of the populations followed up and for the development of social assistance work. In addition, technical action must be contextualized in the social policies in which they are included, maintaining the ethicalpolitical horizon of professional practice committed to the guarantee of rights. This text-based on the theoretical framework of the Marxist tradition-intends to initiate a reflection on the creation of a professional project for occupational therapy, considering the ethical-political dimensions brought to light through work in social assistance, as well as of the societal project to be supported. To this end, the text is organized into three parts: the first addresses how the expressions of the social question determine the demands observed in social assistance work; the second discusses the concepts of societal and professional projects; the third proposes a debate on the construction of a professional project for the work of occupational therapists in the social assistance sector, outlining the contributions of the theoretical-methodological framework of social occupational therapy. --- The Expressions of the Social Question and Social Assistance In social assistance, professional work is required to achieve politically agreed social protection in the country. From a Marxist perspective, it is about intervening in the expressions of the social question that determine the production and reproduction of the social life of the working class. According to Paulo Netto , the term social question began to be used in the third decade of the 19th century to refer to pauperism since, for the first time in history, poverty grew at the same rate as the production of wealth, enabled mainly by the technological advances resulting from the Industrial Revolution. Pauperism differed from the poverty of previous societies, as there was not only an unequal distribution of resources, but also a scarcity of these resources due to the low development of technological conditions to produce material goods in society. The social question is, therefore, a phenomenon typical of capitalism, and it continues to exist even in a society of abundance, where there are technical conditions to eliminate it . In general, the social issue can be defined as the set of expressions of inequalities in a capitalist society that are rooted in its structural logic, which is based on the exploitation of classes and the appropriation, by a minority class, of the wealth produced by a working majority . In society, the social question is materialized in the everyday life of people through different expressions: poverty, misery, violence, unemployment, indigence, inequality, lack of social policies, etc., which, in the capitalist mode of production, require the structuring of social policies to face it, one of which is the social assistance policy -the focus of this study. Paulo Netto emphasizes that the social question can be discussed from different theoretical and political perspectives that can be categorized into two large blocks: conservative and critical-revolutionary. The first block interprets it as a natural phenomenon that can be eliminated in a capitalist society, either through moral measures or through public management planning in the allocation of resources in social policies. The second block understands it as impossible to be eliminated under capitalism because its origin is linked to the production logic of this system, which continually replaces its various expressions. Netto also points out that the recognition both of the structural relationships between the social question and capitalism and that of the impossibility of eliminating the former within the framework of the latter does not nullify the importance of its reformist measures of confrontation through social policies. He claims that: Confronting the social question without considering its economic and social foundations is like drying ice. But this does not mean immobilism. [...] the alternative reform or revolution is wrong. It is not an exclusive alternative. The adequate formulation of the problem would be reform and revolution, for a simple reason: historical experience has shown that struggles that had as a target, as objective, only the recognition of political and social rights had extremely significant impacts on the set of capitalist social organizations . In Brazil, social assistance actions to deal with the expressions of the social question have historically been anchored in the assistance, clientelist and disciplinary logic, which consists of sporadic, fragmented, and discontinuous actions, with little or no professionalization, based on benevolence and goodwill, and are implemented according to the interests of elites and political groups . The universalization of rights in the 1988 Constitution marks the institutionalization of another logic for this and other policies, now understood as a claimable right of the populations that need it and as a duty of the State to guarantee their access to it . Its consolidation as a public, universal, non-contributory social policy of State responsibility is built from some juridical-legal frameworks: the Organic Law of Social Assistance of 1993, the National Policy of Social Assistance of 2004, and the Basic Operational Norm of 2005, which implemented the Unified Social Assistance System . The PNAS aims to ensure social protection for people and their families and provide them with their minimum social needs through organized actions, in addition to acting in an integrated manner with other sectoral policies to combat poverty, meet social contingencies, and universalize social rights . These objectives demand work that has the horizon of building a radically democratic and less unfair society. Silveira points out that qualified social work is "the fundamental mediation in response to human needs, to the repercussions of structurally unequal logic, in facing the social question". The work at SUAS is performed by professional categories established by Resolution No. 17 of the National Council for Social Assistance in 2011, which include occupational therapists, who can be part of the reference teams and/or compose the SUAS management, contributing to carrying out social assistance services . It is understood that the professionals who work at the PNAS respond on the front line to the expressions of the social question in the life of the population followed up. Moreover, Silveira reflects that working in social policy services involves actions that democratize and qualify access to rights, requiring alignment between technical, ethical, and political qualifications, because operating social assistance policy involves recognizing and knowing the contradiction of its nature in mediating the expressions of the social question. Social policies are instruments through which, in a capitalist society, the State responds to the social needs of access to public goods and services of the working class. The characteristics of these policies are directly determined by the ideologies that underlie the understanding of the role played by the State in society, the perspectives on citizenship and social rights, and the ways of understanding and facing the social question . As noted by Lopes , "despite not eliminating the regime's exploitative devices", social policies-and, consequently, the social work resulting from themcombine not only technical dimensions but also, and fundamentally, ethical and political dimensions, since professional action is part of professional projects that operate, in a dialectical way, with ideologies that support projects for society. --- The Inseparability Between Corporate Projects and Professional Projects According to Paulo Netto , societal projects can be understood as collective projects that express class intentions for society, even if they have determinations of other natures, such as cultural, ethnic, gender, etc. These projects present an image of society to be built, advocate certain values to justify this image, and privilege certain means to materialize it and gain the adherence of citizens. Thus, societal projects are essentially political, involve power relations, and are flexible, renewing themselves according to historical and political circumstances. In concrete reality, we observe that the neoliberal project is the currently dominant and ongoing societal project in Brazil. Worldwide, it has gained visibility as of the 1970s, being employed as the main strategy to overcome the structural crisis of capitalism. The neoliberal project is characterized by a resumption of the fundamentals of economic liberalism, which emphasize individual freedoms, criticize State intervention, and praise the virtues of the market. It is mainly characterized by the idea of a smaller State and a larger market, encouraging the transfer of State responsibilities to the private sector. It advocates that the role of the State is to promote favorable conditions for individual competitiveness, without interfering in the spheres of public and economic life. From this perspective, social policies are seen as obstacles to individual initiatives and the market, and should only be aimed at those who cannot access goods and services via the market . In addition to these aspects more suited to the economic dimension, the neoliberal project also brings up new values and rules of behavior aimed at responding to the needs of capitalist production and reproduction. In the ideological field, fragmented and superficial concepts about the most diverse aspects of social life are disseminated, with attacks against critical perspectives of understanding reality in an attempt to devalue ideas that highlight the inherent contradictions of capitalism. Thus, the strengthening of conservative agendas, the criminalization of social movements, and the criticism of social rights unfold in social life . In Brazil, neoliberal ideas were introduced in the 1990s; however, in recent years, they have taken on a harder line. Since the Temer government and that of his successor Jair Bolsonaro , there has been a long cycle of strategic reforms in favor of the market, carried out under the allegation of warming up the economy, in which the working class and the poorest populations are the most affected . In this context, although we recognize the normative advances of social assistance as a social policy, including the professionalization of work in the sector, there are many contradictions and challenges present in its materialization in the context of the advancement of the neoliberal societal project. Social assistance, which was already the fragile link in social security policies, is experiencing critical situations when it has to establish itself as a public policy in the context of dismantling rights and social services. Campos & Ferraz emphasize that the PNAS has been suffering from the budget constraints that have caused the precariousness of services and work at SUAS, the greater focus of actions, as well as from the moralizing attacks on poverty and the poor in Brazil, governed by an unprecedented moral and religious fundamentalism in our new and fragile democracy. These trends, which do not concern only the PNAS, express a project for a conservative and liberal society that re-updates historical issues of the development of social assistance in the country. Faced with the validity of a societal project that directly attacks the populations assisted by occupational therapists in social assistance, deteriorates the services where we work, and dismantles our rights as workers, there is an urgent need to discuss professional projects for work at SUAS. Professional projects are also collective projects, but related to professions2 . They are composed of values, delimit ethical and political principles, offer theoretical and methodological frameworks for professional practice, and establish bases for the relationships with service users, other professions, and organizations and institutions . With these elements, they are not simply professional action plans for a given intervention, but rather broader projects that explicitly include ethical and political dimensions for a profession and define its intentions toward society. In other words, professional projects are in liaison with societal projects, either reinforcing the maintenance of the current order -having a conservative character, or advocating the rupture of that order as the ethical-political horizon of a profession -being criticaltransformative. Thus, in the words of Paulo Netto : Professional projects present the self-image of a profession, elect the values that legitimize it socially, delimit and prioritize its objectives and functions, formulate the requirements for its exercise, prescribe norms for the behavior of professionals, and establish the bases of their relations with the service users, other professions, and private and public social organizations and institutions . Just like societal projects, professional projects are dynamic, as they respond to economic, historical, and cultural changes, as well as to those in the social needs with which professions work. They are also modified according to their theoretical and practical development. We emphasize that given the heterogeneity of the members of a professioncomposed of subjects from different origins, positions, and social expectations, theoretical, ideological, and political behaviors, and preferences are present in the same category of diverse individual and societal projects, configuring it as a plural space from which different professional projects can arise . Thus, even if a project is institutionalized, as professions are a field of struggle, there will always be tension to build new projects in opposition to the dominant project. This debate leads us to reflect on how we can discuss this idea in occupational therapy, especially by encouraging reflection on a professional project for the work of occupational therapists in social assistance. --- Ethical-Political Commitment, Professional Project, and Occupational Therapy We start from the understanding that occupational therapy is not only the result of the interests of the subjects who have built it as a profession, but it also responds to a need for the social and technical division of work in monopoly capitalism, acting within the scope of class relationships and interests. This concept has been used by several authors in the field since the 1980s, including Soares , Lopes , andBezerra . From this perspective, there is also the understanding that, by acting mostly through social policies, occupational therapists respond to antagonistic interests of the dominant class and the working class, which are also expressed in the State, but directly serve the latter, which uses the social services to minimize the impacts of expressions of the social question in the everyday life of members. Therefore, professional work occurs in a concrete and contradictory social context, with a confluence of factors that interfere in professional practice, such as the configuration of social policies at a given historical moment, social legislation that can expand or restrict rights, configurations of power relations in the workplace, working conditions, employment relationships that allow greater or lesser autonomy in professional practice, among other issues . These dimensions concretely and objectively constitute the professional practice, setting limits but also offering possibilities for work capable of promoting changes in the reality of the populations followed up . We understand that, for this, the work must be guided by a critical-transformative professional project. The transformation of the reality of the assisted populations depends on the capacity of the occupational therapist to make a critical reading of their demands and needs and exercise their professional practice to transform them and their everyday life, adopting values such as freedom and human emancipation as the ethical horizon of this action. Thus, as professionals working amid contradictions and limits of State social policies, the work of occupational therapists involves provoking individuals and groups to identify the contradictions in which they live, and "these subjects return to the social environment, beyond the moment of occupational therapy follow-up, with new questions", enhancing their possibilities of transforming the social processes that determine their particular condition, which at the same time is collective because it is a manifestation of social reproduction . Thus, Godoy-Vieira proposes the notion of radical participation as a product of the work process in occupational therapy, which focuses on transforming the way individuals place themselves in the world through their daily activities, considering the dialectic between singularity and social totality. Farias & Lopes discuss the construction of an occupational-therapeutic thinking/doing that is attentive to combating structures of oppression and intended for freedom. Taking the everyday life of subjects and groups as central, those authors propose thinking/doing in occupational therapy along three processes: production of righteous anger that causes indignation with the inequalities and violence experienced, understanding that, in indignation, one can produce movements directed to freedom; suspension of everyday alienation, which produces reflections on the contradictions experienced, having a critical understanding of what is experienced; transformation of reality through the appropriation of the means to exercise citizenship and produce life in a critical way before the world. Thus, it is understood that, in this awareness movement regarding oppression processes, there is the possibility of building action for the "transformation of concrete reality, even within the imposed limits" . In this sense, despite the existence of structural determinations that impose limits to work in social policies, including social assistance, Rizzotti states that the professional condition of the members of the SUAS reference teams, characterized by the intellectual dimension of the work, safeguards a field of freedom and autonomy. The mixture of technical knowledge and ethical-political commitments forms, in daily work, a field where it is possible to combine and define common objectives capable of integrating team professionals, enabling the construction of a social assistance worker identity . This SUAS worker identity goes beyond a certain professionality and dialogues with societal projects. This implies that, among the professional categories of SUAS, there is a sharing of a worldview that values freedom and emancipation of the assisted subjects, equality in differences, real conditions of social justice, and contextualized understanding of vulnerability and social risk, allowing the overcoming of the differences arising from the formation of their origins . In occupational therapy, the debate about professional projects is not common; however, the profession is concerned with the need for an ethical-political commitment in professional practice. This discussion started in the 1970s, when, under the influence of Franco Basaglia's formulations about the technician's social mandate inspired by the debate on the organic intellectual in Antonio Gramsci, occupational therapists began to reflect on the role of the delegates of society's hegemonic values, exercised by its technical practice in the total institutions . From this period, there is also the criticism of the medicalization practices to the social question, the institutional segregation of population groups, and the social control implicit in technical work. This movement of criticism of and reflection on the ideological and political implications of professional practice and the approach and follow-up of people and groups under conditions of social vulnerability and risk unfolded, according to Bezerra , in a critical turn taken by Brazilian occupational therapy toward recognizing the close relationship between the technical, ethical, and political dimensions of professional work. Galheigo et al. states that this historical moment marks the beginning of the "incorporation of critical thinking into Brazilian occupational therapy, inaugurating a process of decolonization of its knowledge and practices, and shifting from a purely technical and procedural role to a social position with technical-political commitment". Reflections on the ethical-political commitment of occupational therapists involve, in general, the questioning of scientific knowledge as a statute of truth, the problematizing interpretation of the macro processes that directly affect the everyday life of the assisted populations, the understanding of the social question and its expressions from a sociohistorical perspective, the appreciation of diversity and culture, and the legitimation of different types of knowledge, in addition to the recognition of inequality in power relations, with a horizon transform society . These elements that constitute the ethical-political dimension in the profession are recurrently discussed in a direct combination between the structural and relational dimensions, understanding that this commitment occurs: through empathetic listening to the subjects and groups with whom it develops projects and through a critical perspective of social and political contexts. Empathetic for accepting the ideas, affections, and experiences of these subjects; critical for the problematizing interpretation of the macro processes in which their everyday life is intertwined. An ethical commitment, for intervening in the plan of life, in its movements of resistance and affirmation; a political commitment, for the continuous explication of the existing macro and micropolitical power games, for the defense of autonomy, citizenship, and rights, and for the search of new strategies to build and/or strengthen social groups . These elements, contained in the idea of ethical-political commitment, dialogue with and integrate the discussion of a professional project. However, the concept of a professional project also involves, as pointed out by Yazbek , explicit directions for the category regarding its ethical norms, professional rights and duties, politicalorganizational resources of the profession, and education. We understand that the ethical-political commitment needs to be included in a collective professional project that, in turn, is related to societal projects. The concept of professional project advances concerning the idea of ethical-political commitment because, depending on how this idea is incorporated by professionals, there is a risk that it will be assumed as a set of individual values for professional action. On the other hand, the professional project is a collective consensus around a set of issues that are even expressed in documents of the professional category . Given the plurality that constitutes a profession and the disputes around which a professional project will be institutionalized by the category, in addition to the debates and constructions previously carried out in social assistance by other categories that work in the sector, it is opportune to reflect on which professional project occupational therapy needs to build to work on this policy to respond to the social needs of the assisted populations. We advocate that the work of occupational therapists in social assistance should be guided by a critical-transformative professional project; but critical of what? Critical of the structure and dynamics of capitalist society that continually replace the expressions of the social question, creating impediments to social participation and touching the everyday life of subjects; critical of the neoliberal model of State management that impacts the PNAS, weakens rights, and makes people's lives and work precarious; critical of the conservative trends of work in social assistance that impact the PNAS, individualize, medicalize, psychologize, and moralize the understanding of problems presented as professional demands; critical of the role and social function of occupational therapy in its relationship with society, breaking with a technical and supposedly neutral view of the profession, and one that assumes freedom and human emancipation as core values. We can state that a critical-transformative project enables occupational therapists to become aware of the macro social issues that cross their work at SUAS and, with that, be able to provide their actions with differentiated technical quality, aiming to defend social and human rights, considering them as historical achievements of the struggles of social movements. In other words, this project connects and responds to a demand from classes subordinated by capitalism, the main demanders of the work of occupational therapists in social assistance. This differentiated quality is manifested through a practice that favors the universalization of rights and that questions focused criteria, which treat rights as privileges; that democratizes access to information that can strengthen subjects and groups, favoring their political mobilization; which reframes institutional requests that often call for uncritical and emergency responses to user demands; that does not comply with institutional productivism and the fulfillment of merely bureaucratic demands. The criticism we are referring to is only possible through the theoreticalmethodological framework adopted within the professional project because, according to Yazbek , knowledge and theoretical choices, the design of research and investigations, and practical choices are components that fundamentally comprise professional projects and give them a conservative or critical-transformative characteristic. In occupational therapy, different perspectives and frameworks have characteristics that can inform a critical-transformative professional project: discussions on occupational therapies of the Global South , the so-called critical perspectives in occupational therapy , community occupational therapy , discussions about human rights , the feminist perspective , Brazilian social occupational therapy , among others. A non-systematic approach to the literature in the area shows that investigations addressing the theme of occupational therapy at SUAS, to a large extent, incorporate the social occupational therapy framework. We have found from the history of militancy to include occupational therapists as a working category at SUAS , to the continuous bibliographical productions on professional practices and the development of strategies, resources, and social technologies to intervene in the various devices of the SUAS network . According to Malfitano , this framework focuses on working with individuals and groups whose socioeconomic condition determines, in their everyday life, difficulties and impediments to social inclusion and participation in conditions of equal rights. It advocates, in a dialectical movement, the inseparability between the micro and macro social dimensions to understand the reality of the assisted population, in a work logic that abdicates individualizing interpretations and interventions, but does not cease to act in the individual sphere: In the individual sphere, strategies that assist the subjects to obtain support for their social inclusion and participation will be developed, respecting their choices and autonomy. In the collective sphere, work is connected to social policies and action in public spaces by maintaining and/or expanding the social recognition of certain needs, and intervening by expanding services and other spaces that allow access to the rights of that group . From this theoretical place, social resources and technologies are proposed to operationalize the work for acting in these two spheres: a) Workshops of Activities, Dynamics, and Projects: professional actions that conceive the activities as a mediating resource to approach, follow up, apprehend the demands, and strengthen the individuals and groups assisted by occupational therapists; b)Individual Territorial Follow-ups: strategies that enable a more real perception of and interaction with the everyday life of the subjects, interconnecting their histories and trajectories, current situation, and network of relationships; c) Articulation of Resources in the Social Field: a range of actions carried out with subjects, groups, and policy management involving the use of possible resources, understood as financial, material, relational and affective, either in the micro or macrosocial spheres, to compose the interventions; d) Dynamization of the Service Network: mapping, disseminating, and consolidating all programs, projects, and services aiming to encourage interaction and integration between them, combining the different sectors and levels of intervention . These work methodologies dialogue with the objectives of the social assistance policy, which proposes to provide groups, families, and individuals who need social protection with guaranteed security: survival security , acceptance security, and family life security . Furthermore, the construction of work from the principles and methodologies of social occupational therapy, associated with the recognition of the inadequacy of health models to inform the practice in social assistance, avoid interventions that individualize, medicalize, and psychologize the expressions of the social question because, as pointed by Almeida & Soares , the "products of social inequality have repercussions on different levels of human experience, forcefully and devastatingly penetrating the large and small movements and actions of everyday life". Thus, social occupational therapy is a theoretical-methodological framework that can be adopted to construct a critical-transformative professional project for work in social assistance. Among its contributions, this framework substantiates a professional practice detached from the health area, discusses the contradictory role of social policies, combines the micro and macro social dimensions in understanding demands and social reality, presents social resources and technologies with a rationale consistent with the work required at SUAS, advocates the inseparability of the technical, ethical, and political dimensions of professional practice, and makes use of theoretical references common to the professions that work in this sector, thus providing a common ground discuss social phenomena. --- Final Remarks The constitution of occupational therapy as a profession is linked to the unfolding of the social question and State strategies to face its expressions, which are presented as demands for the work in social assistance and which, from the Marxist perspective, are insoluble in the order of capital. The work at SUAS has an ethical-political dimension and is influenced by ideologies that support societal and professional projects, which can be conservative or criticaltransformative. The purpose of thinking about professional projects for SUAS lies in the need to: ethically guide the professions [...] by streamlining politics, adding anticipated purposes. While overcoming historical marks, it can strengthen work committed to the expansion of rights, to human emancipation . We argue that social occupational therapy is a theoretical-methodological framework that can coherently subsidize the construction of a professional project for occupational therapy at SUAS, in line with the essential and fundamental skills for working in the sector. This can favor the category with greater incorporation and recognition of its professional practice in social assistance and with the overcoming of its identity as an exclusive health profession, enabling the construction of a working identity shared with the other professions that compose the work teams in the sector. --- Author's Contributions Waldez Cavalcante Bezerra was responsible for the idealization and initial thoughts that gave rise to the text. Ana Carolina de Souza Basso contributed to the maturation of these reflections. All authors approved the final version of the text.

Occupational therapists compose technical teams in social assistance services before the formalization of their insertion in the Unified Social Assistance System (SUAS), which occurred in 2011. As a result, the discussion about the work in this sector and the social question has gained relevance in the profession from an interventional and theoretical-academic point of view. In this study, from a Marxist perspective, we reflect on how the social question determines the demands observed in social assistance work. Next, we present the concepts of societal projectsunderstood as collective projects that express class intentions for society, and of professional projects -defined as collective projects related to professions composed of ethical and political principles, which are theoretical and methodological references for professional practice, and that establish bases for their relationships with service users, other professions, and organizations and institutions. We conclude by proposing that social occupational therapy is a theoretical and methodological framework that can be adopted to construct a critical-transformative professional project: critical of the structure and dynamics of capitalist society that continually replace the expressions of the social question, touching the everyday life of subjects; critical of the neoliberal model of State management that weakens rights and makes people's lives precarious; critical of the conservative trends of work in social assistance that individualize, medicalize, psychologize, and moralize the reading of problems presented as professional From an ethical and political commitment to the design of a professional occupational therapy project: a debate necessary for social assistance work Cadernos Brasileiros de Terapia Ocupacional, 31(spe), e3387, 2023 2 demands; critical of the role and social function of occupational therapy in its relationship with society, breaking with a technical and supposedly neutral view of the profession.

Introduction The rate of genetic evolution viewed over a long time is estimated by counting the number of mutant substitutions and dividing by the elapsed time. Similarly, for studies of cultural evolution based on empirical observations in the archaeological record or inferences from differentiation of contemporary populations, the accumulated differences in many cultural traits are examined to obtain an estimate for the rate of evolution . Drawing on this analogy, we define the long-term theoretical rate of cultural change, R, as R = Nuπ 1 . Eq. is based on the ''infinite site model'' of population genetics, where in the current context N is the population size, u is the innovation rate per individual per generation, and π 1 is the fixation probability of an innovation that is initially made by a single individual. Although this analogy clearly has its limitations , Eq. is conceptually useful because it provides a summary statement relating three important factors that contribute to the cultural evolutionary rate ). More generally, if there are m different social roles , the long-term cultural evolutionary rate should be written as R = m - i=1 N i u i π 1i , where N i is the number of individuals in role i, u i is the innovation rate of an individual in role i, and π 1i is the fixation probability of an innovation made by an individual in role i. In short-term cultural evolution, on the other hand, as illustrated by the ''diffusion of hybrid corn'' , interest attaches to the pattern and timing of the spread of a particular innovation. Here, an appropriate theoretical measure of the rate of cultural evolution is the mean 0040-5809/$ -see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.tpb.2011.02.001 time until fixation . Results on the mean fixation time and mean absorption time are presented for the sake of completeness, but the emphasis in the present paper is on the fixation probability, which determines the long-term rate of cultural evolution. It is also of interest to investigate the patterns of accumulation of cultural traits and the amount of cultural variation that are expected to be maintained within a population when the same innovations to a cultural trait occur repeatedly. Using the methods of Lehmann et al. , we evaluate here the mean number of distinct cultural traits carried by an individual and segregating in the population, as well as the expected level of cultural heterogeneity among individuals at steady state of the cultural accumulation dynamics. A major difference between a gene and a cultural trait lies in their modes of transmission. Whereas genetic information flows solely from parent to child , the inheritance of a cultural trait is not necessarily limited by the degree of relatedness. In addition, some cultural traits may be acquired under the concerted influence of a group of individuals , while for others one particular individual may serve as the exemplar for a large number of individuals . Semi-quantitative arguments have been invoked to predict the theoretical rates of cultural evolution under various modes of social transmission . For example, cultural changes under many-to-one transmission and one-to-many transmission are described as being ''very slow'' and ''very rapid'', respectively. Archaeologists also discuss the amount of cultural variation within a population . In particular, they agree that MtO transmission would result in very low ''cultural variation between individuals''. A major motivation for the present paper is to reevaluate these claims. Theoretical studies have also addressed the consequences of choosing the most adaptive or attractive variant of a cultural trait from among k exemplars . Rates of popular culture change have been investigated, assuming oblique transmission and recurrent innovation in a finite population with discrete generations and simultaneous reproduction using a statistical framework that originated in population genetics ). An important parameter in studies of cultural evolution is the size of the population or social group in which the transmission of cultural traits occurs. Here we are principally interested in interpreting the rates of cultural change and amounts of cultural variation in Pleistocene hominid populations. Hence, we use band size in current hunter-gatherer societies, which averages perhaps a total of 25 individuals of both sexes and all age groups . However, bands do not exist in isolation, and a ''local community'' may be formed by neighboring bands among which networks of social transmission presumably extend. Perhaps five bands comprising approximately 125 individuals are tightly linked into a local community These two estimates of social group size are used to compare the rates of cultural change under different modes of social transmission . In this paper, we obtain detailed quantitative predictions on cultural evolutionary rates and on cultural variation within a population. Stochastic models based on the Moran model are suitable vehicles for this analysis . There are several reasons why this approach is preferable to one based on the Wright-Fisher model . First, Moran-type models are mathematically simpler since they involve asynchronous as opposed to synchronous updating. Second, they result in overlapping generations. Third, they allow an in-depth treatment of OtM transmission . Our major findings are as follows. First, the type of MtO transmission considered in this paper is conformist transmission . With the specific functional form of conformist transmission we assume -), the fixation probability of an innovation is extremely small. In other words, cultural change under MtO transmission is predicted to be very slow ). In addition, this social learning rule results in a very low number of distinct traits carried by an individual and a moderately low number present in the population, but a high level of cultural heterogeneity under the process of cultural accumulation . Second, our best-of-k model posits that each newborn prefers the new cultural variant over the old and adopts the former provided it is carried by at least one individual among k exemplars sampled from the population. Fixation probabilities of the innovation are quite large and depend little on whether it is selectively neutral, adaptive, or maladaptive . With this social learning rule, individuals and the population are both predicted to accumulate a high number of different traits on average, and the cultural heterogeneity is relatively low. Third, in our OtM transmission model, one individual has the special status of teacher and -as long as it remains alive -is copied by each newborn individual. The fixation probability of an innovation then depends on whether it is made by the teacher or some other individual. In the former case the fixation probability is high, but in the latter it is lower than with vertical, oblique, or horizontal transmission -, and Table 4). When Eq. is applied, we do not observe a substantial acceleration of the long-term cultural evolutionary rate unless teachers are highly innovative . For cultural accumulation, OtM transmission results in individuals and the population carrying a low number of distinct traits and a high level of cultural homogeneity. Before proceeding, we wish to point out that the specific models of cultural transmission we introduce and analyze in this paper make extreme simplifying assumptions. This allows us to rigorously demonstrate the large effect that different modes of social transmission can have on cultural evolutionary rates and cultural variation . Predictions based on extreme functional forms do not give realistic cultural evolutionary rates, for example, even if accurate estimates of the innovation rate are available. Hence, the values derived from our models should be used only in a comparative fashion. Moran proposed a model of random genetic drift, which we modify as follows to describe cultural evolution. Assume a finite population comprising N individuals. A cultural trait exists in two variant forms A and B, with i individuals of type A and Ni of type B. At each arbitrary time step, an individual is born and acquires one or the other of these types by social learning , followed by the death of an individual other than this newborn . If the newborn copies a randomly chosen member of the population-oblique or horizontal transmission, there being no conceptual distinction between the two in the Moran model-it acquires type A with probability b A = i/N and type B with probability b B = /N. Similarly, if death occurs at random, the dying individual will be of type A with probability d A = i/N and type B with probability d B = /N. The probability that after this birth-death event the number of type A individuals has increased by one is --- The Moran model and vertical, oblique, or horizontal transmission p i = b A d B = i/N 2 , the probability it has decreased by one is q i = b B d A = i/N 2 , and the probability it has not changed is 1 -p i -q i = 2 + [/N] 2 . Interestingly, these transition probabilities also apply when social learning is vertical -the newborn acquires the cultural type of one of its parents -provided we assume equal fertilities. For brevity, we will refer to this model of social transmission as the oblique transmission model. Differential mortality can be incorporated by setting the mortality of type B to be v times that of type A . Then, the probabilities of death for type A and B individuals are d A = i/[i + v], and d B = v/[i + v], respectively. In Appendix A we present general formulas from Ewens for: the fixation probability of type A, π i ; the mean absorption time, ti ; and the mean fixation time of type A, t * i , if there are initially i individuals of type A. These values can be computed for the oblique transmission model from the transition probabilities Eqs. -. In particular, the fixation probability of type A is π i = i/N in the selectively neutral case , and π i = [1 - i ]/[1 - N ] when there is viability selection . The appropriate formula to be used in computing the long-term rate of cultural evolution is Eq. , since social roles are not distinguished in this oblique transmission model. In the selectively neutral case, Eq. reduces to R = u. Of particular interest is the case where type A is initially represented once in the population, as when it arises by innovation or is introduced by one individual from an external source. The tables illustrate the fixation probabilities, mean absorption times, and mean fixation times for this special but important instance. Table 1 gives results for the oblique transmission model. These values serve as baselines against which the accelerating or decelerating effects of the alternative modes of social transmission we investigate can be compared. Note the equivalence of t * 1 for v = 49/50 and 50/49 and also for v = 9/10 and 10/9. This result -the mean fixation time of type A, initially represented once in the population, depends on the intensity but not the direction of selection -is proved in general in Appendix B. --- Many-to-one transmission As before, let variant cultural types A and B exist in numbers i and Ni, respectively. Suppose each newborn samples k exemplars without replacement from the population of N individuals . The probability that there are j individuals of type A and kj individuals of type B in this sample is given by the hypergeometric distribution, h =  i j   N -i k -j   N k  , where h = 0 if j < 0, j > i, j > k, or j < k -. Pure conformist transmission entails that each newborn adopt the majority cultural type among its k exemplars. Here we modify this condition slightly by introducing a small probability proportional to ε of adopting the minority cultural type provided there is at least one individual of this minority type among the k exemplars. Moreover, when both cultural types are equally represented in the sample, it is natural to assume that the newborn will adopt either type with probability one-half. Then, when k ≥ 2 is even, the newborn will copy type A with probability b A = k/2-1 - j=0 hjε + h/2 + k - j=k/2+1 h[1 -ε], = k/2-1 - j=0 h + h/2 + k - j=k/2+1 hε. Similarly, when k ≥ 3 is odd, the newborn will copy type A with probability b A = /2 - j=0 hjε + k - j=/2 h[1 -ε], and type B with probability b B = /2 - j=0 h + k - j=/2 hε. Note that Eqs. - entail a more extreme form of conformity than the often-used sigmoid function . If the mortality of type B is v times that of type A, the up-and down-transition probabilities for this model are The fixation probability, mean absorption time, and mean fixation time of type A, when there are initially i individuals of this type, can be computed from the formulas of Appendix A. Illustrative results are shown in Table 2 for the case of i = 1. The values 3-5 for parameter k represent the likely number of adults in a band who may serve as exemplars for the newborn. p i = b A d B and q i = b B d A , The most salient feature of Table 2 is the low fixation probabilities. Conformist transmission entails that a newborn is unlikely to adopt the minority cultural type. Hence, an innovation that is initially represented once in the population has only a small chance of spreading through the population. Table 2 shows a sharp drop in the fixation probabilities as the number of exemplars, k, increases from 4 to 5. There is an even larger effect of the population size, N. In this model also, the long-term evolutionary rate is given by Eq. . The mean fixation times are shorter than with oblique transmission, indicating that in the rare cases when fixation occurs, spread of the innovation is fast. When viability selection is imposed, the mean fixation time is identical whether cultural type A is selected for or against . This was also observed in the oblique transmission model. --- Best-of-k transmission As in the previous model, each newborn samples k exemplars from the population of N individuals with i of type A and Ni of type B. Suppose that type A is preferred over type B. Specifically, type A is adopted by the newborn provided there is at least one individual of type A among the k exemplars. Since we regard type A to be the newly arisen cultural type , this implies a pro-novelty bias. The probabilities of adoption of type A and B can be written, respectively, as b A = 1 -  N -i k   N k  , b B =  N -i k   N k  , if i ≤ N -k, and b A = 1, b B = 0, if i ≥ N -k + 1. These assumptions entail an extreme form of pro-novelty bias. Eq. implies that the down-transition probability, q i = b B d A , is 0 when i ≥ Nk + 1. As a result, Eqs. - of Appendix A are ill-defined. Revised formulas applicable to this model are derived in Appendix C and used to calculate the entries in Table 3. Fixation probabilities are quite high whether the preferred cultural type is selectively neutral, adaptive, or maladaptive. Moreover, in contrast to the oblique and MtO transmission models, fixation probabilities are not substantially lower for the larger population size. Mean fixation times are also relatively short. --- One-to-many transmission Assume that the population of size N contains two selectively and preferentially neutral variant cultural types A and B. One individual in this population has the special status of teacher. At each time step, one newborn is produced and adopts the cultural type of the current teacher, followed by the death of one random individual excluding the newborn. If the teacher is the one to die, its social role is taken by another individual, randomly chosen from among the survivors including the newborn. As with the other models considered in this paper, the OtM transmission model entails some extreme assumptions. For example, there is just one teacher that is copied by all newborns while it remains alive. Nevertheless, this assumption receives some empirical support in the case of the transmission of stonetool-making. Based on the evidence from refitted débitage at the Magdalenian site of Etiolles , Pigeot argues that novices may have turned to skilled knappers for instruction. Moreover, she suggests that there may have been just one specialist who served as teacher to all other individuals at this site. The state of the population is completely described by the number i of type A individuals and the type α of the current teacher. We require 1 ≤ i ≤ N when α = A and 0 ≤ i ≤ N -1 when α = B. States 0B and NA are absorbing, whereas the remaining 2 states are transient. Let p iα,jβ be the transition probability from state iα to state jβ. Then, the assumptions of the paragraph before last entail that the positive transition probabilities are p iB,B = i/N, p iB,iB = /N + /N 2 , p iB,iA = i/N 2 , p iA,iB = /N 2 , p iA,iA = /N + i/N 2 , p iA,A = /N, when 1 ≤ i ≤ N -1, and moreover p 0B,0B = p NA,NA = 1. For example, p iA,iA ) is obtained by noting that the newborn will be type A since the teacher is type A , and one of the i -1 type A individuals that is not the teacher dies /N), or the teacher dies and its role is adopted by a type A individual . The fixation probabilities for this model are π iB = i/2N if initially there are i individuals of type A and the teacher is type B, and π iA = /2N if initially there are i individuals of type A including the teacher . In the latter case, the fixation probability exceeds one-half for all population sizes ). With a type B teacher, on the other hand, the fixation probability is one-half that for the oblique transmission model with Eq. ). The mean absorption and mean fixation times from state 1A and 1B given in Table 4 are computed using the formulas in Appendix E. The OtM model posits two social roles, teacher and non-teacher. If the one teacher and each of the N -1 non-teachers innovate at rates u t and u nt , respectively, then from Eq. the long-term cultural evolutionary rate is given by R = [u t + u nt ] /, where we have used Eqs. and with i = 1. Clearly, we recover R = u if we substitute u t = u nt = u in Eq. , which parallels the standard result for oblique transmission that the rate of change of a selectively neutral cultural trait is independent of population size. --- Accumulation and maintenance of cultural variation with oblique transmission In order to address the question of the amount of cultural variation maintained in a finite population, we introduce a different model of cultural innovation which allows for imperfect copying. As before, in a population of size N, assume two cultural types A and B, where we now interpret A as the presence of some trait and B as its absence. At each birth-death event, a newborn may innovate with probability u , resulting in its acquisition of cultural type A. When it does not innovate, it acquires cultural type A by social learning with probability β . Then, with oblique transmission and i individuals of type A, the probability that a newborn is of cultural type A can be expressed as bA = u + βi/N, and the probability it is of cultural type B is bB = [/N + i/N] . The probabilities of death, d A and d B , are unchanged -). If 0 < u < 1 and 0 < β < 1, the stochastic process describing the evolution of the frequency of cultural type A -, -) has a stationary distribution whose properties refer to the cultural variation in the population. Considering a collection of c independent traits, from this stationary distribution we can evaluate the mean number of distinct traits carried by an individual, λ f ; the mean number of traits segregating in the population, λ p ; and the mean number of traits shared between two randomly chosen distinct individuals, λ s . We can also calculate the proportion of shared traits between two randomly sampled individuals, ϕ = λ s /λ f . More precisely, ϕ is the conditional probability that a trait carried by an individual is also carried by another individual. Hence, we may regard ϕ as a measure of the cultural homogeneity of the population , whereas 1 -ϕ is a measure of cultural heterogeneity . With a very large number of independent and selectively neutral traits , we have for large N λ f = U/, where U = cu is the total innovation rate. Moreover, In the first column of panels, λ f , λ p , and ϕ are graphed as functions of population size N for the manyto-one transmission model with U = 0.1 and β = 0.9. Curves from top to bottom correspond to ε = 0.01, 0.001, and 0.0001 , respectively. Similarly, in the second column of panels, λ f , λ p , and ϕ are plotted against N for the best-of-k transmission model, again with U = 0.1 and β = 0.9. From top to bottom, curves are drawn for k = 2, 4, and 8 . When N becomes large we consistently observe that ϕ convergences to β. Finally, in the third column of panels, values of λ f , λ p , and ϕ are given for the one-to-many transmission model, where in this case the curves from top to bottom are for β = 0.9, 0.7, and 0.5. All curves are graphed as functions of N for 1 ≤ N ≤ 25 because this allows for a good compromise between distinguishing the different curves in the panels and letting the variable ϕ approximate its limiting value. The qualitative increase of the remaining variables for larger values of N can be inferred from the panels. By comparison, for the oblique transmission model with U = 0.1 and β = 0.9, we have for large N, λ f = 1, λ p = 0.3 + 0.26N, and ϕ = 9/N. λ p = U 1 -β - NU β  log + ϕ 2  , (6 --- Cultural variation at equilibrium with alternative modes of social transmission With MtO transmission, the probabilities of birth of A and B, respectively, are The stationary distribution is obtained numerically for various parameter sets, which allows computation of λ f , λ p , and ϕ. bA = u + βb A , and bB = [b B + b A ] , Examples are presented in the first column of Fig. 1, where we see that the average number of traits carried by an individual, λ f , is very small . In addition, values of λ p are moderately low while values of ϕ are extremely low , where the latter entails that most individuals carry different traits, implying high cultural heterogeneity among individuals. These results are in qualitative agreement with our previous work . For Bok transmission, we substitute Eqs. - in Eqs. -. The stationary distribution is again obtained numerically for various parameter sets, with an example shown in the middle column of Fig. 1. We find that individuals carry a large number of traits on average , the population has a large number of segregating traits , and the proportion of shared traits is relatively high . Hence, the effects of pro-novelty bias on the accumulation of cultural variation are similar to sensitivity to the minority . Note also that ϕ appears to approach β as N becomes large , which was observed for all parameter sets tested. Finally, for the model of OtM transmission, the positive transition probabilities, piα,jβ , incorporating recurrent innovation and imperfect copying are as shown in Appendix F. From the stationary distribution of the frequency of type A, we can compute the statistics λ f , λ p , and ϕ for any values of N, u, and β. An example is given in the third column of Fig. 1: the average number of traits per individual, λ f , is low, while the average number of traits in the population, λ p , increases with N in an almost linear fashion. The bottom panel shows that the proportion of shared traits between two randomly sampled distinct individuals, ϕ, is high -hence the cultural heterogeneity 1 -ϕ is low -unless the copying fidelity, β, is low. This last result is an intuitively reasonable consequence of one-to-many transmission. --- Discussion In Table 5 we compare the long-term evolutionary rates for selectively neutral cultural variants expected under the various modes of social transmission. In computing R as defined by Eq. , we have arbitrarily assumed that each individual produces, or introduces from an external source, one innovation during a generation . In this case, the cultural evolutionary rate for oblique transmission is R = 1 regardless of the population size. Also, we have set k = 4 and ε = 0.001 where applicable. For the OtM transmission model where Eq. is appropriate, we have assumed that the teacher innovates at rate u t = 2, while each non-teacher innovates at rate u nt = 1. Oblique transmission here includes vertical and horizontal transmission, since in the Moran model either there is no conceptual distinction or their effects are identical. It is often claimed that more rapid cultural evolution is possible with horizontal than vertical transmission . This prediction follows from the assumption of discrete generations, which entails that horizontal transmission can occur repeatedly during a generation while vertical transmission occurs just once. With conformist transmission, the evolutionary rates are exceedingly low. The value of R = 3.38 × 10 -6 for N = 25 means that an innovation that ultimately becomes fixed arises, on average, once every 1 ÷ = 2.96 × 10 5 generations. This is approximately how long hominids have walked the earth. When the population size is N = 125, the mean interval between successful innovations is even longer. The functional form of Eqs. - and the setting for parameter ε result in a low probability that a minority cultural variant will spread. Larger values of ε or alternative functional forms for conformist transmission would produce higher evolutionary rates. Our numerical examples show how extreme the decelerating effect of conformist transmission can be and support the suggestion of Lycett and Gowlett that ''a many-to-one system could have operated . . . within Acheulean populations, at least . . . with handaxe manufacture'', which they made to explain the extremely conservative nature of this lithic tradition. Table 5 shows that Bok transmission, on the other hand, accelerates cultural evolution relative to oblique transmission. In the Bok model, each newborn has a preference for a newly introduced cultural variant , which it adopts regardless of its fitness effects, provided there is at least one individual of this type among the k exemplars. Table 3 records the results of a more detailed numerical analysis. We see that fixation probabilities are high even when only two exemplars are sampled and when type A is selected against . Hence, the cultural evolutionary rates are also high. . Tables 3 and5 also reveal that a larger population size implies a higher cultural evolutionary rate. This is because the fixation probability in the Bok model decreases with N, but at a rate slower than 1/N. OtM transmission has different consequences depending on whether the innovator is the teacher or some other individual. In the former case, the fixation probability exceeds one-half for all population sizes /2N), whereas in the latter case, the fixation probability is exactly one-half that with oblique transmission . If the one teacher and each of the N -1 non-teachers innovate at rates u t and u nt , respectively, then the long-term cultural evolutionary rate is given by Eq. -a weighted average of the innovation rates. In particular, if we make the arbitrary assumption as in Table 5 that u t = 2 and u nt = 1, we obtain R = 3/2 + 1/ ≈ 3/2. It is arguable whether teachers are more innovative than nonteachers. One may be a teacher by virtue of a large fund of knowledge, which may facilitate innovation. Similarly, the teacher may have many contacts outside the population and hence be more aware of externally occurring innovations. On the other hand, a non-teacher may be more willing to experiment with new techniques and styles. If all individuals are equally likely to innovate, then substituting u t = u nt = 1 yields R = 1, i.e., the OtM model predicts the same cultural evolutionary rate as the oblique transmission model. Thus, evidence for teaching such as has been found at the Magdalenian archaeological site of Etiolles cannot necessarily be invoked to explain the rapid turnover of lithic traditions during the Upper Paleolithic. Let us now turn to the question of the amount of cultural variation at equilibrium when there is recurrent innovation and imperfect copying of each trait. The various social transmission models we investigated produce remarkably different predictions. For example, the cultural heterogeneity or trait diversity, 1 -ϕ, expected under MtO transmission is appreciably greater than with oblique transmission. This is because a moderately but not negligibly small number of traits accumulate in the equilibrium population , but since each exists at a low frequency , they are unlikely to be copied. The result is a decrease in ϕ , or equivalently an increase in the trait diversity, 1 -ϕ. This prediction would appear to contradict prevailing theory in archaeology, where the intuitive claim is often made that MtO transmission will produce little cultural variation among individuals . If by little variation among individuals it is meant that artifacts produced by different individuals tend to share the same traits, then this claim requires revision, as we have shown. Lehmann et al. discuss in more detail why intuition should fail in this case. On the other hand, Fig. 1 shows that the mean number of traits segregating in the population, λ p , will be smaller than with oblique transmission. Hence, MtO transmission should result in reduced intra-site variability, in the sense that fewer cultural traits are present and variable at one location. By contrast, the Bok transmission yields large values of λ f , λ p , and ϕ relative to oblique transmission. Hence, the trait diversity, 1 -ϕ, will be low. This model of social transmission is similar to sensitivity to minority studied by Lehmann et al. . It also effectively constitutes a stochastic version of the model suggested by Enquist et al. , in which each individual has more than one cultural parent and traits are copied with probability β ). Fig. 2 shows the results for k = 2 and k = 5 and with β = 0.9, 0.7, 0.5 . The predictions of the two models are consistent; in particular the cultural homogeneity remains substantial for β ≥ 0.5 and is very high for β = 0.9. The other transmission scheme that we consider, OtM , represents the maximum possible level of cultural inbreeding , giving here an effective size of approximately 2.26, which is independent of population size . In other words, one and the same individual, i.e., the teacher, acts as the cultural parent for all individuals that are born during its lifetime. We obtain the intuitively reasonable result that a substantial level of cultural homogeneity is maintained for each of the three values of β . Thus, our teacher model - of Appendix F) does not require multiple cultural parents in order for cultural homogeneity to evolve. On the other hand, when just one cultural parent is chosen at random as in our oblique transmission model, cultural homogeneity will be low in a large population at equilibrium ), which agrees with Enquist et al. . Fig. 2. Measures of variation for a selectively neutral trait in the best-of-k transmission model. In the first column of panels, λ f , λ p , and ϕ are graphed as functions of population size N for the case of U = 0.1 and k = 2. The top to bottom curves correspond to transmission rates β = 0.9, 0.7, and 0.5, respectively. The second column of panels is the same as the first except that k = 5. --- Appendix C Derivation of mean times in the best-of-k transmission model. It is convenient to view the state space as comprising two overlapping subspaces, S 1 = {0, 1, . . . , Nk, Nk + 1} and S 2 = {N -k + 1, Nk + 2, . . . , N -1, N}. Eqs. - entail that S 1 has two absorbing states, 0 and Nk + 1, either of which can be reached from any of the remaining, Nk transient, states within S 1 . On the other hand, only up-transitions are possible in S 2 , and state N will eventually be reached . Let the initial state i lie between 1 and Nk. For j such that 1 ≤ j ≤ Nk, the mean sojourn time at state j conditional on fixation of type A can be computed from Eqs. - with N -1 replaced by Nk. Similarly, the mean sojourn time conditional on loss can be obtained from Eqs. , and . When Nk + 1 ≤ j ≤ N -1, the mean sojourn time conditional on fixation is equal to the mean waiting time until an up-transition, since down-transitions are impossible. Hence, t * ij = p -1 j . On the other hand, given that type A is eventually lost, we clearly have t * * ij = 0 when N -k + 1 ≤ j ≤ N -1. Thus, the mean fixation time from state i ≤ Nk has the form t * i = i - j=1  π j j-1 - l=0 ρ l   ρ j-1 q j   • N-k - l=i ρ l   π i N-k - l=0 ρ l  + N-k - j=i+1  π j N-k - l=j ρ l   ρ j p j   × i-1 - l=0 ρ l   π i N-k - l=0 ρ l  + N-1 - j=N-k+1 p -1 j , and similarly the mean loss time can be expressed as t * * i = i - j=1   1 -π j  j-1 - l=0 ρ l   ρ j-1 q j   × N-k - l=i ρ l   N-k - l=0 ρ l  + N-k - j=i+1   1 -π j  N-k - l=j ρ l   ρ j p j   × i-1 - l=0 ρ l   N-k - l=0 ρ l  . The mean absorption time can then be obtained from Eq. . --- Appendix D Recursions satisfied by the fixation probability of type A in the one-to-many transmission model. From Gale , the fixation probabilities satisfy the recursions -). Also, let I be the 2 × 2 identity matrix, and let T =  tiα,jβ  be the 2 × 2 matrix of mean sojourn times . State 1B must be visited when type A is ultimately lost. π iA = [/N] π A +  /N + i/N 2  π iA +  /N 2  π iB , Similarly, state A must be visited when type A is ultimately fixed. Hence, tiα,1B = t1B,1B 1 -π iα 1 -π 1B , tiα,A = tA,A π iα π A , where π iα is the fixation probability from state iα (see Eqs. -) . We have T = I . Equating the 1B, 1B elements on both sides yields N + 1 N 2 t1B,1B - 1 N 2 t1A,1B = 1. Then, using Eq. gives N + 1 N 2 t1B,1B - 1 N 2 • N -1 2N -1 t1B,1B = 1. Hence, t1B,1B = /2 and using Eq. again gives tiA,1B = /2, tiB,1B = /2. By symmetry, tA,A = t1B,1B = /2 and hence tiA,A = /2, tiB,A = i/2. Thus, the first and last columns of matrix T are readily obtained. The other elements can be obtained by reversing the order of matrix multiplication-from T = I . The mean sojourn times can be defined recursively by this method, but closed formulas are not generally forthcoming. In what follows, we first derive recursions in t1B,iB and t1B,iA and then in t1A,iB and t1A,iA . Using the former permits us to evaluate the mean absorption time and the mean fixation time from initial state 1B, and the latter gives us the corresponding values for state 1A. The elements of the first row of T = I are t1B,1B N + 1 N 2 + t1B,1A  - N -1 N 2  + t1B,2B  - 2 N  = 1 t1B,1B  - 1 N 2  + t1B,1A N 2 = 0 t1B,iB i N 2 + t1B,iA  - N -i N 2  + t1B,B  - i + 1 N  = 0 t1B,A  - N + 1 N 2 = 0. Eq. holds for 2 ≤ i ≤ N -2 whereas Eq. holds for 2 ≤ i ≤ N -1. N -i + 1 N 2  + t1B,iB  - i N  + t1B,iA( --- Appendix A The fixation probability and mean times for a finite birth-death chain with two absorbing states and with positive probabilities of transition from each of the transient states . Let and assume there are initially i type A individuals. Then the fixation probability of type A is In terms of the number of birth-death events, the mean sojourn time in state j is Finally, the mean fixation time, mean loss time, and mean absorption time are, respectively, --- Appendix B Equality of t * 1 with positive and negative selection in the oblique transmission model. Applying the formulas of Appendix A with ρ 0 = 1 and ρ l = l , we can write the mean sojourn time at j ≥ i conditional on fixation as which differs subtly from Eq. . Hence, the equality of mean sojourn times for positive and negative genic selection as predicted by diffusion theory does not hold in the Moran model. Nevertheless, substituting the dummy variable l = Nj and summing over the transient states gives Rearranging Eq. and renumbering (i valid for 3 ≤ i ≤ N -1. Using Eq. , we rewrite Eq. as t1B,iA = t1B,B Eqs. -, together with t1B,1B = /2, t1B,1A = / [2] , t1B,2B = /4, and t1B,2A = / [], are sufficient to compute the mean absorption time from state 1B When the initial state is 1A, we equate the elements of the second row on both sides of T = I . Then, we obtain on repeating the above analysis, t1A,iB = t1A,B --- Ni -t1A,A where t1A,1B = /2 and t1A,1A = /2. --- Appendix F Positive transition probabilities, piα,jβ , for the one-to-many transmission model, incorporating recurrent innovation and imperfect copying . piB,B = i/N, From the terminal states, 0B and NA, we have --- Largest non-unit eigenvalue of the one-to-many transmission model In order to find the rate of approach to either fixation or the stationary distribution , we examined the numerically obtained largest non-unit eigenvalue of the transition matrix specified by Eqs. -. For each of a large number of values of N, we obtained this eigenvalue as a function of u and β using Mathematica. The inductively determined eigenvalue has the form where In particular, when u = 0 and β = 1 this eigenvalue reduces to From Eq. , we may estimate the corresponding effective population size, N e , as follows. Clearly, after N birth-death events the cultural variation declines to  λ * 0  N of its previous value . Then, because N birth-death events in this Morantype model are analogous to one generation in the Wright-Fisher model, we can set 1 - . Note that we cannot use the standard formula given by Felsenstein , since 1 -λ * 0 is of order of magnitude 1/N. Hence, as N becomes large we see that for this one-to-many transmission model the dependence on N disappears and which we can interpret as extreme cultural inbreeding-the cultural variation is dictated by the equivalent of only 2.26 individuals, a small number indeed.

Cultural variation in a population is affected by the rate of occurrence of cultural innovations, whether such innovations are preferred or eschewed, how they are transmitted between individuals in the population, and the size of the population. An innovation, such as a modification in an attribute of a handaxe, may be lost or may become a property of all handaxes, which we call ''fixation of the innovation.'' Alternatively, several innovations may attain appreciable frequencies, in which case properties of the frequency distribution-for example, of handaxe measurements-is important. Here we apply the Moran model from the stochastic theory of population genetics to study the evolution of cultural innovations. We obtain the probability that an initially rare innovation becomes fixed, and the expected time this takes. When variation in cultural traits is due to recurrent innovation, copy error, and sampling from generation to generation, we describe properties of this variation, such as the level of heterogeneity expected in the population. For all of these, we determine the effect of the mode of social transmission: conformist, where there is a tendency for each naïve newborn to copy the most popular variant; pro-novelty bias, where the newborn prefers a specific variant if it exists among those it samples; one-to-many transmission, where the variant one individual carries is copied by all newborns while that individual remains alive. We compare our findings with those predicted by prevailing theories for rates of cultural change and the distribution of cultural variation.

Introduction 1.Ecosystem Services Ecosystems provide benefits to humans, and these services are known as ecosystem services [1]. They are, for example, forest purification of surface water [2], carbon sequestration [3,4] and psycho-physical well-being from frequenting green spaces [5]. These services underpin the achievement of some of the Sustainable Development Goals defined in the U.N. 2030 Agenda [6]. There are several classifications of such services in the literature, the first of which is provided below [1]. Ecosystem services can be divided into: "supporting services" that have an indirect effect on society but serve for the creation of all other categories of services; these are, for example, nutrient cycling and soil formation [7]; "provisioning services" that provide a tangible ecosystem benefit to society, such as drinking water and firewood [8]; "regulating services" that results from the proper management of ecological processes within an ecosystem, such as the reduction in natural disasters, the purification of surface water and the mitigation of ongoing climate change [9]; "cultural services" that provide intangible benefits, such as spiritual value, the aesthetic beauty of a landscape and recreational activities [10,11]. Ecosystem services now find an increasing interest in civil society [12] and also in the digital domain, as evidenced by the average search frequency of the term "ecosystem services" in the last year globally on the Google search engine, which is about 64% [13]. However, only a little research has addressed the public perception of such services, focusing instead mainly on technical and economic aspects [14]. A better understanding of such perception would help in forest planning processes, green space design and natural capital protection and enhancement [15]. --- Social Media Data Social media can be useful for extracting information and insights into subjective perceptions of a particular topic [16] and finding various uses in the academic context [17,18]. According to the Digital 2022 Global Overview Report [19], there are over 4.6 billion users on social media, three times as many as a decade ago and 10% more than in 2021. This figure is growing exponentially every year, at a faster growth rate than with the Internet, and translates into a huge amount of data at our disposal [20]. Among the different social media present nowadays, Twitter has been highly successful in academic research, as it is the main social platform where scientists and researchers disseminate their work and raise awareness [21,22], although it ranks 14th in the ranking of the most active social media platforms in the world [23]. In 2017, 1-5% of Twitter users were still active scientists [24,25], a number that is still growing today [26]. Another important aspect is that on Twitter, unlike other social media, fake news and disbeliefs find it difficult to spread [27]. --- Literature Review Several studies in the literature have analysed ecosystem services using social media [28][29][30]. Most authors, however, have focused exclusively on cultural ecosystem services by mapping and quantifying them using images [31][32][33][34], and few of them have analysed their social perceptions [15,35,36]. With reference to Twitter, several authors have analysed the social perception of natural capital, that is, nature and its ecosystem services and protected areas [37][38][39] or people's emotions [40], but few have focused exclusively on ecosystem services and specifically cultural services [41][42][43][44]. The only study that analysed the perception of ecosystem services in a broad sense focused on a specific geographical area, namely the Laurentian Great Lakes [45]. --- Research Questions and Hypothesis In light of the findings of the literature review, this study aims to investigate and fill research gaps regarding the global-scale social perception of ecosystem services on social media. Specifically, the research was conducted on Twitter with the aim of answering the following questions: Based on the empirical evidence found in the literature [38,39], the topic under research and our theories, the respective hypotheses for the research questions were identified: --- • In the social network, there are good interactions between users and high content sharing. --- • Most influential users in the network are scientists or researchers. --- • The most discussed topics are related to the latest research in the literature on ecosystem services, while the keywords are related to both technical terms used in research and more generic terms known to society. --- Sections The manuscript presents a first section devoted to a brief description of the subject matter and the context of the research. Section 2 describes the theoretical foundations on which the social and semantic network analyses are based, and the methodological flow adopted. Sections 3 and 4 report and comment on the results obtained. The last section concludes with the limitations of this study and its possible developments. --- Materials and Methods --- Data Collection This study gathered 4398 tweets that were sent between 17 January 2022 and 3 February 2022, using the keyword of "ecosystem services" and the hashtag "ecosystemservices". The Microsoft Excel plugin, NodeXL [46] was used to retrieve data. The data collection was conducted in two phases. The first phase was to retrieve data on 25 January, building the first dataset that had a time range of 17 to 25 January and a second dataset was captured on 3 February 2022, that had a time range of 26 January to 3 February NodeXL has access to Twitter's Search Application Programming Interface , which provides official access to data. The Search API provides data going back in time around seven days. For low-tweet volume topics, as in our case, the Search API was able to retrieve all the available tweets sent during our time period of data collection. --- Data Analysis The retrieved data was analysed in NodeXL by drawing upon social network and semantic network analysis , and the social network analysis visuals were produced in the Gephi software application. Social network analysis is the examination of social structures using networks and graph theory. It characterises networked structures in terms of nodes or entities that are connected by ties or relationships between the entities. In sociology, it is often used to study social groups and has also been utilised in management consultancy to study relationships among employees. In the context of Twitter, SNA can shed light on how users connect with one another. More specifically, by using SNA and analysing how users interact with each other it is possible to build an understanding of the most popular users and groups of users. SNA methods also help to visualize the users and interaction patterns in social networks. In this study, we provide a visual representation of the interaction of users and are able to show the most popular groups and the most influential users within the network . Newcomers to SNA may want to consult an overview and classification of social media networks published by Himelboim et al. [47]. The network graph was produced by analysing relationships between Twitter accounts. Accounts that conversed with each other were assigned into distinct groups within the network. Influential users were identified by using a centrality algorithm called betweenness centrality, which ranks users according to the number of shortest paths that pass through them [48]. Figure 1 provides a visual summary of this study's methodology. The manuscript presents a first section devoted to a brief description of the subject matter and the context of the research. Section 2 describes the theoretical foundations on which the social and semantic network analyses are based, and the methodological flow adopted. Sections 3 and 4 report and comment on the results obtained. The last section concludes with the limitations of this study and its possible developments. --- Materials and Methods --- Data Collection This study gathered 4398 tweets that were sent between 17 January 2022 and 3 February 2022, using the keyword of "ecosystem services" and the hashtag "ecosystemservices". The Microsoft Excel plugin, NodeXL [46] was used to retrieve data. The data collection was conducted in two phases. The first phase was to retrieve data on 25 January, building the first dataset that had a time range of 17 to 25 January and a second dataset was captured on 3 February 2022, that had a time range of 26 January to 3 February No-deXL has access to Twitter's Search Application Programming Interface , which provides official access to data. The Search API provides data going back in time around seven days. For low-tweet volume topics, as in our case, the Search API was able to retrieve all the available tweets sent during our time period of data collection. --- Data Analysis The retrieved data was analysed in NodeXL by drawing upon social network and semantic network analysis , and the social network analysis visuals were produced in the Gephi software application. Social network analysis is the examination of social structures using networks and graph theory. It characterises networked structures in terms of nodes or entities that are connected by ties or relationships between the entities. In sociology, it is often used to study social groups and has also been utilised in management consultancy to study relationships among employees. In the context of Twitter, SNA can shed light on how users connect with one another. More specifically, by using SNA and analysing how users interact with each other it is possible to build an understanding of the most popular users and groups of users. SNA methods also help to visualize the users and interaction patterns in social networks. In this study, we provide a visual representation of the interaction of users and are able to show the most popular groups and the most influential users within the network . Newcomers to SNA may want to consult an overview and classification of social media networks published by Himelboim et al. [47]. The network graph was produced by analysing relationships between Twitter accounts. Accounts that conversed with each other were assigned into distinct groups within the network. Influential users were identified by using a centrality algorithm called betweenness centrality, which ranks users according to the number of shortest paths that pass through them [48]. Figure 1 provides a visual summary of this study's methodology. --- Results --- Search Results Figure 2 shows the different types of tweets published in the two time periods analysed. An initial comparison of the two datasets reveals an almost similar situation between the different types of tweets published. The first dataset collected 2317 different types of tweets in total, while the second 2081 over a data collection time of 7 days backwards in time. This first output allows us to make a consideration, namely a low interest in the term ecosystem services on Twitter during the analysed period, since out of 18,000 types of tweets, which was the maximum threshold set in the analysis, only 4398 were collected. The ratio between tweets and retweets, shifted more towards the latter, suggests instead an excessive noise of information circulating in the network, as the number of shared contents is higher than new contents. --- Search Results Figure 2 shows the different types of tweets published in the two time periods analysed. An initial comparison of the two datasets reveals an almost similar situation between the different types of tweets published. The first dataset collected 2317 different types of tweets in total, while the second 2081 over a data collection time of 7 days backwards in time. This first output allows us to make a consideration, namely a low interest in the term ecosystem services on Twitter during the analysed period, since out of 18,000 types of tweets, which was the maximum threshold set in the analysis, only 4398 were collected. The ratio between tweets and retweets, shifted more towards the latter, suggests instead an excessive noise of information circulating in the network, as the number of shared contents is higher than new contents. --- Social Network Analysis --- Network Overviews Table 1 shows the macroscopic characteristics of the social network. Relationships are the connection between two Twitter users. We speak of relationships with duplicates in the case where there are multiple connections between two users. In both analysed datasets, the number of relationships with duplicates is very low, which probably suggests that users merely share content but do not engage in discourse or debate about it, and this is also confirmed by the reciprocated Twitter users pair ratio, as only 2 out of 100 users are mutually connected. The number of isolated Twitter users, namely those who have no relationship with other users, is low for both datasets. The diameter shows the size of the social network is, in our case, both diameters are high, so the speed of information dissemination is slow within the network. However, the low average shortest path suggests the opposite, namely that there are influential users in the network, referred to as "gatekeepers", which speed up the information dissemination process. Finally, the density and modularity metrics, both of which have a value range from 0 to 1, explain the structure of the network. Specifically, the density of both networks is very low, suggesting that most users have no relationship with others. Modularity, on the other hand, is very high, thus, there are well-structured user groups within the networks that are strongly intra-connected but weakly interconnected with other groups. In summary, information circulates slowly in the entire network but quickly in the groups due to the presence of gatekeepers. --- Social Network Analysis --- Network Overviews Table 1 shows the macroscopic characteristics of the social network. Relationships are the connection between two Twitter users. We speak of relationships with duplicates in the case where there are multiple connections between two users. In both analysed datasets, the number of relationships with duplicates is very low, which probably suggests that users merely share content but do not engage in discourse or debate about it, and this is also confirmed by the reciprocated Twitter users pair ratio, as only 2 out of 100 users are mutually connected. The number of isolated Twitter users, namely those who have no relationship with other users, is low for both datasets. The diameter shows the size of the social network is, in our case, both diameters are high, so the speed of information dissemination is slow within the network. However, the low average shortest path suggests the opposite, namely that there are influential users in the network, referred to as "gatekeepers", which speed up the information dissemination process. Finally, the density and modularity metrics, both of which have a value range from 0 to 1, explain the structure of the network. Specifically, the density of both networks is very low, suggesting that most users have no relationship with others. Modularity, on the other hand, is very high, thus, there are well-structured user groups within the networks that are strongly intra-connected but weakly interconnected with other groups. In summary, information circulates slowly in the entire network but quickly in the groups due to the presence of gatekeepers. 2 and3 show the top ten most influential users across the two datasets. The first aspect that emerges is that the users between the two networks are different, as the analyses are made about a week apart. The stakeholders involved are multiple, such as civil society, scientific journals, non-profit organisations, government initiatives and, to a greater extent, academia with professors, scientists and scholars related to environmental issues. Most influential users belong to a few groups that are recurrent in social networks, which probably indicates that these groups have the largest size. The important metric for a user's influence is betweenness centrality, since the higher it is the more influence the user has in allowing or blocking the exchange of information between one person/user and another in the network. The number of followers, instead, is not an important metric for determining the influence of a user. High follower numbers may be there because users pay for fake followers or bots. It was intended to include this metric to show how there is no causal link between a user's influence and his or her number of followers. In Table 2, the most influential users are part of the academy, suggesting a research interest in the topic of ecosystem services. In Table 3, the most influential users are a global network related to ecosystem services and a trust fund dedicated to environmental threats, most likely because of the anniversary of World Wetlands Day, promoted annually on 2 February by the United Nations. --- Content Analysis --- Top Contents Hashtags in Twitter serve to categorise topics and allow users to follow them more easily. Table 4 shows the top ten hashtags of the two datasets. What emerges, excluding the search term, is that some popular keywords today are common to both social networks, such as "ecosystem", "biodiversity" and "climatechange". There is also the presence of specific hashtags linked to events or anniversaries, such as "india" and "tiger", in reference to the news about the increase in tiger reserves in India, which as protected habitats indirectly protect ecosystem services, or "worldwetlandsday", "wetlands", "actforwetlands" and "worldwetlandday" linked to raising awareness about the importance of wetlands and the services they provide to society and the environment. The hashtag "groundedinsoil" is used by the Canadian Society of Soil Science to inform about the ecosystem services provided by soil, and probably the keyword "soil" was matched, as it has the same occurrence. --- Semantic Analysis Table 5 shows the total number of word-pairs and their frequency classes for the two time periods analysed. In the first dataset, for example, 4053 word-pairs appeared from 0 to 2 times, 1711 from 3 to 10 and so on. The more frequent a word-pair is, the more important it is. The reverse trend is clear, namely, as the frequency increases, the word pairs decrease and vice versa. No major differences emerge between the two datasets in terms of the number of word-pairs per frequency. The most frequent word-pair obviously refers to the search term "ecosystem services". For the semantic network analyses, only word-pairs with a frequency of at least ten times were considered to avoid too much noise in subsequent processing. Table 6 shows the most frequent word pairs in the analysed tweets. Due to space constraints, only the top ten results are shown for both datasets. Knowing the word-pairs helps to identify the topics discussed within social networks [49]. In general, the results that emerge are in line with the type of influential users within the two networks, since in the first case, we talk about human-nature interactions and social-ecological systems, which are much-studied topics in academia today, and in the second case, on the other hand, aspects related to climate change, biodiversity and policies emerge that are often discussed or addressed by non-governmental and non-profit organisations, activists and trust funds with the aim of raising awareness of the civil society with respect to these issues. Several word pairs then confirm what has already been identified with hashtags, namely the event in India on the increase in tiger reserves and World Wetlands Day 2022. The users' most influential appear towards the centre of the network, and their size highlights their influential nature. The network visual highlights how discussions occur in different groups and clusters. The different colours represent the different groups conversing in the network, with a few very influential users driving some larger discussions The most dominant cluster is coloured in purple and makes up 7.29% of the network. This is then followed by clusters in green and blue . The network can be characterised as a community of users who connect over a shared interest. Figure 4 below is a network visualisation of the dataset retrieved on 3 February 2022 The users' most influential appear towards the centre of the network, and their size highlights their influential nature. The network visual highlights how discussions occur in different groups and clusters. The different colours represent the different groups conversing in the network, with a few very influential users driving some larger discussions. The most dominant cluster is coloured in purple and makes up 7.29% of the network. This is then followed by clusters in green and blue . The network can be characterised as a community of users who connect over a shared interest. Figure 4 below is a network visualisation of the dataset retrieved on 3 February 2022. The network retrieved on 3 February 2022 appeared similar in structure to the network in January. There appear to be several key groups and only a handful of users who are particularly influential in shaping the discussion. The most dominant cluster was coloured purple, which made up 7.43% of the network, followed by green and blue . It is worth mentioning the dark grey cluster of the network because it contained the user with the highest betweenness centrality score, which is a measure of network influence. The users' most influential appear towards the centre of the network, and their siz highlights their influential nature. The network visual highlights how discussions occu in different groups and clusters. The different colours represent the different groups co versing in the network, with a few very influential users driving some larger discussion The most dominant cluster is coloured in purple and makes up 7.29% of the network. Th is then followed by clusters in green and blue . The network can be chara terised as a community of users who connect over a shared interest. Figure 4 below is a network visualisation of the dataset retrieved on 3 February 202 The network retrieved on 3 February 2022 appeared similar in structure to the ne work in January. There appear to be several key groups and only a handful of users wh are particularly influential in shaping the discussion. The most dominant cluster was co oured purple, which made up 7.43% of the network, followed by green and blu In both network visualisations, there are multiple smaller communities and conversations taking place with a few larger discussions. There are several medium-sized communities but no large or very large clusters of users. This finding, alongside the volume of tweets, indicates that it is a niche topic as popular topics tend to be much larger in volume and contain larger size communities conversing with one another. These results can be interpreted alongside Tables 2 and3, which provide insight into the specific users who were influential. --- Semantic Networks Different topics are discussed within the two networks analysed. In some cases, it is also possible to read pieces of sentences, pairing the different words that emerged from the semantic analysis. In Figures 5 and6, several topics related to specific aspects of ecosystem services emerge, each discussed in a different group. Some of these main topics are then recurrent in both social networks, such as sustainable food production, protection service against natural hazards and disasters , and regulation service related to carbon sequestration. Other topics, however, are unique to the networks, such as social-ecological interactions, support services and natural capital and the event on the expansion of tiger reserves in India in Figure 5 ; the ecosystem services value chain and its markets, cultural services related to tourism and Wetlands Day in Figure 6 . are then recurrent in both social networks, such as sustainable food production, protection service against natural hazards and disasters , and regulation service related to carbon sequestration. Other topics, however, are unique to the networks, such as social-ecological interactions, support services and natural capital and the event on the expansion of tiger reserves in India in Figure 5 --- Discussion Social media, as argued by Wang et al. [50], are a data source for analysing the behavioural patterns of individuals and for a better understanding of public opinion about specific issues. In this regard, several studies have made use of social media for applica- --- Discussion Social media, as argued by Wang et al. [50], are a data source for analysing the behavioural patterns of individuals and for a better understanding of public opinion about specific issues. In this regard, several studies have made use of social media for applications on ecosystem services. Wang et al. [51] have performed research on the prioritisation of ecosystem services for cost-efficient governance. Xuezhu and Lange [14] explored the public perception of ecosystem services for nature-based solution projects. Hausmann et al. [52] analysed 'tourists' preferences in protected areas for nature-based experiences. Other authors have focused on the assessment and mapping of such services [41,53,54]. However, there are no studies in the literature that have been carried out exclusively to understand the social network and public opinion on social media regarding the term ecosystem services. Yet, it is an increasingly popular topic used to indicate the public role of natural resources when moving from multifunctionality to the perceived usefulness of resources for society. This is the context of our research, and the choice of using Twitter as a case study is due to multiple reasons, as reported by several authors [55][56][57][58]. These include the ability to have open communication between users and therefore favour the sharing of opinions, compared to other private social networks, the increasing use of Twitter for research purposes and public opinion studies. Based on these considerations, our work tried to shed light on the following questions: 1. The first one concerned the type and characteristics of the network within the term ecosystem services. Based on the density and modularity metrics and the number of isolated Twitter users, it can be concluded that the structure of our network is "community clusters", as defined by Smith et al. [59]. This result partially satisfies our hypothesis since the "community cluster" structure, as evidenced by the low density and number of isolated users and high modularity, restricts the flow of information to small groups of users who, in turn, rely on one or a few influential users in that cluster and for that specific topic related to ecosystem services. Himelboim et al. [47], however, identify advantages of this structure such as the stability of the small groups over time and the possibility of identifying different opinions in the network regarding the analysed topic since there is no single source of information and multiple conversations can be established each with its own community and points of view. Another demonstration is the low relationship and reciprocal relationships in the network. Even though users made efforts to create relationships with other users, as evidenced by the high retweeting, the interactions were unsuccessful. Ahn and Park [60] state that the function of retweeting news is not only to facilitate the exchange of information but also to create relationships with other users and express their agreement in public. However, according to Wiiava and Handoko [61], a conversation is active when community users interact with each other. This network structure, however, is in line with the study conducted by Pan and Vira [39] on natural capital, a term akin to ours, and differs from another environmental issue, namely climate change, in the study conducted by Williams et al. [62] on climate change, perhaps because the latter issue creates a strong polarisation and alignment in this regard. Although, a more recent study by Bridge [63] on climate action showed that the debate was not so polarised. The type of network probably changes depending on the time period, the topic analysed, and the importance and role of the users participating in the debate. 2. The second research question was about the type of users who play a key role in the information flow of the network. Here again, our hypothesis was partially confirmed, as in the social network obtained from the first dataset, the majority of the ten most influential users belonged to the academic world and published posts related to the promotion and approval of their research work, thus supporting what Côté e Darling [21] defined that Twitter is the main social platform for the dissemination of results from research. The second network, on the other hand, had a majority of governmental bodies, nonprofit organisations and activists, probably because in that analysed time period, there was World Wetlands Day, and several institutions were therefore intent on raising awareness on the topic. Aguilar-Gallegos et al. [64], in a study analysing the social networks on Twitter regarding an agricultural research and training centre in Mexico, point out that central institutions should not only create content but also identify and promote relevant information in the networks, act as intermediaries and encourage the dissemination of content. This concept is shared by several authors [65][66][67]. 3. The last research question investigated the most discussed topics on the topic of ecosystem services. In order to do this, the analysis of the ten most used hashtags in tweets was useful, as it made it possible to identify topics of general interest, as stated by Fedushko [68] and remove disambiguation between tweets that do not have a commonality and obtain information about them, as reported by Palos-Sanchez et al. [69]. The main hashtags were the forerunners of semantic network analysis, and topics emerged that were both topical in research, such as social-ecological systems, nature-based solutions and markets for ecosystem services, and topics related to events or anniversaries. This diversity of topics affirms what was said before, namely that in "community clusters" networks, there are different sources of information, each with its own audience and conversations. --- Conclusions This study analysed the most discussed topics around ecosystem services on a social media platform, specifically Twitter. To our knowledge, this is the first study to use SNA and semantic analysis to analyse the social network and public opinion of the term ecosystem services on Twitter. Furthermore, this is one of the first studies using this methodology applied to ecosystem services in the post-pandemic period. What emerges is that the social network is characterised by having small, well-focused discussion groups on precise aspects related to ecosystem services, but that the flow of information throughout the network is weak. There are several influential users within the groups, and these vary depending on the time analysed and the topics discussed. The topics are also diverse and cover ecological, social, economic, and political aspects related to ecosystem services. This study aimed to investigate and fill research gaps regarding social perceptions of ecosystem services on a global scale on social media, with the objectives of: characterising the social network on Twitter in relation to ecosystem services; identifying users who play a key role in the dissemination/dissemination of information on the network; and highlighting the most discussed topics and keywords that emerge in discussions. The user analysis showed that there are many stakeholders involved, such as civil society, scientific journals, non-profit organisations, government initiatives and, to a greater extent, academia with professors, scientists and scholars related to environmental issues. Within this network, the most influential users are part of academia, suggesting the growing research interest in the topic of ecosystem services. Content analysis summarises today's most popular keywords in discussions, such as "ecosystem", "biodiversity" and "climate change". Semantic analysis confirms these results, and the most frequent word pair obviously refers to the search term "ecosystem services". --- Limitations There are five potential limitations to the study. The first concerns the use of social media for research purposes, as such analyses are often difficult to replicate due to data access constraints and the need to rely on private companies specialising in data storage. The second relates to the choice of using only Twitter and its inherent biases, as each social media outlet has specific types of users and content posted, and this can influence the research. In any case, for our public opinion research, Twitter was the most suitable social platform, compared to others such as Instagram, which is mainly used to share recreational activities and experiences, Flickr to publish professional photos or Facebook, which with private and semi-private profiles makes it difficult to have open communication. The third limitation concerns the period considered, as our research only focused on two weeks, but it might be interesting to analyse more distant and relevant periods. The fourth limitation is due to the difficulty, in recent years, of analysing public opinion on ecosystem services in normal times, i.e., without the presence of special events or emergencies, such as political issues and socio-economic aspects, which can influence opinion and spark extraordinary debates and discussions. For this reason, in this work, an attempt was made to analyse two different time periods, to try to extract both the spot event component and the common component typical of normal times . The last limitation is due to the choice of the English language only, but we only wanted to focus on the international level, and taking into consideration other languages would have drastically increased the number of tweets to be analysed by the software, with its consequent technical limitation of data processing. In fact, beyond the threshold of 18,000 tweets, the software is not able to collect any more data. --- Implications for Research This study can be used as a bibliographic source for future studies on public opinion of ecosystem services, as it is the first of its kind after the pandemic period. At the same time, it can be cited for the methodology applied, that is, SNA and semantic analysis on innovative tools such as social media. The results obtained can then be used in a combined approach with more traditional public opinion survey methods, such as interviews or questionnaires. Furthermore, our study builds upon previous literature using social network tools to analyse Twitter data [70][71][72]. Researchers can use the results of this study to conduct new analyses on ecosystem services by extending the research time period and refining the methodology employed. --- Implications for Practitioners The results of this study can be used by public institutions and social organisations to improve their initiatives, programmes and policies related to nature and its services. Social media, such as Twitter, can then be a good channel for communicating environmental awareness actions, activities and campaigns by policymakers and activists, such as World Wetlands Day. In addition, being aware of the topics discussed in public opinion can be useful for companies in promoting ad hoc marketing campaigns, segmented by type of user, even more so if the social network is presented, as in our case, to many small groups, each dealing with a different aspect of ecosystem services. Finally, our results highlight the importance of the role of research and professional training in guiding public decision making on these issues and of complete and correct information to civil society, which is often easily influenced by the media on alarmist topics about the environment and its services as a subject [73,74]. --- The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph192215012/s1, Table ---

Social media data reveal patterns of knowledge, attitudes, and behaviours of users on a range of topics. This study analysed 4398 tweets gathered between 17 January 2022 and 3 February 2022 related to ecosystem services, using the keyword and hashtag "ecosystem services". The Microsoft Excel plugin, NodeXL was used for social and semantic network analysis. The results reveal a loosely dense network in which information is conveyed slowly, with homogeneous, mediumsized subgroups typical of the community cluster structure. Citizens, NGOs, and governmental administrations emerged as the main gatekeepers of information in the network. Various semantic themes emerged such as the protection of natural capital for the sustainable production of ecosystem services; nature-based solutions to protect human structures and wellbeing against natural hazards; socio-ecological systems as the interaction between human beings and the environment; focus on specific services such as the storage of atmospheric CO 2 and the provision of food. In conclusion, the perception of social users of the role of ecosystem services can help policymakers and forest managers to outline and implement efficient forest management strategies and plans.

Background Physical inactivity is a public health concern that is considered a potential risk factor for adverse health outcomes worldwide. The World Health Organization reports the prevalence of PI to be more than 80% in adults and 23% in adolescents. This rate varies considerably across countries. Reports indicate that the prevalence of PI is higher in the eastern Mediterranean region, the US, Europe and the Western Pacific region than in other parts of the world [1]. Urban and industrial life, advances in technology, economic development and globalization have led to rapid changes in lifestyle and PI in individuals, leading to an increase in the prevalence of related chronic diseases such as coronary heart disease, colon cancer, hypertension, stroke, breast cancer, type 2 diabetes and osteoporosis [2]. Childhood is a duration of fast physical and cognitive development [3]. Studies show that many behavioral risk factors for chronic diseases, including PI, are rooted in childhood and that the patterns of behavior developed during this period remain relatively stable in adolescence and adulthood [4,5]. PI leads to obesity, reducing working memory and brain health in children [6]. Therefore, it seems that children are an important target group for interventional and preventive activities. Any intervention in this group can be very effective in controlling the chronic diseases epidemic in adulthood [3]. Today's adolescents are less active than their peers decades ago. Watching television and computer games are known to be the main causes of PI in adolescents. There are also other factors in Iran, such as living in small houses and apartments, the phenomenon of single adolescent or the low number of adolescent in each family, lack of appropriate sports facilities in the community and schools, pressure of lesson plans for high school students, preference of education than other aspects of life, disregard for sports at school and home and lack of security in the community that affect inactivity and lower activity of children and adolescents [3,7]. One of the most important health indicators that influence one's attitude, behavior, and exposure to risk factors is socioeconomic status . There is evidence to suggest that the lower the SES of a person, the worse their health status. There is, in fact, a social gradient in health that moves from the top to the bottom of the socioeconomic spectrum and creates gaps in health outcomes such as infection rates, mortality and disability across different social classes worldwide [8]. In the 2014 study by Finger et al., the physical activity and SES of parents of German children aged 11-17 years were examined. The results showed that higher education of parents was negatively correlated with the level of PI of children [9]. Rey-López's study [10] had similar results; its results showed an inverse relationship between socioeconomic factors such as parental education and child PI. A systematic review was also conducted to investigate socioeconomic factors affecting physical activity by Ragna Stalsberg et al. The results indicated a positive relationship between socio-economic factors and physical activity, especially the impact on leisure time [11]. To our knowledge, there is little information available on the status of physical activity in adolescents in Iran. So far, very few studies have been conducted to investigate the effect of socio-economic inequalities on PI in Iranian adolescents. Therefore, the aim of the present study was to determine the prevalence of PI in Iranian adolescents aged 10-12 years and the effect of socioeconomic factors and economic inequalities on it. --- Methods --- Study setting and participants The population of this cross-sectional study was 10-12 years old children in fourth, fifth and sixth grades of elementary school. The final sample size was 1590 individuals. The multi-stage sampling method involving systematic classification, clustering and random methods were used. For the first step, the provinces of Iran were divided into three geographical clusters of 9, 9 and 13 provinces. One province was randomly selected from each cluster including Kurdistan province in the west, Markazi in the center and Fars in the east. In the next step, one city was randomly selected in the selected provinces including Marivan in Kurdistan, Saveh in Markazi and Garash in Fars. The number of children in the age range were 13,513 , 9864 and 2561 . In order to reduce the error, the sample size of each cluster was selected proportional to the size of that cluster. Six elementary schools were randomly selected from the primary schools in each city and 526 samples were randomly selected from each of the 10 to 12-year-old students . Prior to the study, written consent was obtained from all parents and trained public health experts were used to collect data. Questionnaires used in this study included demographic information questionnaire, modifiable activity questionnaire [12] and socioeconomic status questionnaire which were completed by adolescents and parents at home based on the protocol. In the MAQ questionnaire, each physical activity was weighted according to its metabolic equivalent . One MET was the amount of energy consumed by a resting person every minute. Then, the level of physical activity was calculated in terms of hours of activity per week and the subjects were divided into three groups based on their overall physical activity: light activity , moderate activity and heavy activity . Based on the studies, the reliability of this questionnaire is 0.97 and its validity is 0.47 [12]. To determine SES, questions were asked about parental assets. The SES of each province was measured based on an assets index computed for each province using the data on assets ownership of households method, an asset index was calculated and SES has been determined by the assets index of the households computed based on the asset index. --- Statistical analysis To determine the SES, we used the method proposed by O'Donnell and Moradi, et al. [13,14]. Accordingly, a socioeconomic status questionnaire consisting of a number of assets was used. PCA first identified the variables that had the most impact on the variance of all variables and then a new SES variable was constructed based on these variables. The PCA gives each asset a specific coefficient or weight then extracts a linear combination of the variables with the most variance. It then eliminates this variance and finds the second linear combination that describes the largest proportion of the residual variance, and continues this procedure. In this study, we developed an asset index using the PCA method. Five quintiles were constructed using the mean of this variable which divided the statistical population into 5 groups: very poor, poor, middle, rich and very rich. Finally, the PI variable was compared between the two very poor and very rich groups. Logistic regression was used to evaluate the relationship between prevalence of outcome variables in different socioeconomic subgroups. The highest socioeconomic group was considered as the baseline. The model began with the baseline model and then added just significant determinants and explained the effect of each determinants contributing to inequality. Concentration index and concentration curve methods were used to measure inequality. The CI quantitatively shows the degree of inequality at the income distribution level of a health variable. To calculate the relative CI, individuals are first sorted by socioeconomic status, and then the cumulative percentage of the population is plotted against a cumulative percentage of the health variable to create the CC. CI values vary from + 1 to -1 [14]. It is one of the most common indices in calculating income inequality and socioeconomic status. Negative values indicate that the PI variable is concentrated among people with poor socioeconomic status and the CC is above the equality line. But positive values indicate that the PI variable is concentrated among the rich, and when the distribution of health among all individuals is the same, the CI will be zero. The Kakwani method was used to calculate the CI based on the following formula: C ¼ 2 μ COV y i; R ið1Þ In this formula, C is the concentration index, Cov is covariance, yi is health variable, Ri is economic rank, and μ is mean health variable. Due to the limitations of this index for binary data, CI was normalized based on the method used by Wagstaff according to the following formula in which C is the standard concentration index and μ is mean health variable [15]. Wc ¼ C= 1 -μ ð Þð2Þ After measuring the inequalities, the decomposition concentration index was used to determine the contribution of each of the determinants to inequality. According to the Wagstaff method, we considered a linear regression model in relation with PI variable and a set of determinants : y i ¼ α þ X k β k x ki þ ε ið3Þ i is the average i th person, β k is the regression coefficient, ε i is error coefficient or interpersonal changes. Based on the relationship between y i and x k in formula 3, we write C for PI as follows: C ¼ X β k x k μ C k þ GC ε μð4Þ In this formula, inequality generally consists of two explained and unexplained parts. In the explained part, μ is the mean health variable, β k is the regression coefficients for PI on the available determinants, x k is the mean of the determinants or x k , C k is the concentration index C for the x k determinant and in the unexplained part, GC ε μ is the residual error in the total C for ε i [14]. In this study, PI was considered as the response variable and Chi-square test was used to estimate the prevalence of response variable at each level of demographic variables. Multivariate logistic regression was used to estimate the final model based on variables with p < 0.1 in chi-square test and calculation of univariate OR and AOR. All analyses were performed in Stata/SE 14.0. --- Results Of the 1590 participants, 410 were physically inactive, 116 had moderate physical activity and 1064 had heavy physical activity during the week. Among the participants, 696 cycling, 633 football or handball, 613 went running, 485 volleyball, 347 swimming, 226 hiking or outings, 212 skating, 199 did martial arts, 133 ping pong or badminton, 124 dance or rhythmic movements, 116 gymnastics and endurance sports, 100 basketball, 78 wrestling, bodybuilding or weightlifting, 8 skiing, and 90 people did other physical activities. The mean of total physical activity in terms of was 40.72 . The mean of physical activity among girls was 30.87 and among boys was 49.61 MET hours per week. The mean of physical activity among fourthgraders was 40.1 , fifth graders 38.7 and sixth graders 43.5 . Table 1 shows physical activity in terms of sex, educational grade, BMI, maternal education, father education, age of father, age of mother, highest education level of both parents, and household size. Mean physical activity during 1 week was significantly different between boys and girls . In Table 2, univariate OR is presented for those variables in Table 1 that had a significant relationship at P = 0.2 with physical activity level, using univariate and multivariate logistic regression. The OR for the variables of sex and maternal education had a significant relationship with physical activity level . In multivariate logistic regression, sex and maternal education at secondary and high school level had a significant relationship with physical activity level . Also, the concentration index for PI was negative , indicating inadequate activity in adolescents with low socioeconomic status. The Fig. 1 indicated propoor inequality for PI, CC of PI in is above the equality line, indicating a greater concentration of PI in the poorer group of the community. Table 3 shows the results of CI decomposition for PI in adolescents. The reference group of 10-year-old male students with parents under 35 years of age and high school diploma or academic education was considered 3). --- Discussion Based on the findings of this study, about 26% of adolescents were physically inactive. Female adolescents and adolescents with less educated mothers were significantly more physically inactive. Also, according to the centralized index, inequality was seen in the PI rate, so that the PI was concentrated in lower socioeconomic classes. Adverse economic conditions, paternal education, and household size were the most important factors associated with PI inequality. PI is one of the most important risk factors for mortality worldwide. It is one of the pests of industrial life that is increasing every year due to advances in science and technology [16]. The results showed that about one quarter of adolescent in the study had PI. The results of other studies conducted in Iran and other parts of the world also show a high prevalence of PI in adolescent all over the world [17][18][19]. The results of the study by Kelishadi et al. on 6-18 years children showed that 24% of them were physically inactive and the average physical activity in boys was higher than in girls, which is consistent with the present study [19]. A systematic review to assess PI in Arab countries also found that the prevalence of PI was higher in girls than in boys [20]. Differences in physical activity can be influenced by gender-defined roles or societal norms and values, which may result in women having less access to resources for exercise and have fewer choices for leisure time [21,22]. Also, differences in psychological and behavioral aspects can be effective in making these differences [23]. The findings showed that adolescents with mothers with low levels of literacy had a higher chance of PI than others. Mothers with more education appear to be more aware of the importance of regular physical activity in Fig. 1 The concentration curve for physical inactivity in adolescent in Iran adolescents [24,25]. However, the results of some studies are different and show that the level of physical activity in adolescent decreases as mothers' education and employment levels increase [26]. These differences may be justified by different cultural and family backgrounds. Providing proper awareness and education about the necessity of physical activity in adolescents is one of the critical issues that should be considered for families. One message of this study to improve physical activity is to increase mothers' education level. Another determinant of health and well-being in all periods of life is socio-economic status, which is considered an important factor in creating health inequalities in populations [27]. In this study, the factors affecting inequality were investigated using the decomposition model. The findings showed that low economic status, level of education of the father and the family size were the most important factors in creating inequality. CI analysis showed that people with low economic status had less physical activity than other adolescent. These adolescent may not have access to sports facilities and equipment due to their cost constraints or may be reluctant to do so due to lack of time and lack of access to safe and appropriate sports facilities [28]. A study by Prins et al. on Dutch adolescents showed that adolescents' lack of involvement in sport activities was highly correlated with lack of access to sports venues, parks and green spaces, and the socioeconomic status of households in a neighborhood [29]. On the other hand, some scholars believe that people in the upper classes are most aware of the goals of a regular physical activity promotion program [30,31]. However, some studies have had different results and have shown that children in the upper classes may have less physical activity for reasons such as having a more prosperous life, more consumption of fast foods, more use of vehicles, computer games, mobile phones, etc. [32,33]. In the present study, fathers' education level also played an important role in creating inequalities: children of fathers with low educational level were more inactive. This difference may be due to the low level of literacy, awareness and attitude of the head of household that who is mostly the father towards the benefits of physical activity of adolescent and not providing suitable conditions for physical activity in children according to their life preferences. It seems that when education level of parents increases, they have more information and they can better encourage their adolescents to do physical activity. Similar results were obtained in the Eduardo Gonzalo-Almorox study, which examined the effect of socio-economic inequalities on leisure-time inactivity of Spanish adolescents. The results of this study showed that the education level of household head and family income were the main socio-economic inequalities affecting child inactivity [34]. Also, in the present study, household size or the number of individuals in the family was one of the factors affecting the inactivity of adolescent. Adolescents with more than 4 family members had more PI than families with less than 4 members. In a study carried out in Kurdistan in 2014, the factors affecting inequality in PI of adolescent aged 10-12 were investigated using Oxaca model. The results of the study showed that the PI of individuals in the lower socioeconomic classes was higher than the others. In the initial model, increased maternal literacy, improved socioeconomic status, and improved living place were all effective in increasing physical activity. Environmental problems such as lack of open space, parks, or sports fields often make it difficult that some adolescent engage in sufficient physical activity. It seems that, given the contradictions in the results of various studies, education on the importance of public health and adopting healthy behaviors during childhood and adulthood due to the dangers of inadequate physical activity are essential to have a healthy community away from economic and social differences. School-based interventions, with the involvement of students' families, and multidisciplinary interventions such as active commuting from/to school, active Physical training lessons, active school break, sleep health promotion can ultimately lead to increased physical activity in children and adolescents [35,36]. In the present study, cycling and sports such as soccer or handball were more common than other physical activities, and skiing accounted for the lowest percentage of physical activity in adolescents. This difference, on the one hand, may be due to the availability of sports facilities for both sexes, the costs and seasonality of a sport, and, on the other hand, due to parents' views on its impact on adolescents future. These results indicate that boys and girls have different expectations and interests in choosing their physical activities, which is very important for policy-making and planning. One of the limitations of this study may be recall bias due to questions about subjects' activities in a past period, which were attempted to be reduced as much as possible by training and protocol preparation. Another limitation is uncertainty about the measure of association in a cross-sectional study. However, the present study with a large sample size provides valuable evidence on family and behavioral risk factors affecting PI in Iranian adolescents, which can justify further related studies, effective interventions, and planning and policy making. --- Conclusion One quarter of the study population had PI, which is alarming for this age group. The situation of the girls was worse. PI was more concentrated in adolescents with poorer socioeconomic status. Socioeconomic levels, parental literacy, and children's sex were factors affecting the level of physical activity. Improving the socioeconomic status of households, increasing the education of parents and especially mothers, and paying more attention to groups with lower socioeconomic status are recommended to improve physical activity in adolescent. Large-scale national policies are recommended to promote physical activity in adolescents. Increase in parental knowledge through local mass media or other communication strategies may be associated with more support for participation in physical activity. Domestic governments have a key role to play in promoting physical activity. Design and construction of safe places like parks, soccer fields, and walking trails are associated with more outdoor play especially in adolescents with lower economic levels that needs national policies. --- --- Supplementary information Supplementary information accompanies this paper at https://doi.org/10. 1186/s12889-020-09618-0. --- Additional file 1. Abbreviations PI: Physical inactivity; MAQ: Modifiable activity questionnaire; SES: Socioeconomic status; PCA: Principal component analysis; SD: Standard deviation Authors' contributions GM and BP conceived and designed the study. BZ, FM, BP and GM analyzed and interpreted the data, and drafted the manuscript. GM, BP, MM, BZ and MAR were involved in the composition of the study tool, supervision of the research process and critical revision and review of the manuscript. All the authors read and approved the final manuscript. --- --- --- Competing interests The authors have no conflicts of interest to declare. ---

Background: This study aims to determine the prevalence of physical inactivity in Iranian adolescents aged 10-12 years and the impact of socioeconomic inequality on it. Methods: In this descriptive study, the study population consisted of 10-12 years old adolescents from an Iranian population from Kurdistan, Fars and Markazi provinces in 2018. The sample size was 1590 individuals. The sampling method was cluster sampling. Data was collected using demographic questionnaire, modifiable activity questionnaire (MAQ) and socioeconomic questionnaire. Cut points on the MAQ for light activity, moderate activity and heavy activity were MET< 3, MET = 3-6 and MET> 6, respectively. Linear and logistic regression were used to estimate the final model and the Oaxaca analysis method was applied. All analyses were performed in Stata/SE 14.0. Results: Of the 1590 participants, 52.82% were male. The results showed that 25.79% of the subjects were physically inactive and 7.30% engaged in moderate physical activity during the week. The average physical activity during 1 week was more in boys than in girls (P-value< 0.05). Adolescents of mothers with secondary and high school education were more likely to have physical inactivity than mothers with a high school diploma or higher (AOR: 1.35, 95% CI: 1.02-1.77). The concentration index was -.11, indicating a greater concentration of physical inactivity in adolescents with low socioeconomic levels.One-fourth of the study population had physical inactivity in this age group. Socioeconomic levels, parental literacy, and sex of adolescents were associated with the level of physical activity.

INTRODUCTION Technological development is the main driver of international competitiveness and economic growth. Therefore, technological progress has become a policy priority for many countries, and considerable financial resources are invested with the expectation of long-and short-term returns. Since the 1950s interest in the development and use of new technologies has steadily grown, together with the attempts of decision makers to regulate related processes at the institutional and national levels. To describe the ongoing changes in global technological development, a selected number of terms emerged to tag similar groups of technologies . The most popular labels are emerging, disruptive, enabling and advanced technologies. These terms are also widely used in the academic literature and provide relevant classification categories or elaborate an operational definition to clarify the meaning a widely used umbrella concept. Among the four mentioned technological categories, two deserve special attention, namely the 'advanced' and 'emerging' technologies. The former is broader and older, comprising various technologies from flexible manufacturing, computer-aided design to robotics and automated storage systems. The term appeared in the literature in the 1950s and became the foundation for the future development of other types of technologies. Advanced technology is a key category that emerged together with a set of definitions and distinct criteria for R&D. Immediately following WWII, advanced technologies were primarily associated with the growing need for sufficient training of technical specialists . Early definitions of the term addressed various features of advanced technologies. Before 1980s it implied a certain complexity of production processes, including the required equipment and organizational methods, with the aim to increase efficiency . Advanced technologies were embedded in discussions of developments in energy storage , automation processes and computer utilization . Later definitions used the term primarily in controlling, tracking or monitoring related manufacturing processes . The mid 1990s were marked by the first conventional definition of advanced technology, which was proposed in the Frascati Manual in relation to manufacturing processes and the occurrence of the term in statistical and other surveys . Despite the growing interest in the role of advanced technology in economic growth, the definition vanishes from later editions of the Manual as well as from other OECD manuals. The term seemed to contradict the dominant term relating to the measurements and evaluation of technology. Still, the definition proposed in the Manual was replicated in some later works . Despite the variety of interpretations, there still is no common understanding or a notion of advanced technologies. There are however some common features that the majority of authors share: novelty, relation to information technology, and improvement in efficiency of current operation or production processes. However, in the literature one can hardly find robust criteria for distinguishing advanced technology from others. Moreover, some of the characteristics used in the definitions such as novelty or efficiency, are universal and therefore relevant for the definitions of other groups of technologies. As opposed to the vague concept of 'advanced technology', the literature on emerging technology is more coherent. Numerous papers present a definition of the concept of 'emergence' in general and 'emerging technology' in particular. The first appearance of 'emerging technology' is associated with the late 1960s. The two early studies using the term belong to different research areaspsychometry and engineering. Despite this fact, both authors perceive 'emerging' as 'new and superior' to existing practices . Later studies equate emergence with uncertainty of outcomes, unobvious social impact or limitations of availability . One of the recent studies review defining emerging technologies include five features of emerging technologies: radical novelty, fast growth, coherence, prominent impact, uncertainty and ambiguity. Since the 1990s, emerging technologies received much attention from scholars in the field of both strategic management and science and technology policy setting. Due to potential of these technologies for discontinuous and pervasive change, particular attention has been paid to their role in socio-economic development. Emerging technologies such as spin transistors, gene therapy, digital imaging, micro-machines are referred as originators of new industries , with an influence on a country's capability to introduce technological innovations and, therefore, compete at the international level . Moreover, the concept has been examined by technological foresight studies to identify the most promising technologies for the socio-economic development of different countries . Over time perceptions of certain technological categories have undergone remarkable change in the professional discourse. This paper is focusing on a meta-analysis of discussions on emerging technologies as they are represented in journal articles and asks whether social studies in advanced and emerging technology construct consistent research fields. While conventional approaches focus on the analysis of terminology usage and construction of definitions in order to provide certain conceptual framing of a growing area this study suggests looking at corps of texts that set up a research agenda for a certain period. If a semantically connected group of papers appear once in a network and demonstrate a certain level of stability over time we can conclude that there is a self-generating communication core in professional discussions and therefore a consistent research field exists. To identify such a 'communication core' and its elements it is suggested that one take a closer look at reproduction of co-citation networks through analysis of transitionary references. The rise of the same co-cited papers in a longer period can indicate ongoing institutionalization processes in academic discussions. Moreover, it is proposed to look at centrality characteristics of transitionary references to clarify if they are filling gaps in knowledge or acting as gatekeepers and consolidating accumulated theoretical or practical experience on a crossroad of different frameworks. Identification of these papers in dynamics may bring a better understanding of the evolution of weakly conceptualized and growing research areas. The overall scheme of the proposed approach is demonstrated at Figure 2. --- Fig. 2. A proposed approach to analysis of a research field consistency The paper is structured as follows. First, we address approaches used for the analysis of scientific networks and the identification of a knowledge base in a selected research field or tradition. Here we pay special attention to the idea of spatial configuration of research networks and the notion of 'betweenness centrality'. In the next section, we describe our method in greater detail. In the findings section, the results of our analysis are presented. Finally, we discuss the main findings of the paper and draw conclusions on the limitations of the study and further areas of research. --- LITERATURE REVIEW Ways to identify the emergence of new scientific and technological areas has recently received great attention from the scientific community . Various contributions offer distinctive data mining methods for investigation of the overall dynamics of technology development or growth of specific technology domains. Some scholars suggest looking at bursts of academic publications to detect the emergence of a wider range of technologies , while others focus on citation network analysis to identify research fronts in a specific technology domain . In these studies authors elaborate their best methods for identification of emerging technologies by modeling milestones [a good review on this can be found in ], while the question as to whether this area constructs a separate field of research remains untouched. The answer to such a question requires some reflection on theories concerning the social nature of knowledge networks and the evolution of research fields. Numerous of social studies of science have demonstrated that a research field exists when a group of scholars act within a common paradigm that concentrate their effort on a certain set of cognitive problems and form collective identities, research goals and practices . This approach works properly until one is considering any of the problem-oriented fields developed within or against an established standard of research work supported by a certain level of cognitive and social organization of scientific effort . In this case, specific patterns of the development of new disciplines and re-organization of scientific fields can be identified through an analysis of organizational change ) as well as through the growing interdisciplinarity of science . For both cases authors regularly adopt the strategy of exploring the 'knowledge base' of the recognized field by looking at key contributors and/or publications that are setting the research agenda for a certain period . In a fewer number of cases authors suggest looking at the emerging knowledge base that is associated with some vague concepts such as 'knowledge economy' . Specific methods of such an analysis may include the identification of keyword combinations, describing the growing research domain or an exploration of the listed number of journals primarily associated with the selected research field . The latter is crucial for the completeness and fullness of an observation as long as the use of some concepts might be a challenge when one is dealing with a weakly defined concept that may have different connotations in different disciplines. One example is the 'knowledge economy' that was mentioned above. Another example is the term 'institution', which despite its common use, has different interpretations in the tradition of classical and new institutionalism used in organizational analysis . Therefore, reliance only on methods that search for a common terminology or use of term is not enough to argue for the existence of a research field. Another strategy to use when considering the development of a scientific field lies in the analysis of communication structures. To identify stable communication networks that feed further academic discussions, scholars suggest using Social Network Analysis and particularly co-citation networks . Specifically, attention is paid to the co-citation analysis initially introduced in 1973 for studies of linkages between the documents to show their semantical relatedness. Further this method was adopted for mapping invisible colleges and clusters of science as well as searching for emerging topics . As a measure of the frequency with which at least two papers are cited together in other documents, co-citation analysis allows for the identification of cognitively related knowledge clusters accepted by a wider network of followers. Similar ideas can be found in graph theory showing that the spatial configuration of elements might characterize their role in a network. For instance, Bavelas ) showed the relationship between centrality and communication processes in small networks. He demonstrated that optimally positioned actors could accumulate information flows from dislocated parts of a network and therefore play a gatekeeping role. Smith and Leavitt suggested that authors holding central positions might also influence the behavior of other linked members. For the aim of this study it is assumed that those actors in academic networks that gain more attention from their colleagues through received citations will have a higher social status and therefore hold more central positions . Subsequently, a centrality position would mean the author's association with a communication core of a field. Recent papers suggest using betweenness centrality algorithms for placing actors and journals in citation networks to identify core elements of knowledge networks. Papers with higher betweenness are essential in a network as long as they mark parts of the network known as 'structural holes', i.e., the separation between non-redundant contacts that provide opportunities for mediating knowledge flows in a wider community of actors . On the other hand, betwenness centrality might be taken as a characteristic of gatekeeping papers . While the original theory focuses on competition and social capital in actors' relationship, we took the idea of information flows among actors and further elaborated it for the co-citation networks. In line with Burt, this paper focuses attention on the actors who adopt relatively more important positions and therefore affect co-citation characteristics. --- METHOD An earlier study by the authors on the evolutionary analysis of advanced technology as a research field showed the penetration of the concept from social sciences to natural and engineering disciplines . The analysis covered 10-year periods and identified at least three phases with foci on different issues: R&D and technology management , technology diffusion and innovation development , and discussions on specific technologies and materials . The method suggested in the paper was initially adopted from , who investigated co-authorship networks of library and information science through centrality analysis. The authors used different centrality metrics to compare the stability of positions of the top 30 authors in a network moving from one period to another. For the aim of this study, it is proposed to use the idea of transitionary papers to a co-citation network in order to identify a communication core in discussions about advanced and emerging technologies. It is assumed that such a core exists when there is an observable, significant number of references moving from one co-citation period to another. In other words, there is an observable and stable citation track that connects discussions in more than one follow-up period. This is expected to be especially relevant for discussions in social sciences as long as they have larger citation windows and longer citation half-life . For a better understanding of the role of transitionary references it is suggested to compare their centrality characteristics of which specific attention is to be paid to betwenness centrality as long as this measure is associated with the idea of gatekeeping in science . To summarize, in order to verify existence of a separate research field in discussions about advanced and emerging technologies, the following five steps were taken. 1. Data extraction and the creation of a data set. For the purpose of this survey, data was extracted from the SCOPUS database accessed: 20.11.2016 for the period from 1955 until the end of 2015 , TITLE-ABS-KEY ). This allowed for the extraction of documents using the term explicitly or specific technologies directly associated with it. Papers published in the social sciences and humanities were considered the only data source for the current study. Previous studies showed that the majority of attempts to elaborate a definition or conceptual framework for analysis of vague technology domains came from the social sciences . --- Identification of periods for comparison. Using a frequency analysis of the number of documents already reveals several shorter periods of publication activity growth. While the emergence of early papers goes back to 1960s, the overall number of studies remains insufficient for the co-citation analysis until mid-1990s for both advanced and emerging technologies . Due to the increased attention given to citation-based indicators, scholars discuss differences in citation patterns across the fields of science and various disciplines. As shown in , the ageing of publications in social sciences and humanities is slower than in natural sciences. Thus, Wang showed that for SSH journals optimal citation windows go between 5 and 15 years, shorter or longer periods do not allow for tracking publication influence. The literature agrees , that threeyear citation windows help to balance fields rapidly becoming obsolete and the fields that become obsolete more slowly . The current study attempts to cover both types of citation behavior among publications in SSH. Therefore, longer citation windows were used. Despite that, the decision was made to keep the periods the same for both groups of technologies to assure comparability. --- Construction of co-citation networks. In a third step, VOSviewer version 1.6.5 1 was used to construct cocitation networks . Due to the low density of both networks papers cited at least twice were taken for further consideration. Table 2 provides details about the networks. Betweenness centrality was further calculated in UCINET . 4. Identification of papers co-cited in more than one period. Following the recommendations provided in Yan and Ding , networks characteristics were compared to identify two types of references: those that have high centrality values, and 'transition references'papers co-cited in more than one period. Identifying such elements in a network signals the existence of continuous research questions that are one of the crucial elements of research field . --- Comparison of centrality metrics. In order to understand the role of transitionary references in co-citation networks for two or more concurrent periods it was suggested to compare their centrality characteristics of which specific attention is to be paid to betwenness centrality. The latter was selected as a key measure to look at the group of authors connected by the shortest paths passing through a vertex. Compared to other metrics , these centrality characteristics may be applied to any collection of entities with reciprocal quotations and references. It assigns a numerical weighting to each element of a linked set of documents with the purpose of measuring its relative importance within the set. Betwenness centrality characterizes gatekeeping papers that are given priority attention in this study. Other commonly used metrics such as degree centrality and closeness are considered to provide additional characteristics of the core elements. --- RESULTS Following the proposed strategy, first the overall publication dynamics was considered. As Fig. 2 shows, the most active period of growth in both areas is observed in the decade from 2001 until 2010. Since 2005 publications on emerging technologies have begun to dominate those on advanced technologies and after 2010 trends go in different directions. While the latter area seems to lose its importance, the former continues to grow at the same rate. The number of papers written in each period for further co-citation analysis is provided in Table 1. Before the 1990s, publication sources did not belong to only social or natural sciences . As such, 'American Helicopter Society Proceedings', 'Science', 'Rail International', 'European Journal of Operational Research' were primarily sources of information on both areas during the first period. Starting in the second period , a certain differentiation of sources began to take place. While papers on advanced technology were mostly concentrated in natural sciences journals studies, emerging technologies were wider cited among periodicals in social sciences. The identification of sources relevant for both domains such as 'Annual Forum Proceedings -American Helicopter Society' indirectly points out that at this stage, advanced and emerging technologies could have been considered similar categories or even synonyms. The co-citation networks of the two technological areas had not differed significantly from the perspective of networks metrics . For all periods discussions on emerging technology demonstrate higher growth of intensity of communications. However, the density of a network that characterizes the speed at which information diffuses among the nodes as well as the extent to which actors have high levels of constraint is higher for advanced technology networks in all of the considered periods. While for the earlier periods this can be explained by a lower number of papers and therefore tighter communication in cliques, higher density of co-citations in advanced technology against other parameters of the network indicates a slower growth rate and poorer penetration in a wider range of discussions. Established once in the last decade of the twentieth century, this area of research seems to remain constrained from further integration into the agenda of social sciences. Since the primary intention of this paper was to assess the evolution of the academic discussion about advanced and emerging technologies, the authors focused on the similarities and differences of the knowledge base between concurrent periods. This allowed for tracking the dynamics of network structures, including the identification of key actors and therefore sets of problems connecting discussions about emerging and advanced technologies. Table 5 describes the dynamics of co-citation networks: rows indicate periods from which references move, while columns show the ones to which they move. The area of emerging technology studies demonstrates a higher number of transitionary references. Moreover, the number of such references for the two decades increased by more than 10 times, which confirms the previous observations on better penetration of this area within the social sciences discourse. Five references from the emerging technology co-citation networks moving from 1991-2000 to 2001-2015 include three papers that conceptualize technological evolution in economic terms and two works on the dynamics of innovation . Studies cited both in 2001-2010 and 2011-2015 were focused on a variety of areas. Authors wrote about the various aspects of technological development, including questions concerning the philosophy of science and technology , technology acceptance , innovation development , approaches to technology forecasting , public perception of technology , and specifically nanotechnology acceptance . Out of the six references moving from 1991-2000 to 2001-2015 in the co-citation network of advanced technology three coincide with the transitionary references in the emerging technology network of the same period , three others refer to the critical analysis of innovation typology , reflections on knowledge development of a firm and the problem of international technology diffusion . Descriptive statistics and centrality values have not shown a significant difference between the two analyzed technology areas. The majority of references were cited from two to three times with no substantial discrepancy between minimum and maximum number of citations. Betweenness centrality values as well as other centrality metrics taken into account did not affect the referencing of papers in the two concurrent periods. As Figure 4 shows, centrality values are distributed more or less equally for stable and transitionary references. Therefore, referring to a paper during a longer period is not related to social networks properties . As seen from the contents of papers appearing in two concurrent periods, both in the case of advanced and emerging technologies, authors refer to studies addressing a wide range of problems concerning technological and innovative development starting from general philosophical issues to more narrowly focused questions concerning the adoption of specific technologies. Although one can see slightly better dynamics in core papers that inspired discussions on emerging technologies, this can hardly be associated with their gatekeeping position in co-citation networks. Moreover, other centrality metrics do not demonstrate any significant difference between transitionary and stable references. The latter means a coincidental use of both concepts and therefore the natural variety of their use in the literature. All of this demonstrates that discussions on advanced and emerging technologies can hardly be considered separate research fields. In both cases, one can observe stochastic referencing or its synthesis in relation to a specific problem with which a group of technologies with some common, but abstract characteristics can be associated. --- CONCLUSIONS New technologies are crucial for national economic growth and the improvement of our daily lives. In order to better understand the effects of technology implementation, reveal technological trends and manage technology growth, new developments require proper assessment and monitoring. However, as it was shown, the boundaries of the concept of 'technology' especially when it is connected with some advancements are vague, which leads to difficulties in measurements and increases confusion among decision makers. The paper presented the evolution of academic discussions concerning two groups of technologies namely 'advanced' and 'emerging', which were identified through the analysis of transitionary references in co-citation networks for several decades throughout the overall period from 1955 to 2015. This allowed for the identification of key alterations in the professional discourse in social sciences about these groups of technologies in order to assure their development as separate research fields. It was shown that conceptual framing of both concepts is a stochastic process and that social studies in related areas can hardly constitute separate research fields. Papers on advanced technology, though addressing social issues, have a stronger connection with natural sciences while studies in emerging technologies demonstrate a wider penetration among periodicals in social sciences journals. A further analysis of co-citation networks pointed to a lack of common background for advanced technology papers in almost all of the observed periods, only those which were cognitively connected through a limited number of transitionary references. Studies in emerging technologies demonstrated a better consistency in terms of reproduction of citations in related co-citation networks. In particular, three connected periods were identified for which the number of transitionary references was growing. Despite this, centrality characteristics of such references as well as other characteristics of co-citation networks can hardly confirm existence of separate research fields of study. Moreover, traditions that feed the current research agenda on emerging technologies are too diverse and weakly connected. In conclusion, the analysis showed that at the current stage, existing discussions in both areas cannot be considered separate research fields. Though discussions in both areas demonstrate stochastic referencing and lack replicability of co-citation networks, emerging technologies demonstrate the primary attributes of a developing research field. The suggested approach including the identification and measuring of network characteristics of transitionary references can be used for the analysis and predictive modeling of specific technological domains. However, a keyword analysis or advanced text-mining techniques may provide more evidence about the technological content of promising areas. --- Advanced technology Emerging Multidisciplinary journals such as 'Nature' and 'Science' became the primary publication sources for advanced technology studies after 2000, while emerging technology studies were concentrated either in the conference proceedings of engineering sciences or in management journals. The overall use of the term 'emerging technology' is higher across social sciences discourse while the term 'advanced technology' is better established in natural and engineering sciences journals. In order to identify existence of a communication core in each of the subsets of data, eight cocitation networks were constructed (

, How the analysis of transitionary references in knowledge networks and their centrality characteristics helps in understanding the genesis of growing technology areas. Scientometrics,

Background Sleep disorder is common in later life and various epidemiological studies have reported that it affects 36-70% of community-dwelling older adults [1,2]. The prevalence of sleep disorders increases dramatically as people reach older adulthood [3], and sleep disorders are strongly linked to risks of mortality, falls, cognitive impairment, and depression [4][5][6][7]. The American Geriatrics Society recommends the use of nonpharmacological tools to improve sleep quality [8]. Exercise is a representative nonpharmacological intervention, and high levels of daily physical activity have been found to be associated with a lower prevalence of sleep disorder in older adults; however, the effects of exercise in this regard depend on the intensity, duration, and time of day at which it is performed [9][10][11]. In addition, some epidemiological studies have suggested that high levels of housework and/or work-related activities are associated with a lower prevalence of sleep disorder [12,13]. These studies suggest that, similar to exercise, housework and/or work-related activities are positively associated with thermoregulation and/or regulation of the circadian system, which consequently improves sleep quality in older adults [12,13]. Physical activity, including exercise, is a free or low-cost method of promoting health benefits, especially physical and cognitive function, mood, and sleep, and has a low risk of side effects in older adults [14]. Sleep disorder is associated with poorer physical functioning among older adults [15,16]; however, the mechanism by which disturbed sleep impacts physical functioning is currently unknown. It is possible that the relationship between physical activity and sleep quality is mediated by physical functioning [16]. Previous studies have found that a narrow range of social relationships is associated with an increased risk of mortality [17][18][19]; this is because long-term social isolation can create chronic stress [17]. Conversely, some epidemiological studies have reported that a wide range of positive social relationships is related to lower prevalence of sleep disorder [20,21]. An epidemiological study revealed that older adults who have a large number of positive social relationships for at least 15 years have better sleep quality than those with shorter-duration relationships [21]. Furthermore, combining low-intensity physical activity with social interaction has been found to increase the prevalence of slow-wave sleep and improve cognitive function among older adults [22]. In later life, older adults naturally have fewer social relationships ; the above findings suggest that older adults should seek to counteract this by endeavoring to maintain or expand their range of social relationships, as this is an important factor for their quality of life. Physical activity and social relationships for older adults are positively related to not only physical and cognitive functions and mental health but also sleep quality. Both variables are positively correlated [23]; however, there is insufficient previous research considering the confounding factors for each of these variables [9,10,20]. Furthermore, some older adults lack either social relationships or physical activity, or both. Considering this finding and previous results, we hypothesized that the prevalence of sleep disorder is lower among people with a low level of one of these variables and a high level of the other, compared to those who are inactive and have a narrow range of social relationships. Thus, to identify the relationship between sleep quality and a high level of one of these variables with a low level of the other, we examined the combined effect of physical activity and social relationships on the prevalence of sleep disorder among community-dwelling older adults in Japan through a cross-sectional and longitudinal study . --- Materials and methods --- Participants and data collection Both Studies 1 and 2 were based on data from the Kasama Study, which was a community-based cohort study conducted in Japan between 2011 and 2019 [24]. Participants for each year of the study were randomly selected from the Basic Resident Register using the following eligibility criteria: aged 65 years or older, not a recipient of long-term care insurance, and living in Kasama City, Japan. A total of 1094 community-dwelling older adults participated in Study 1, of whom 106 were excluded because of incomplete data. Ultimately, 988 participants were included in the analysis. In Study 2, we excluded 575 older adults who had not participated more than twice during 2011-2019, a further 199 participants who had a sleep disorder, and 25 who had incomplete data. Finally, 351 participants were included in Study 2's analysis. --- Daily physical activity The Japanese version of the Physical Activity Scale for the Elderly [25] was used to assess participants' daily physical activity. The PASE is a 12-item questionnaire that measures the average hours per day spent performing leisure-time, housework, and work-related physical activity, respectively, over the previous 7 days. Leisuretime physical activity includes walking; light-, moderate-, and vigorous-intensity recreational activities; and muscle-strength training. Housework includes light and heavy housework, home repair, lawn work or yard care, outdoor gardening, and caring for other people. Finally, work-related physical activity includes paid and volunteer work. These items are weighted based on the intensity of each activity, and the total PASE score is the sum of the 12 weighted items [25]. This study used the PASE total score to determine participants' overall physical activity levels [26]. --- Social relationships To assess social relationships, we used the Japanese version of the Lubben Social Network Scale-6 [27]. The LSNS assesses social relationships by measuring the strength of respondents' family ties and friendship ties, respectively, in terms of each of the following three categories: "How many people do you see or hear from at least once a month?" "How many people do you feel sufficiently at ease with such that you can talk with them about private matters?" "How many people do you feel sufficiently close to such that you could call on them for help?" All questions were answered using a six-point Likert scale, where 0 = "none, " 1 = "one, " 2 = "two, " 3 = "three or four, " 4 = "five through eight, " and 5 = "nine or more". The LSNS score was determined by summing the scores for all six items [27]. Due to the quantitative estimates of social relationships, this study used continuous data divided into median or tertiles [28,29]. --- Sleep disorder We used the Pittsburgh Sleep Quality Index to identify sleep disorder; this tool has been used in both clinical research and epidemiological studies [30]. The PSQI measures seven components: subjective sleep satisfaction, sleep efficiency, sleep onset latency, sleep duration, sleep disturbances, use of sleeping medication, and daytime dysfunction. The score for each component is weighted by 0 to 3, and the PSQI global score is determined by summing all items . A previous study revealed that with a cut-off of 5/6, the PSQI global score has a sensitivity of 89.6% and a specificity of 86.5% for identifying cases of sleep disorder [30]. In this study, this cut-off value was used to define an individual having a sleep disorder. --- Physical functions To assess physical function between the groups in Study 1, we conducted five physical performance tests: grip strength, one-leg standing duration, time-up and go test, 5-m walk test, and 5 times sit-to-stand test. Grip strength was measured twice on each hand by a grip dynamometer and the average of the best values of each hand was adopted. To evaluate static balance ability, we conducted a one-leg standing with eyes open test. Participants put both hands on their waist and gradually raised their preferred foot in front of them to approximately 20 cm above the floor. They maintained this position as long as was possible . The records were captured twice, and the highest record was adopted. To assess dynamic balance ability, we used the time-up and go test. Participants rose from a chair, walked 3 m as quickly as possible, turned around, walked back, and sat down. We assessed usual gait speed using the 5-m walk test. Participants walked at their typical speed on an 11-m straight course. To eliminate acceleration phases, we calculated walking time between the 3-and 8-m marks of the course. To assess lower-limb muscle strength, we conducted the 5-times sit-to-stand test. Participants sat on a chair with their arms over their chest. They carried out 5 consecutive sit-to-stand cycles as quickly as possible. The time-up and go test, 5-m walk test, and 5 times sit-to-stand test were conducted twice, and the faster record was adopted [31,32]. --- Potential confounders To identify potential confounders, we referenced previous studies [33] and included measures of age, sex, and body mass index ; use of hypertension, psychotropic, diabetes, and sleep medication, respectively ; medical history of lower back pain, knee pain, and hip pain, respectively ; alcohol consumption ; tobacco-smoking status ; and presence of depressive syndrome [34]). --- Statistical analysis --- Cross-sectional study Linear trends in the prevalence of sleep disorder were computed using ordinal scoring for the PASE and LSNS scores, respectively [35]. For both scales, scores were divided into tertiles; for the PASE score, 0-94.3 points represented the "first tertile"; 94.5-140.4 points represented the "second tertile"; and 140.6-462.3 points represented the "third tertile"; meanwhile, for the LSNS score 1-15 points represented the "first tertile"; 16-19 points represented the "second tertile"; and 20-30 points represented the "third tertile. " Additionally, to test the relationship between the combined two factors and sleep disorder, the two factors' medians were calculated for the respective scores, and these were used to develop the following participant groups: low activity group with low social relationships ; low activity group with high social relationships ; high activity group with low social relationships ; and high activity group with high social relationships . To compare the four groups, we used one-way analyses of variance with Bonferroni post-hoc tests for continuous variables and chi-square tests for categorical variables including physical functions. We used logistic regression analysis to examine the association between physical activity and social relationships, individually and in combination with each other, on the risk of sleep disorder. First, in the tests of linear trends, we adopted two models. Model 1 was adjusted for age, sex, BMI, use of hypertension and sleeping medication, smoking and drinking habits, and depressive syndrome. There was a significant correlation between the PASE and LSNS scores . Thus, along with the adjustments contained in Model 1, Model 2 featured additional adjustments for physical activity variables when examining social relationships. Further, additional adjustments for social relationship variables when examining physical activity were applied. Second, only Model 1 was used to examine the combined association between physical activity and social relationships on sleep disorders. We calculated odds ratios and 95% confidence intervals . --- Longitudinal study Similar to Study 1, to test the combined effect of the two factors, the medians were calculated for the respective scores, and these were used to develop the following participant groups: low activity group with low social relationships ; low activity group with high social relationships ; high activity group with low social relationships ; and high activity group with high social relationships . To investigate the combined effect of the two factors on sleep disorder, we conducted Cox proportional hazards regression analysis and calculated hazard ratios and 95% CIs. The independent variable was the low activity group with low social relationships as the reference, and duration of existence was determined for this cohort. Further, sleep disorder occurrence was initially defined when diagnosed for the first time during the follow-up period. Study 2 adopted only Model 1. All analyses were performed using IBM SPSS Statistics version 26.0 . --- Results The mean age of the sample in Study 1 was 73.1 ± 5.5 years, and 56.3% of the participants were female. Individuals with sleep disorder accounted for 36.8% of the total sample. Table 1 shows the demographic characteristics of the four participant groups. The active group with a narrow range of social relationships featured a significantly lower ratio of females, while the active group with a wide range of social relationships featured a significantly higher ratio of females than the other groups. Both high activity and low activity groups with high social relationships showed significantly lower GDS scores than the low activity with low social relationships. In addition, the high activity group with high social relationships showed a significantly lower GDS score than the high activity group with low social relationships . Naturally, the LSNS and PASE scores significantly differed among the groups because of the participant group formation methods used. Notably, however, the high activity group with high social relationships had a significantly higher PASE score than the high activity group with low social relationships . In terms of physical functions, the high activity group with high social relationships shows significantly positive value compared to the low activity group with low social relationships on all of the physical functions . Moreover, compared to the low activity group with low social relationships, the high activity group with low social relationships on the one-leg standing duration shows a longer duration, and low activity group with high social relationships on the 5 times sit-to-stand test shows shorter duration . Table 2 presents the association between physical activity and social relationships on sleep disorder. In the unadjusted model and Model 1, both variables show linear trends regarding the high risk of sleep disorder. However, the linear trends are significantly different in Model 2, which features adjustments for the other variable . In Model 1, the unadjusted model, and Model 2, the high risk of sleep disorder is significantly lower in the third tertile of both physical activity and social relationships than in the first tertile of the respective variable. Table 3 shows the combined association between physical activity and social relationships on sleep disorder. The only significant difference in Model 1 concerns the high activity group with high social relationships, which features a 41.5% lower risk of sleep disorder than the low activity group with low social relationships. The other groups show no significant differences from the low activity group with low social relationships . The mean age of the sample in Study 2 was 72.4 ± 5.0 years, BMI was 23.1 ± 2.9 kg/m 2 , and 49.6% of the participants were female. During a mean follow-up of 3.7 years , sleep disorder occurrence was 96 participants . Figure 1 and Table 4 show the combined effect of physical activity and social relationships on sleep disorder. Figure 1 shows the adjusted Kaplan-Meier curves for sleep disorder by group among all participants in Model 1. Table 4 shows that the HRs of the high activity group with high social relationships are significantly lower than that of the low activity group with low social relationships in Model 1. --- Discussion This study examined sleep disorder among older adults in terms of its association with daily physical activity and social relationships, both individually and in combination with each other using cross-sectional and longitudinal studies. Our study showed that a high level of physical activity and high level of social relationships may, independently of each other, positively affect sleep quality . This result is similar to previous findings that low levels of physical activity and narrow ranges of social relationships are related to poor sleep quality in older adults [9,20]. However, unlike these previous studies, we adjusted for the effect of one variable when measuring the other. The risk of sleep disorder was significantly lower in the high activity group with high social relationships, compared with the low activity group with low social relationships . These results do not support our hypothesis that high levels of one variable with low levels of the other results in a lower prevalence of sleep disorder than low activity levels and a narrow range of social relationships. Sleep disorder and depressive syndrome are strongly linked [36], and the rates of both are significantly higher among older adults than younger people [37]. Overall, 83% of patients with depression have experienced some form of sleep disturbance [36] and, although the effects of daily sleep quality on next-day mood are larger than the effects of daily mood on sleep quality, sleep quality and mood share a bilateral relationship [38]. Some previous studies have shown that strategies for regulating mood control can improve sleep quality [39]. In this study, although no significant differences were observed, the groups with low levels on one variable showed 14-26% lower depressive syndrome scores than the low activity group with low social relationships. Meanwhile, the high activity group with high social relationships showed a significantly lower value than the low activity group with low social relationships and the high activity group with low social relationships in Study 1. The significant difference in sleep disorder between the low activity group with low social relationships and the high activity group with low social relationships disappeared after full adjustment for covariates, including depressive syndrome. However, the difference between the low activity group with low social relationships and the high activity group with high social relationships remained significant . Similarly, Model 1 in Study 2 shows a significant difference in sleep disorders between the low activity group with low social relationships and the high activity group with high social relationships . Researchers have highlighted several potential mechanisms by which physical activity affects sleep quality, including energy conservation, body and central nervous system temperature increases, and anxiety reduction [40][41][42]. In addition, some previous studies have mentioned that the relationship between physical activity and sleep disturbances might be mediated by physical functioning [15,16]. High levels of both variables indicated significantly better health across all physical functions compared with low levels. These results demonstrate that high levels of physical activity and sleep disturbance may have a positive effect on sleep quality mediating physical functions . Meanwhile, the effects of social relationships on sleep quality are not fully understood. Theoretically, however, strong social relationships provide an evolutionarily adaptive function, creating a safe environment in which sleeping persons feel they are protected by others from dangers [20]. Recent studies have also hypothesized that social isolation causes chronic stress [17], directly affects anxiety, and decreases sleep satisfaction [20,21]. Social isolation is related to lower physiological functioning and higher risks of physical disorders [17]. In particular, addressing deficits in terms of social relationships and physiological activity may directly arrest early progression toward chronic diseases, and could also delay disease onset and lessen disease burden in later life [17]. The above findings suggest that physical activity and social relationships produce overlapping, synergistic anti-anxiety and anti-depressant effects. In a previous study, the present authors revealed that communal exercise has a lower hazard ratio for five-year morality than no exercise [43]. The present study suggests that physical activity and exercising with others [44,45] may have a synergistic effect on not only physical functions, but also mental health and sleep quality in older adults. A previous epidemiological study showed that intellectual, social, and recreational activities are correlated with cognitive function [23]. Indeed, an experimental study revealed that low-intensity exercise with social interaction not only has positive effects on sleep quality, but also improves cognitive function, particularly attention ability, in older adults [22]. It has been postulated that exposure to social relationships and engaging in physical activity increase blood flow to the brain, improving neuronal function and facilitating deep sleep [22,23]. We suggest that future studies should examine whether the effects of physical activity and social interaction on older adults' sleep quality are mediated by improved cognitive function; methods such as polysomnography and functional near-infrared spectroscopy could be used to perform this examination. Our study has several limitations. First, although this study examined the lower prevalence of sleep disorder among individuals who are active and have a wide range of social relationships , we could explore the causal relationship; however, we could not explain the mechanism. Thus, this aspect warrants further investigation through an experimental study or an intervention study, for example. Second, we did not examine whether supportive and aversive social relationships have different effects on sleep disorder. In this study, we only focused on supportive relationships because a previous study revealed no significant association between aversive relationships and sleep quality [20]. Generally, the LSNS cut-off is used for social isolation with less than 12 points [27], and high scores on the LSNS indicate abundant social networks [46]. However, since we aimed to ensure statistical power and for ease in interpretation of the results, we used continuous data divided into median or tertile [28,29]. Especially, the PASE has subscales such as exercise types , household activity, and work-related activity [12,26]. Nevertheless, further studies should consider the kind of activity and adverse relationships in this regard. Third, all variables were assessed using self-report scales; thus, our findings may have been influenced by recall and reporting bias. Fourth, physical activity and social relationship factors may be influenced by individual differences. Thus, future studies should consider gender differences and implement stratification in terms of physical function and the presence of depressive syndrome. Further, the type of physical activity and/or social relationship is not only affected by individual differences, but also by cultural and/or regional differences [27]. Thus, the results of the present study need to be carefully interpreted. Finally, although a previous epidemiological study involving a Japanese population [1] reported the same prevalence of sleep disorder as in the present study, the generalizability of our results is uncertain because of the small sample size. --- Conclusions In conclusion, high levels of physical activity and social relationships are independently related to good sleep quality. Our findings suggest that being highly active with a wide range of social relationships is favorably associated with sleep quality. However, a high level of one variable with a low level of the other, has not been confirmed to improve sleep quality among older adults. These results suggest the importance of considering both, physical activity and social relationships, when seeking to improve sleep quality among older adults. • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year --- --- --- --- --- Competing interests The authors have no conflicts of interests to declare. ---

Background: This study investigated whether daily physical activity of older adults, combined with social relationships, is associated with the risk of sleep disorder. Further, it determined whether a high level of one variable with a low level of the other, leads to a significantly lower risk of sleep disorder than low levels of both.The sample comprised 1339 community-dwelling older Japanese adults: 988 in Study 1 and 351 in Study 2. The level of daily physical activity and range of social relationships were assessed using the Physical Activity Scale for the Elderly and the Lubben Social Network Scale, respectively. The Pittsburgh Sleep Quality Index was used to assess sleep disorder. To test the combined relationships and effects in Studies 1 and 2, the medians for the respective scores of each of the following four groups that the participants were categorized into, were calculated: (1) low activity group with low social relationships, (2) low activity group with high social relationships, (3) high activity group with low social relationships, and (4) high activity group with high social relationships. After adjusting for potential confounders, a logistic regression analysis was conducted in Study 1. After adjusting for potential confounders, a Cox proportional hazards regression analysis was conducted in Study 2. Results: Study 1 revealed that the high activity group with high social relationships showed a significantly lower risk of sleep disorder (ORs: 0.585, 95% CI: 0.404-0.847) than the low activity group with low social relationships. Study 2 also revealed that the high activity group with high social relationships showed a significantly lower prevalence of sleep disorder (HRs: 0.564, 95% CI: 0.327-0.974) than the low activity group with low social relationships.Our findings suggest that for older adults with high social relationships, being physically active is favorably associated with sleep quality. However, a high level of one variable with a low level of the other has not been confirmed in improving sleep quality among older adults.

Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. --- Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request. Social justice organizations play an important role in challenging systems of groupbased inequality, as they work to coordinate resources and labor to advocate for the interests of low-status groups . Social psychological theories of collective action have traditionally assumed that participants in such social justice efforts come from the low-status group that stands to benefit from challenges to the status quo . These frameworks have viewed members of high-status groups, in contrast, as seeking to maintain the social hierarchies that benefit their groups . The theoretical focus on social change actors from low-status groups mirrors historical trends in social movement participation . However, some members of high-status groups do take solidarity actions and participate in social justice efforts that benefit low-status outgroups . Increasingly, members of high-status groups also take up positions of leadership in these organizations. While 58% of American non-profit groups serve racial/ethnic minority communities, for example, White people make up 82% of executives and 83% of board members in these organizations . Similarly, trustees of British not-for-profit organizations typically belong to high-status groups, with 92% being White and 67% being men . Members of high-status groups who lead social justice efforts wield a great deal of power and influence. Leaders oversee the development of these organizations' goals and strategies , and can serve as the public face of the movement . To our knowledge, however, no social psychological research has investigated the impact of having high-status group leaders take on such influential roles in social justice efforts. Does the presence of high-status group leaders facilitate efforts to mobilize observers to contribute to the social justice effort, or does it instead dampen observers' collective action intentions? This question has not been addressed in the literature, which to date has focused on two other issues. One line of work has delineated the factors that predict high-status groups' participation in social justice efforts . Other studies have examined low-status group members' responses to individual allies from high-status groups who support the social justice effort . This paper considers the intergroup dynamics that underpin the impact of high-status group leaders' involvement in social justice efforts that seek to benefit a low-status group. Because high-status groups occupy positions of power in society , their participation can contribute to the success of a social justice effort by providing access to resources and networks , and by showcasing the broader justice values represented by the movement . However, high-status group leaders' contributions to the social justice effort can be limited because they do not have lived experience of the systemic disadvantages being tackled. We propose that when observers perceive these kinds of problems with high-status group leaders, they will be less willing to join the social justice effort. We further expect these negative responses to be especially pronounced among observers from low-status groups. We develop these two arguments in turn below. --- Collective Action Intentions in Response to Leaders of a Social Justice Effort We expect that observers will be less willing to join a social justice effort whose influential leaders belong to the high-status group , because they perceive particular problems with leaders from high-status groups and experience more negative emotional responses. Our model specifies emotions to be the direct predictors of collective action intentions, because emotions motivate actions in response to stimuli that are relevant to an individual . Discrete emotions, in particular, have been linked to goals and action intentions in the context of interpersonal relations and intergroup relations . Most relevant to the context of social justice efforts, emotions such as anger and hope have been shown to predict individuals' collective action intentions . We propose that observers' perceptions of particular problems with high-status group leaders will elicit specific emotional responses. Appraisal theories propose that emotional responses are based on distinct cognitive judgments about a situation or relationship . Research has also shown that emotions mediate the relationship between appraisals and collective action intentions . Below we set out the specific appraisals and emotions that should be most relevant in predicting observers' collective action intentions to work with a social justice effort led by high-status group leaders. Figure 1 presents the conceptual model. --- Proposed Process Perceived Leaders' Awareness of Inequality. Research on responses to inequality suggests that when high-status group leaders are engaged in a social justice effort that benefits a low-status group, observers are likely to perceive a particular problem: low awareness of inequality. Leaders from the high-status group by definition have not experienced the disadvantages being tackled by the social justice effort, and thus may be perceived as not fully understanding the experiences of the disadvantaged group. For instance, research shows that members of high-status groups are not well attuned to subtle instances of bias experienced by low-status groups . Compared to members of low-status groups, members of high-status groups are typically less aware of prejudice and perceive inequality to be more legitimate and less pervasive . Indeed, members of high-status groups are perceived to be less willing to engage in a social justice effort than are members of low-status groups . Thus, observers may believe that high-status group leaders lack a full understanding of the problem being tackled by a social justice effort, compared to leaders from the low-status group. Hope. Perceiving high-status group leaders as being less aware of inequality should result in lower levels of collective action among observers because they feel less hope . Hope is experienced when an individual believes that a situation can change and has positive feelings about the anticipated outcome . Appraising high-status group leaders as having lower awareness of inequality should reduce observers' belief that social change will occur, and thus reduce feelings of hope . Because hope combines a desire for change with a sense of efficacy, it is conceptualized as an action-oriented emotion in interpersonal contexts and intergroup contexts . Research has shown that hope predicts support for social change and collective action intentions , especially when individuals are focused on social justice . Given the positive orientation of hope, it should be a strong predictor of collaborative collective action to work with a social justice effort. --- Alternative Processes The literature suggests two alternative processes within our framework that might underpin observers' unwillingness to contribute to a social justice effort led by members of the high-status group . --- Perceived Alienation of Low-Status group and Perceived Leaders' Ulterior Motive. Research on the psychology of relative advantage suggests that observers could perceive high-status group leaders as alienating members of the low-status group whose interests they should be promoting. For instance, high-status group leaders might support goals and strategies that are inconsistent with low-status groups' preferences: Research shows that, compared to members of low-status groups, members of high-status groups are less willing to challenge inequality , and prefer more conventional-and in some cases non-political-strategies to address inequality . Alienation of the low-status group can also occur because high-status group leaders are accustomed to holding a position of power in society and thus may engage in behaviors that reproduce status distinctions within the social justice effort ; examples include dominating discussions and crowding out voices of the low-status group . Observers might also perceive high-status group leaders' motives as not entirely focused on achieving social equality. While members of high-status groups likely believe in the principles of fairness and justice, research shows that their prosocial contributions can be based in other strategic aims , such as advancing their career or protecting the moral reputation of their high-status ingroup . Observers who perceive such ulterior motives are likely to find them problematic, as the leaders will not appear to be genuine representatives of the core values associated with the social justice effort. Anger. These two perceptions of high-status group leaders-alienation of low-status group members and ulterior motive-point to clear transgressions that violate accepted norms of fairness and justice, and thus should increase observers' feelings of anger. Appraisal theories posit that anger is based in judgements of injustice or unfairness, with evidence for this relationship documented in the context of interpersonal and intergroup relations. Anger is associated with the goal of challenging the agent responsible for the transgression or harm . This work suggests that anger about high-status leaders' shortcomings might predict lower levels of collective action to work with the organization they represent. Given our focus on collaboration with a social justice effort, however, we suspect that the processes via anger will not be sufficient to explain observers' willingness to take action. In the literature, anger has been linked to confrontational action that seeks to disrupt the status quo . Thus, we would expect anger to predict a more direct form of challenging the problematic leaders . In contrast, collaborative collective action intentions to work with the justice effort should be predicted by a positive emotion such as hope. --- Observers' Group Status Shapes Responses to Leaders of a Social Justice Effort A social identity analysis posits that observers' responses should be moderated by their own group membership . Building on this perspective, we propose that observers from the low-status group will report strong negative responses to high-status group leaders relative to low-status group leaders. Members of low-status groups should be motivated to consider the implications of high-status group leaders' presence in social justice efforts that benefit the low-status group, because their ingroup is directly affected by this action. For instance, research shows that women report stronger emotional responses than do men when they read about gender inequality , which suggests that such inequality is more relevant to members of the low-status group . Observers from the low-status group are also likely to recognize the particular problems created by high-status group leaders. Compared to members of highstatus groups, members of low-status groups are more aware of bias and discrimination in contexts of racial inequality and gender inequality . Low-status group members have lower threshold standards than do high-status group members for identifying injustice against the low-status group and harm done by the high-status group . Thus, we expect observers from the low-status group to report lower levels of collective action intentions, more negative emotions, and higher levels of perceived problems in response to high-status group leaders, compared to low-status group leaders. In contrast, we expect observers from the high-status group to report more equivalent responses to leaders from the high-status group and low-status group. The aforementioned research indicates that members of high-status groups will be less aware of high-status group leaders' limitations in the context of a social justice effort that seeks to benefit a low-status group . High-status group members may also pay less attention to the potential for problems created by leaders from the high-status group, because they are less directly affected by the disadvantage . As such, we expect that observers from the high-status group will not differentiate between high-status group leaders and low-status group leaders in their responses, with respect to collective action intentions, emotions, and perceived problems. --- Alternative Accounts of Low-Status Groups' Negative Responses to High-Status Group --- Leaders Theoretical frameworks in the extant social psychological literature would suggest that low-status group members' negative responses to high-status group leaders are not unique to contexts of solidarity, but instead could be explained by two more general processes. Ingroup Bias. One account posits that low-status group members might simply prefer leaders from their own group, compared to leaders of an outgroup. Individuals' bias towards their ingroup has been documented among both minimal groups and real-world groups . People also respond more favorably to leaders from their ingroup compared to leaders from an outgroup . According to the ingroup bias explanation, low-status group members would report negative responses to high-status group leaders in any organizational context, regardless of whether their presence represents solidarity with the low-status group or a more general social justice orientation . --- Role Congruity. A second plausible explanation for observers' preference for lowstatus group leaders draws on the concept of role congruity . High-status group leaders could be seen as generally unsuitable to lead social justice efforts that benefit an outgroup, because they do not have the physical, intellectual, or personality characteristics that are stereotypically considered to be effective in this area. This perspective predicts that negative responses to high-status group leaders' presence in solidarity contexts would be equivalent to the negative responses documented for other types of low role congruity . We investigate the extent to which these alternative accounts can explain the pattern of results predicted by our model, as described below. --- Overview of Hypotheses and Studies Our framework sets out three hypotheses regarding observers' responses to influential high-status group leaders in a social justice effort that benefits the low-status group. Compared to observers from the high-status group, we expect observers from the low-status group to more clearly differentiate between influential leaders from the high-status group and influential leaders from the low-status group, with respect to collective action intentions and the hypothesized mediator variables: perceived leaders' awareness of inequality and hope . We also predict that perceived leaders' awareness of inequality and hope will mediate observers' lower collective action intentions in response to high-status group leaders compared to low-status group leaders , because of our focus on collaborative actions; furthermore, we evaluate the extent to which the alternative process variables-perceived alienation of low-status group, perceived leaders' ulterior motive, and anger-mediate this relationship. The first two studies test these hypotheses in the context of a non-profit organization that addresses gender or racial/ethnic inequality. Participants from either the low-status group or high-status group read that influential leaders belong to either the low-status group or the high-status group, and then complete measures of collective action intentions, perceived problems with the leaders, and emotions. Study 3 provides an experimental test of the hypothesized process underpinning lowstatus group members' lower collective action intentions in response to high-status group leaders. In the context of a social justice effort that benefits the low-status group, we orthogonally manipulate the influential leaders' group background and their awareness of inequality. We expect that observers who read about highstatus group leaders with high awareness of inequality will feel more hope, which will predict increased collective action intentions . Lastly, we consider alternative explanations for our hypothesized effects, by investigating whether low-status groups' negative responses to high-status group leaders' role in solidarity contexts can be explained by accounts of ingroup bias and role congruity. Study 4 compares low-status group members' responses to high-status group leaders in a solidarity context versus two nonsolidarity contexts that represent either high or low levels of general role congruity. We predict that low-status group members will report more negative responses to high-status group leaders in the solidarity context compared to the non-solidarity contexts , with respect to perceived problems, felt emotions, and collective action intentions. We also expect that these differences will hold regardless of the role congruity associated with each domain in the non-solidarity contexts . Findings consistent with these hypotheses would indicate that negative responses to high-status group leaders are especially pronounced in solidarity contexts, and thus extend beyond preferences predicted by frameworks of ingroup bias and role congruity. --- Study 1 We predicted that observers' group membership would moderate their responses to leaders from the high-status group compared to leaders from the low-status group, with respect to collective action intentions as well as perceptions of the leaders and emotions . In the context of a non-profit organization that combats gender inequality, the composition of the leadership team was manipulated to include either more women or more men . We designated one gender group as the numerical majority in order to demonstrate that leaders from this group had meaningful influence in the organization. Participants who self-categorized as male or female reported their responses to the leaders who made up the numerical majority on the leadership team. Our prediction regarding the mediation model had not been developed when we designed this study. Thus, our investigation of the process underpinning lower collective action intentions in response to high-status group leaders was fully exploratory. This approach dictated two aspects of how we report this study. First, we do not differentiate between the hypothesized mediators and alternative mediators in the Method section. Second, we report all analyses for the tests of experimental effects and mediation models in the main text of the Results section. --- Method --- Design The study employed a 2 x 2 between-participants personby-treatment quasi-experimental design. Participants were randomly allocated to one of the two leadership composition conditions and reported their gender group. Every participant self-identified as either male or female, with none choosing the "other" option. --- Procedure Participants read a description of a fictional gender equality organization that manipulated the leadership team's gender composition. They then completed a manipulation check item and measures of their perceptions, emotions, and collective action intentions. --- --- Materials Participants read about a fictional organization that sought to combat gender inequality in the United States. The first paragraph described the aims and strategies of the organization: The Gender Equality Network is a national grass-roots organization that works to improve the lives of millions of women and their families in the United States. GEN advocates for policies that advance equity for girls at all levels of education; challenges gender bias and discrimination in higher education and the workplace; and provides leadership development opportunities and mentoring for women in higher education. The second paragraph described the leadership team's responsibilities in the organization, and the final sentence manipulated the gender composition: The Gender Equality Network is run by an Executive Leadership Team that is responsible for identifying strategic priorities, developing and implementing campaigns, and mentoring staff. This leadership team currently includes [more women than men or more men than women]. --- Measures The measures reported below were created using 32 out of the 95 items presented in the full questionnaire. The 63 omitted items were discarded for one of two reasons. First, 23 items originally intended to be included in a composite measure were dropped based on exploratory analyses; details are provided in relevant sections below. Second, 40 items were determined to be tangential to the aims of this paper, as they assessed either individual differences or general evaluations of the leaders and the organization . Further information can be obtained by contacting the first author. Manipulation Check. One item asked participants to recall the gender composition of the leadership team described in the manipulation: "The leadership team of the Gender Equality Network currently includes: more men than women; equal numbers of men and women; more women than men. Participants could choose only one response option. As noted in the Participants section, four participants provided responses that did not match the condition to which they had been assigned and were thus excluded from the sample. Collective Action Intentions . Five items were developed to assess the extent to which participants wanted to take action to support the organization: "I would like to donate money to this organization," "I would like to help this organization achieve its goals," "I would like to work with this organization," "I would like to contribute my time to this organization," and "I would like to get involved in campaigns run by this organization." Participants indicated the extent to which they agreed with each statement using a seven-point scale ranging from 1 to 7 . --- Perceptions of Leaders . To ensure that participants appraised the leaders who were in a position of power and influence in the organization, we directed them to focus on the leaders who made up the numerical majority on the leadership team. The instructions for the measures reminded participants which gender group the majority of the leaders belonged to, and asked them to focus on these leaders when responding to the items: "Using the response scale below, indicate the extent to which you agree with each statement about the leadership of the Gender Equality Network. Recall that the leadership team is made up mostly of {men or women}. Focus on these {male leaders or female leaders} when considering your answers." Responses were provided on a seven-point scale . Perceived Awareness of Inequality. Four items were developed to assess the extent to which participants believed that the leaders were aware of gender inequality and the experiences of women, and the extent to which the leaders were able to apply this knowledge to their role: "The leaders truly understand the struggles faced by victims of gender inequality," "I think that the leadership team represents those who are disadvantaged by gender inequality," "The leadership team probably has blind-spots that limit its ability to work for gender equality ," and "The leadership team will be able to speak for those who are disadvantaged by gender inequality." 1 Perceived Alienation of Women. Four items were developed to measure the extent to which participants believed that the leadership prevented women from taking a central role in the social justice effort and instilled distrust in the leadership team: "The make-up of the current leadership team reduces the voice available to women," "The current leadership team will make women feel marginalized," "Women might question the work done by the leadership of the Gender Equality Network," and "The leadership team probably faces skepticism from those who are victimized by gender inequality." We originally developed four additional items to assess the extent to which participants believed that the leadership would empower women . In an Exploratory Factor Analysis, these four items loaded onto a distinct factor from the four items assessing perceived alienation of women. However, the two measures were strongly correlated and produced nearly identical patterns of results in the main analyses. To reduce unnecessary duplication, we have dropped these items from the paper. Perceived Ulterior Motive. Three items measured the extent to which participants perceived a personal ulterior motive for the leaders' involvement in the organization: "The leaders probably took these jobs in order to advance their own careers," "I think that the leaders gain some personal benefits by being part of this organization," and "I wonder if the leaders have selfish reasons for being part of this organization." We had created a reverse-scored item , but dropped it as it reduced the reliability of the measure to below 0.60. --- Emotions . Participants were asked to indicate the extent to which they felt each of 27 emotions about the leadership of the Gender Equality Network, using a sevenpoint response scale . As with the previous set of measures, the instructions reminded participants of the gender composition of the leadership team and asked them to focus on the leaders in the numerical majority group when considering their responses. The emotion terms spanned a range of positive and negative emotions that represent a broad range of responses to inequality and efforts to challenge it . All 27 items were submitted to an Exploratory Factor Analysis using Maximum Likelihood extraction and Oblimin rotation. Two distinct factors emerged, which could be interpreted as hope and anger . Nine items loaded strongly and uniquely on the Hope factor: hopeful, inspired, optimistic, excited, thankful, confident, appreciative, pleased, grateful . Six items loaded strongly and uniquely on the Anger factor: angry, annoyed, irritated, outraged, furious, fed up . The remaining 12 items did not load strongly onto these factors or any other distinct interpretable factors, and thus were dropped from further analyses. --- Results --- Bivariate Correlations The measured variables were significantly associated with each other in expected directions, 2 with the exception of anger and collective action intentions . --- Experimental Effects A series of 2 x 2 between-participant Analyses of Variance were conducted to assess the impact of the two factors on each measure. Table 2 presents the tests of omnibus main effects . Full details of the interaction effects-including omnibus and follow-up tests-are reported in the text below, as they provide direct tests of Hypotheses 1 and 2. Significant two-way interactions were followed up by testing the simple effects of the leadership composition manipulation for each participant gender group; figures displaying the results are included in the Appendix . --- Collective Action Intentions . There was a significant main effect of the leadership composition manipulation, and no significant main effect of participant gender group . These effects were qualified by a significant two-way interaction, F = 9.30, p = .003, hp 2 = .046. Simple effects analyses indicated that female participants in the majority female leaders condition reported significantly stronger collective action intentions than did the female participants in the majority male leaders condition , F = 13.13, p < .001, hp 2 = .063. Among male participants, there was no effect of the leadership composition manipulation: there was no evidence of a difference in collective action intentions between men in the majority female leaders condition and men in the majority male leaders condition , F = 0.40, p = .531, hp 2 = .002. --- Perceptions of Leaders . Perceived Awareness of Inequality. There was a significant main effect of the leadership composition manipulation, and no significant main effect of participant gender group . These effects were qualified by a significant two-way interaction, F = 24.31, p < .001, hp 2 = .111. Simple effects analyses indicated that female participants in the majority female leaders condition perceived these female leaders to be significantly more aware of gender inequality than did the female participants who evaluated the male leaders in the majority male leaders condition , F = 130.17, p < .001, hp 2 = .402. Among male participants, the same pattern of results emerged though with a smaller effect size: men in the majority female leaders condition perceived the female leaders to have significantly more awareness of gender inequality than did men who evaluated the male leaders in the majority male leaders condition , F = 22.75, p < .001, hp 2 = .105. --- Perceived Alienation of Women. There was a significant main effect of the leadership composition manipulation, and no significant main effect of participant gender group . These effects were qualified by a significant two-way interaction, F = 9.34, p = .003, hp 2 = .046. Simple effects analyses indicated that female participants in the majority female leaders condition perceived significantly less alienation of women , than did the female participants in the majority male leaders condition , F = 102.70, p < .001, hp 2 = .346. Among male participants, the same pattern of results emerged with a smaller effect size: male participants in the majority female leaders condition perceived significantly less alienation of women than did male participants in the majority male leaders condition , F = 37.81, p < .001, hp 2 = .163. --- Perceived Ulterior Motive. There was a significant main effect of the leadership composition manipulation, and no significant main effect of participant gender group . These effects were qualified by a significant two-way interaction, F = 24.48, p < .001, hp 2 = .112. Simple effects analyses indicated that female participants perceived less ulterior motive among the female leaders in the majority female leaders condition than did the female participants who evaluated the male leaders in the majority male leaders condition , F = 35.78, p < .001, hp 2 = .156. Among male participants, there was no significant effect of the leadership composition manipulation: there was no evidence of a difference between men in the majority female leaders condition and men in the majority male leaders condition in their perception of leaders' ulterior motive F = 0.83, p = .363, hp 2 = .004. --- Emotions . Hope. There was a significant main effect of the leadership composition manipulation, but no significant effect of participant gender group . These effects were qualified by a significant two-way interaction, F = 17.611, p < .001, hp 2 = .083. Simple effects analyses indicated that female participants responding to the female leaders in the majority female leaders condition reported significantly more hope than did the female participants who responded to the male leaders in the majority male leaders condition , F = 28.84, p < .001, hp 2 = .129. Among male participants, there was no significant effect of the leadership composition manipulation: there was no evidence of a difference between men in the majority female leaders condition and men in the majority male leaders condition in their levels of hope, F = 0.22, p = .643, hp 2 = .001. Anger. There was no significant main effect of the leadership composition manipulation or participant gender group . The two-way interaction effect was significant, F = 6.40, p = .012, hp 2 = .032. Simple effects analyses indicated that female participants responding to the female leaders in the majority female leaders condition reported significantly less anger than did the female participants who responded to the male leaders in the majority male leaders condition , F = 9.71, p = .002, hp 2 = .048. Among male participants, there was no effect of the leadership composition manipulation: there was no evidence of a difference in reported levels of anger between men in the majority female leaders condition and men in the majority male leaders condition , F = 0.16, p = .686, hp 2 = .001. --- Mediation Analyses Moderated Mediation Model. To investigate the process underpinning lower collective action intentions in response to a majority male leadership team, we designed a custom mediation model for the PROCESS macro in SPSS using a model builder tool . The model specified the leadership composition manipulation as the exogenous predictor variable in a serial mediation model with collective action intentions as the final outcome variable. The three measures assessing perceptions of leaders were specified as the first set of parallel mediators, and the two emotions as the next set of parallel mediators. Participant gender was specified to moderate the relationships between the leadership composition manipulation and the first set of mediators . This model enabled an exploratory test of the proposed mediation pathway against the alternative mediator variables. We tested the model using estimates based on 10,000 bootstrap samples . Participant gender moderated the direct effect of the leadership composition manipulation on each perception of the leaders: perceived awareness of inequality , perceived ulterior motive , and perceived alienation of women . To decompose these interaction effects, we tested the custom serial mediation model separately among male and female participants using estimates based on 10,000 bootstrap samples . This strategy enabled us to examine the indirect effects of the leadership composition manipulation on the collective action outcome measure via the various mediators separately for each gender group. Mediation Model among Female Participants. The total effect of the leadership composition manipulation on collective action intentions was significant: lower 95% CI = -1.7815, upper 95% CI = -0.5314. Figure 2 presents the direct effects tested in this model. The leadership composition manipulation had a significant effect on each of the three leader perception variables. However, the manipulation did not have a significant effect on the emotions or collective action intentions once the three perception variables were also included as predictors. Only perceived awareness of inequality was associated with hope, and only perceived alienation of women was associated with anger. Collective action intentions were significantly associated with only one variable: hope. The indirect effects for all pathways from the leadership composition manipulation to the collective action outcome variable are presented in Table 3, All indirect effects from the leadership composition manipulation to the collective action outcome variable are presented in Table 3, with results for male participants in the bottom row for each indirect effect. Two of the 11 indirect effects were significant: Path 1, which tested the effect of the manipulation on collective action via perceived awareness of inequality, and Path 6, which tested the effect of the manipulation on collective action intentions via perceived awareness of inequality and hope. Alternative Model. Our design measured all variables at a single time-point and thus does not allow us to make strong claims about the causal order specified in the mediation model. However, the fact that hope was the only consistent direct predictor of collective action intentions gives us some confidence in our interpretation of the findings. Our model is also consistent with appraisal theories of emotion and previous research on the role of groupbased emotions in predicting collective action intentions. While comparing our model with alternatives in which the order of variables is altered does not provide strong evidence that one order should be preferred , we report the test of a plausible alternative model in the Supplemental Materials. --- Discussion Study 1 provides a first empirical demonstration of the intergroup dynamics involved in observers' responses to high-status group leaders in a social justice organization that seeks to benefit the low-status group. The presence of influential male leaders, compared to influential female leaders, reduced collective action intentions among female participants . In contrast, there was no evidence that the presence of influential male leaders lowered collective action intentions among male participants . These results provide support for Hypothesis 1, and the social identity analysis of the role of group membership in shaping observers' willingness to collaborate with a social justice organization. Results also support Hypothesis 2, whereby participants' group membership moderated their responses on the mediator variables. Female participants responded more negatively to the male leaders on a majority male leadership team than they did to the female leaders on a majority female leadership team, with respect to all three perceptions of the leaders and both emotions . Male participants, in contrast, evaluated influential high-status group leaders negatively only with respect to two mediator variables: they perceived the male leaders in the majority male leadership team as being less aware of inequality and more likely to alienate women, compared to the female leaders on the majority female leadership team. There was no evidence that influential male leaders of a gender equality organization evoked more negative emotional responses among male observers. Consistent with Hypothesis 3, exploratory mediation analyses showed that lower collective action intentions to work with a gender equality organization led by a majority male leadership team were mediated by a specific perceived problem and emotional response . Furthermore, there was no evidence that the alternative mediator variables that were tested in the model-perceived alienation of women, perceived ulterior motive, and angermediated participants' collective action intentions. The nature of the indirect effect was moderated by participant gender. Only one indirect effect was significant among female participants . In contrast, two indirect effects were significant among male participants . The stronger impact of the emotional response among members of the low-status group suggests that they perceived the information about the organization as more personally relevant to them . Study 1 thus provided evidence for the hypothesized pattern of experimental effects and the process underpinning collective action intentions. To replicate these results, we conducted a second study in a different intergroup context: racial inequality in the United Kingdom. --- Study 2 White and ethnic minority participants in the United Kingdom were presented with information about a non-profit organization seeking to achieve racial equality, whose leadership team was manipulated as being either majority White or majority racial/ethnic minority. We tested Hypotheses 1-3 in order to replicate results from Study 1. The hypothesized mediation model was supported in Study 1, and thus we now differentiate between hypothesized mediators and alternative mediators . To conserve space when reporting the experimental effects, we focus on the outcome variable and the hypothesized mediators in the main text; results for the alternative mediators are reported in the Supplemental Materials. Our test of the mediation model includes the hypothesized mediators as well as the alternative mediators, so as to assess the independent effects of the hypothesized mediators. --- Method --- Design The study employed a 2 x 2 between-participants person-by-treatment quasi-experimental design. Participants were randomly allocated to one leadership composition condition, and reported their racial/ethnic group membership. We only retained participants who self-identified as either solely racial/ethnic minority 3 or solely White, to ensure that they clearly identified with either the low-status racial/ethnic group or the high-status racial/ethnic group. --- Procedure As in Study 1, participants read about a fictional racial equality organization that manipulated the racial composition of the leadership team. They then completed a manipulation check item and a set of measures assessing responses to the influential leaders. --- --- Materials The materials from Study 1 were adapted to describe a fictional organization-the Race and Equality Centre-that sought to combat racial inequality in the United Kingdom. The last sentence manipulated the racial group composition of the leadership team: "This leadership team currently includes [more White people than members of ethnic minority groups or more members of ethnic minority groups than White people]." --- Measures The questionnaire was identical to that used in Study 1, except for minor modifications to fit the context of racial inequality in the United Kingdom. The measures reported below were created using 32 out of the 95 items in the full questionnaire, based on the rationale reported in the first study. Further information can be obtained from the first author. Manipulation Check. A single item asked participants to recall the racial composition of the leadership team presented in the manipulation: "The leadership team of the Race and Equality Centre currently includes: more White people than members of ethnic minority groups; equal numbers of White people and members of ethnic minority groups; more members of ethnic minority groups than White people. Participants could choose only one response option. As noted in the Participants section, 18 participants provided responses to this question that did not match the experimental condition to which they had been randomly assigned; they were thus excluded from the sample. --- Outcome Variable. The five items assessing Collective Action Intentions in Study 1 were adapted to assess the extent to which participants wanted to take action to help the Race and Equality Centre . Hypothesized Mediators. Items from Study 1 were adapted to create the four-item measure of leaders' Perceived Awareness of Racial Inequality and the nine-item measure of Hope . Following the format of the first study, instructions reminded participants which racial/ethnic group the majority of the leaders belonged to, and asked them to focus on the leaders in the numerical majority group when responding to the items. --- Alternate Mediators. The Study 1 items were adapted to create the four-item measure assessing Perceived Alienation of Racial/Ethnic Minorities by the leaders, the threeitem measure of Perceived Ulterior Motive of the leaders, and the six-item measure of Anger . As in Study 1, instructions reminded participants which racial/ethnic group the majority of the leaders belonged to, and asked them to focus on these leaders. --- Results --- Bivariate Correlations All six variables were significantly associated with each other and the pattern of results was generally consistent with Study 1. --- Experimental Effects A series of 2 x 2 between-participant Analyses of Variance were conducted to assess the impact of the two factors on each measure. Significant two-way interactions were followed up by testing the simple effects of the leadership composition manipulation for each participant racial/ethnic group. Below we report the results for the main outcome variable and the two hypothesized mediators; figures displaying these results are included in the Appendix . Results for the three alternative mediator variables are reported in the Supplemental Materials. Collective Action Intentions. The main effect of the leadership composition manipulation was not significant, F = 3.44, p = .065, hp 2 = .016: there was no evidence of a difference in collective action intentions between participants in the majority racial/ethnic minority leaders condition and participants in the majority White leaders condition . Furthermore, there was no significant effect of participant race/ethnicity, F = 3.40, p = .067, hp 2 = .016, indicating no evidence of a difference between racial/ethnic minority participants and White participants in collective action intentions. The two-way interaction was not significant, F = 2.88, p = .091, hp 2 = .013. However, this omnibus test is relatively insensitive, particularly when one simple effect is non-significant -as we had expected to be the case for members of the high-status group . Given that we did find distinct simple effects among men and women in Study 1, we sought to replicate this specific set of results . Thus, we proceeded to test the simple effects of the leadership composition manipulation for each participant racial/ethnic group . Among racial/ethnic minority participants, those in the majority racial/ethnic minority leaders condition reported significantly stronger collective action intentions than did those in the majority White leaders condition , F = 6.26, p = .013, hp 2 = .029. Among White participants, the effect of the leadership composition manipulation was not significant, F = 0.12, p = .911, hp 2 < .001. There was no evidence of a difference in collective action intentions between White participants in the majority racial/ethnic minority leaders condition and White participants in the majority White leaders condition . --- Perceived Awareness of Inequality. There was a significant main effect of the leadership composition manipulation, F = 146.81, p < .001, hp 2 = .409: participants in the majority racial/ethnic minority leaders condition perceived these racial/ethnic minority leaders to be significantly more aware of racial inequality than did participants who evaluated the White leaders in the majority White leaders condition . The main effect of participant race/ethnicity was not significant, F = 0.86, p = .356, hp 2 = .004. Thus, there was no evidence of a difference between racial/ethnic minority participants and White participants in their perception of leaders' awareness of inequality. These effects were qualified by a significant two-way interaction, F = 9.87, p = .002, hp 2 = .044. Simple effects analyses indicated that racial/ethnic minority participants in the majority racial/ethnic minority leaders condition perceived the racial/ethnic minority leaders to be significantly more aware of racial inequality than did racial/ethnic minority participants who evaluated the White leaders in the majority White leaders condition , F = 115.35, p < .001, hp 2 = .352. Among White participants, the same pattern of results emerged though with a smaller effect size: those in the majority racial/ethnic minority leaders condition perceived the racial/ethnic minority leaders to be significantly more aware of racial inequality than did those who evaluated the White leaders in the majority White leaders condition , F = 40.65, p < .001, hp 2 = .161. Hope. There was a significant main effect of the leadership composition manipulation, F = 27.20, p < .001, hp 2 = .114: participants in the majority racial/ethnic minority leaders condition reported significantly more hope than did participants in the majority White leaders condition . The effect of participant race/ethnicity was not significant, F = 0.37, p = .542, hp 2 = .002. Thus, there was no evidence of a difference in reported levels of hope between racial/ethnic minority participants and White participants . These effects were qualified by a significant two-way interaction, F = 9.37, p = .002, hp 2 = .042. Simple effects analyses indicated that racial/ethnic minority participants in the majority racial/ethnic minority leaders condition felt significantly more hope than did racial/ethnic minority participants in the majority White leaders condition , F = 33.93, p < .001, hp 2 = .138. Among White participants, there was no significant effect of the leadership composition manipulation, F = 2.34, p = .127, hp 2 = .011: there was no evidence of a difference in levels of hope between White participants in the majority racial/ethnic minority leaders condition and White participants in the majority White leaders condition . --- Mediation Analysis Moderated Mediation Model. We investigated the process underpinning lower collective action intentions in response to a majority White leadership team with the serial moderated mediation model developed in Study 1. We aimed to test the hypothesized mediation pathway in the context of the alternative mediator variables . We also assessed whether participant race/ethnicity moderated the relationship between the leadership composition manipulation and the first set of mediators . We tested the model using estimates based on 10,000 bootstrap samples . Participant race/ethnicity moderated the direct effect of the leadership composition manipulation on perceived leaders' awareness of inequality . There was no evidence that the effect of the As in Study 1, we next tested the custom serial mediation model separately among racial/ethnic minority group participants and White participants using estimates based on 10,000 bootstrap samples . This approach enabled us to examine the indirect effects of the leadership composition manipulation on the collective action outcome measure via the various mediators separately for each racial/ethnic group. Mediation Model among Racial/Ethnic Minority Participants. The total effect of the leadership composition manipulation on collective action intentions was significant: upper 95% CI = -1.1619, lower 95% CI = -0.1011. Figure 4 presents the direct effects tested in the model. The leadership composition manipulation had a significant effect on all three perceptions of the leaders, but not on emotions or collective action intentions once the other predictors were also included. Only perceived leaders' awareness of inequality was associated with hope and with anger. Hope was the sole variable that was significantly associated with collective action intentions. The indirect effects for all pathways from the leadership composition manipulation to the collective action outcome variable are presented in Table 5, with results for racial/ethnic minority participants presented in the top row for each indirect effect. As in Study 1, only one indirect effect was significant: Path 6, which tested the effect of the manipulation on collective action intentions via perceived awareness of inequality and hope. Mediation Model among White Participants. The total effect of the leadership composition manipulation on the collective action intentions was not significant . Figure 5 presents the direct effects tested in this model. The leadership composition manipulation had a significant effect on perceived awareness of inequality, perceived alienation of racial/ethnic minority groups, and hope. Perceived awareness of inequality was associated with hope, and two variables-hope and perceived ulterior motive-were associated with collective action intentions. Indirect effects from the leadership composition manipulation to collective action intentions are presented in Table 5, with results for White participants presented in the bottom row for each effect. Path 6 was the only significant indirect effect, and included the hypothesized mediator variables: perceived awareness of inequality and hope. Alternative Model. Our serial mediation model was based in theories of emotion and collective action, and replicated results from Study 1. Nevertheless, we acknowledge that the cross-sectional design did not allow for strong claims regarding causal order. We report a test of a plausible alternative model in the Supplemental Materials. --- Discussion We replicated the findings from Study 1 in a different intergroup context: racial inequality in the United Kingdom. First, there was qualified support for the prediction that observers' status group membership moderated their collective action intentions in response to influential high-status group leaders, compared to influential low-status group leaders . Although the omnibus interaction test did not meet the conventional level of statistical significance, the tests of simple effects were consistent with the prediction : Participants from racial/ethnic minority groups reported lower levels of collective action intentions in the presence of influential high-status group leaders, compared to the presence of influential low-status group leaders. In contrast, there was no evidence that White participants varied their collective action intentions based on the group membership of the influential leaders. Tests of experimental effects on the hypothesized mediator variables provided support for Hypothesis 2. Racial/ethnic minority participants perceived influential White leaders to be less aware of racial inequality compared to influential racial/ethnic minority leaders, and reported less hope in response to the White leaders compared to the racial/ethnic minority leaders. White participants did perceive influential White leaders as being less aware of racial inequality compared to influential racial/ethnic minority leaders, but there was no evidence that they differentiated between the types of leaders in their emotional response. Lastly, mediation analyses provided evidence for Hypothesis 3 and corroborated the Study 1 finding that lower levels of collective action in response to influential high-status group leaders were mediated by a specific perceived problem and emotional response . This hypothesized pathway was an important predictor of collective action intentions even after accounting for the effects of the alternative mediators in the model. In contrast to Study 1, the significant indirect effect was consistent for all observers regardless of their status group membership. However, both studies did demonstrate that the hypothesized mediation pathway was significant for both high-status groups and low-status groups. While there is a clear theoretical rationale for this hypothesized model, the results offer limited evidence for the specified causal pathway in our serial mediation model, because all variables were measured at a single time-point . Thus, we conducted an experimental study to provide more conclusive evidence for the specified causal pathway. --- Study 3 Mediation analyses in the first two studies indicated that a specific perception of leaders and emotional response mediated low-status group members' lower levels of collective action intentions in response to influential highstatus group leaders. Study 3 provided a stronger test of this hypothesized model by manipulating the mediator variable that was directly affected by the leadership composition manipulation: leaders' awareness of inequality. The present study was set in the domain of gender inequality, and recruited women as members of the low-status gender group. We presented participants with information about a gender equality organization, which included two independent manipulations: leaders' awareness of gender inequality and gender group composition of the leadership team . Participants then completed measures of hope and collective action intentions. All aspects of this study-including the design, hypotheses, sampling procedure, materials, measures, and data analytic strategy-were pre-registered before the start of data collection . Below we note any deviations from the pre-registration plan based on feedback obtained during the peer review process. All pre-registered measures and analyses that have been omitted from the text are reported in the Supplemental Materials. --- Method --- Design and Procedure The study employed a 2 x 2 between-participants design. The control condition in the awareness of inequality factor specifically did not mention awareness, because a statement of low knowledge would not be plausible in an ostensibly public description of an organization's leaders. Participants were randomly allocated to one of four conditions. The procedure was identical to that of the previous studies: Participants read a description of a fictional organization that included the manipulations, and then completed a series of measures. --- --- Materials Participants read about a fictional gender equality organization, which included the manipulations of the two independent variables. The first paragraph presented the focus and aims of the organization: "RBH is a British not-for-profit organisation that focuses on gender equality. RBH develops policies and programmes to improve opportunities for women." The second paragraph described the role of the leadership team in the organization, with the final sentence manipulating the gender composition of the team: The organisation is run by an Executive Leadership Team that is responsible for identifying strategic priorities, developing concrete goals and key performance indicators, and evaluating the success of these campaigns. This leadership team current includes [more men than women or more women than men]. The final paragraph manipulated the leaders' awareness of gender inequality. Participants in the high awareness condition read the following text: The leadership team brings extensive experience to the organisation: All members have worked on issues relevant to gender equality for more than 10 years. Every leader also undertakes ongoing training and independent study to further develop their knowledge of subtle gender bias in contemporary society. All members of the leadership team are thus well-equipped to understand the challenges facing women, and to develop appropriate policies and programmes to address these barriers. The no statement control condition did not mention the leaders' awareness of gender inequality. Participants in this condition only read the first two paragraphs. --- Measures The questionnaire included the final set of items from Studies 1 and 2 to assess the relevant variables. Thus, all items were used to create the composite measures described below. Manipulation Checks. As in Studies 1 and 2, participants were first asked to identify the composition of the leadership team, as a check of whether they had correctly understood the materials. A total of 84 participants incorrectly identified the gender composition of the leadership team they had read about; this included 51 participants in the majority female leaders condition and 33 participants in the majority male leaders condition. 4 These participants were omitted from the final sample . The four items assessing perceived awareness of inequality in Studies 1 and 2 were adapted for use as a check for the second manipulation . Following the format of the previous studies, instructions reminded participants which gender group the majority of the leaders belonged to, and asked them to focus on these leaders when responding to the items. --- Hope. Instructions reminded participants which gender group the majority of the leaders belonged to, and asked them to focus on these leaders when responding to the items. The items used in Studies 1 and 2 to assess Hope were again used here. The questionnaire also included items to assess anger, because we had planned to include it in the mediation model to test the independent effect of hope in predicting collective action intentions. In the interest of brevity, however, we have decided to focus solely on the variables in the hypothesized mediation model variables. Details regarding the anger variable are reported in the Supplemental Materials. Collective Action Intentions. As in Studies 1 and 2, five items were used to assess the extent to which participants wanted to work with the gender equality organization . --- Results --- Bivariate Correlations The three measured variables were correlated significantly with each other in the expected directions. Perceived leaders' awareness of inequality was positively associated with hope and collective action intentions (r [405] = 0.35, p < .001. Collective action intentions were positively associated with hope . --- Manipulation Check: Perceived Awareness of Inequality A 2 x 2 between-participant Analysis of Variance was conducted to assess the impact of the two independent variables on the perceived awareness of inequality manipulation check. The main effect of the awareness of inequality manipulation was significant, F = 12.76, p < .001, hp 2 = .03: The leaders presented in the high awareness condition were perceived as being significantly more aware of inequality than were the leaders in the control condition whose awareness of inequality was not explicitly mentioned . Furthermore, there was a significant main effect of the leadership composition manipulation, F = 339.00, p < .001, hp 2 = .46: Participants evaluated the female leaders in the majority female leaders condition as being significantly more aware of gender inequality than did participants who evaluated the male leaders in the majority male leaders condition . These effects were qualified by a significant two-way interaction, F = 5.14, p = .024, hp 2 = .09, which was followed up by testing the simple effects of the awareness of inequality factor at each level of the leadership composition factor 5 . When considering the majority male leadership teams, participants perceived the male leaders in the high awareness condition to be significantly more aware of gender inequality than did participants who evaluated the male leaders in the control condition , F = 18.01, p < .001, hp 2 = .04. In the majority female leaders condition, the simple effect was not significant, F = 0.81, p = .369, hp 2 < .01: there was no evidence of a difference in perceived female leaders' awareness of gender inequality, whether participants were in the high awareness condition or the control condition . The simple effects indicate that the awareness of inequality manipulation successfully addressed the perceived problem of male leaders' low awareness of inequality. This was the main comparison of interest, given our focus on explaining negative responses to influential male leaders in a gender equality organization. However, there was no evidence that the awareness manipulation influenced perceptions of the female leaders on the majority female leadership team. We return to this point in the Discussion section. --- Experimental Effects We conducted a 2 x 2 betweenparticipant Analysis of Variance that assessed the impact of the two independent variables on hope and collective action intentions. 6 Significant two-way interactions were followed up by testing the simple effects of the awareness of inequality factor at each level of the leadership composition factor. 5 Hope. The main effect of the awareness of inequality manipulation was significant, F = 12.31, p = .001, hp 2 = .03: Participants reported more hope in response to leaders with high awareness of inequality than they did in response to leaders whose awareness of inequality was not mentioned . Furthermore, there was a significant main effect of the leadership composition manipulation, F = 359.44, p < .001, hp 2 = .47: Participants reported more hope in the majority female leaders condition than did participants who in the majority male leaders condition . These effects were qualified by a significant two-way interaction, F = 10.00, p = .002, hp 2 = .02. Simple effects analyses indicated that, in the majority male leaders condition, participants reported significantly more hope when the leaders were described as having high awareness of inequality than did participants in the control condition that did not mention leaders' awareness of inequality , F = 23.52, p < .001, hp 2 = .06. In the majority female leaders condition, the simple effect was not significant, F = 0.06, p = .811, hp 2 < .01: there was no evidence of a difference in participants levels of hope, whether they considered female leaders in the high awareness condition or female leaders in the control condition that did not mention leaders' awareness . --- Collective Action Intentions. The main effect of the awareness of inequality manipulation was not significant, F = 0.403, p = .526, hp 2 < .01: there was no evidence that participants reported different levels of collective action intentions in response to leaders with high awareness of inequality or leaders whose awareness of inequality was not explicitly mentioned . There was a significant main effect of the leadership composition manipulation, F = 16.87, p < .001, hp 2 = .04: participants reported higher levels of collective action intentions in the majority female leaders condition than did participants in the majority male leaders condition . The interaction effect was not significant, F = 0.49, p = .482, hp 2 < .01. Thus, there was no evidence that the effect of the leadership composition manipulation was moderated by the extent to which leaders were described as being aware of inequality. --- Mediation Analysis Our hypothesized model posited that the indirect effect of the awareness of inequality manipulation on collective action intentions via hope would be significant . We first investigated whether there was evidence for moderated mediation, such that the leadership composition manipulation influenced the first direct effect in the mediation model, between the awareness of inequality manipulation and hope. This model was tested using estimates based on 10,000 bootstrap samples . The leadership composition manipulation moderated the direct effect of the awareness of inequality manipulation on hope . The indirect effect from the awareness of inequality manipulation to collective action intentions via hope was also moderated by the leadership composition manipulation . We thus tested the model separately in the two leadership composition conditions. The total effect of the leadership awareness manipulation on collective action intentions was not significant in either the majority male leaders condition or the majority female leaders condition , but was not significant in the majority female leaders condition . --- Discussion Low-status group members' negative responses to influential male leaders in a social justice effort can be improved by showcasing these leaders' high awareness of the inequality. The manipulation of leaders' awareness of inequality increased female participants' perception that male leaders on a majority male leadership team were aware of inequality, as well as participants' reported feelings of hope. Drawing attention to these male leaders' high awareness of inequality also indirectly increased participants' collective action intentions via increased levels of hope, which was the sole predictor of collective action intentions. Taken together, these results provide more conclusive evidence for the causal order specified among the mediators in the first two studies: perceived leaders' awareness of inequality à hope à collective action intentions. Results also indicated an important caveat to this conclusion: participants responded more negatively to influential male leaders than they did to influential female leaders, even when male leaders' increased awareness of inequality was highlighted. This suggests that observers perceive low-status group leaders to be especially well suited to leading a social justice effort that benefits their ingroup. Most likely this is because observers assume lowstatus group leaders to have lived experience of the disadvantage that is being tackled, which arguably gives these leaders the insight to develop the most effective goals and strategies to challenge the inequality. Indeed, the manipulation check results indicated that female leaders were perceived to be highly aware of gender inequality, even when the study materials did not explicitly mention this awareness. Furthermore, the awareness manipulation influenced participants' perceptions of male leaders only to a certain point: Perceptions of male leaders' awareness of inequality in the high awareness condition were only slightly above the midpoint of the response scale . We posit that this pattern of results showcases the intergroup dynamics that underpin any system of intergroup inequality: High-status groups and low-status groups occupy discrete structural positions within a status hierarchy that produce divergent material and psychological experiences . Observers are likely to take notice of leaders' particular perspectives in the context of a social justice organization, where the work itself makes the concept of group status salient. In such cases, a single paragraph that showcases leaders' awareness of inequality is not likely to substantially change observers' views about these leaders. For this reason, our manipulation of leaders' awareness of inequality should not be interpreted as an intervention that can elicit equally accepting views of highstatus group leaders and low-status group leaders in a social justice effort that benefits the low-status group. Rather, the present study sought to demonstrate that systematically varying leaders' awareness of inequality is sufficient to shift observers' reported feelings of hope and, indirectly, their willingness to contribute to the social justice effort. Taken together, the first three studies demonstrate that observers from low-status groups respond more negatively to influential leaders from the high-status group in a solidarity context where the social justice effort seeks to benefit the observers' low-status ingroup. As noted in the Introduction, alternative accounts could explain these results using the well-established effects of ingroup bias and role congruity. We propose that our effects extend beyond such preferences, because members of a high-status outgroup should be viewed as especially unsuitable to lead an organization that champions the interests of a low-status ingroup. To test this prediction, the final study compared low-status group observers' evaluations of influential high-status group leaders in various organizations that either serve to benefit an ingroup or an outgroup with varying degrees of role congruity. --- Study 4 This study recruited individuals who belong to the low-status group in the context of gender inequality: women. Participants read a description of a non-profit organization that included two manipulations. First, the gender composition of the leadership team was presented as either majority male or majority female . Second, we manipulated the group that would benefit from the work done by the organization, in order to vary the extent to which the high-status group leaders' presence could be interpreted as solidarity. One organization worked in the context of gender equality to improve opportunities for women ; thus, the male leaders' presence in this organization could be interpreted as solidarity. The other two conditions were non-solidarity controls, as they described organizations that aimed to benefit outgroups. One organization worked in the context of education to improve students learning outcomes. The other organization worked in the context of engineering to improve construction workers safety. The two contexts that benefited outgroups were chosen because they represent distinctly gendered domains. Education is a female-dominated occupation in which workers are perceived as stereotypically feminine . Education is also characterized as a stereotypically female occupation in research across the social sciences, including organizational behavior and political science . In contrast, engineering/construction is a male-dominated occupation in which workers are perceived as stereotypically masculine . As such, male leaders' role congruity should be low in the education domain and the high in the engineering domain. We report a pilot study to demonstrate that an independent group of participants perceived these sectors in stereotypically gendered ways. The questionnaire in the main study included all six measured variables reported in Studies 1 and 2, so that we could assess the full range of observers' responses to leaders. Since the previous studies demonstrated the key role of the hypothesized mediators , we focus solely on these variables in the main text. Information about the alternative mediators is provided in the Supplemental Materials. In our tests of experimental effects, we expect to find a significant 2-way interaction between the leadership composition manipulation and the beneficiary group manipulation, which will be followed up by examining the simple effects of the beneficiary group factor at each level of the leadership composition factor. To demonstrate that our results cannot be explained solely by ingroup bias , we expect to show that participants' responses to influential male leaders are significantly more negative in the organization that benefits women , compared to the organizations that benefit students or construction workers . Furthermore, we expect to find a significant difference between the solidarity context and both non-solidarity contexts, regardless of the broader role congruity implied for high-status group leaders in each domain ; this result would demonstrate that the role congruity explanation alone cannot account for our effects. Lastly, we examine whether lower collective action intentions in response to influential male leaders in the solidarity context can be explained by perceived leaders' awareness of the problem and reported levels of hope-the hypothesized process documented in the previous studies. --- Method --- Design The study employed a 2 x 3 between-participants design. All participants were randomly assigned to one of six experimental conditions. --- Procedure As in previous studies, participants read a description of a fictional organization that included the manipulations. They then completed two manipulation check items and a set of measures that assessed their responses to the leaders. --- --- Materials and Manipulations Participants read a brief description of a non-profit organization. The first two sentences manipulated the beneficiary group, with the following text in each condition: Female beneficiaries condition: "RBH is a British not-for-profit organisation that focuses on gender equality. RBH develops policies and programmes to improve opportunities for women." Student beneficiaries condition: "RBH is a British not-for-profit organisation that focuses on education. RBH develops policies and programmes to improve students' learning outcomes." Construction worker beneficiaries condition: "RBH is a British not-for-profit organisation that focuses on engineering. RBH develops policies and programmes to improve workers' safety in the construction industry." Participants then read about the leadership team in the organization: "The organisation is run by an Executive Leadership Team that is responsible for identifying strategic priorities, developing concrete goals and key performance indicators, and managing staff." The final sentence manipulated the gender composition of the leadership team: "This leadership team currently includes more [more women than men or more men than women]." Pilot Study: Perceptions of Occupational Sectors. We sought to demonstrate that our descriptions of the education and engineering organizations accurately represented stereotypically female and male occupational sectors, respectively. Thus, we conducted a pilot study in which an independent sample of participants provided their views of each sector, with respect to the gender composition of the workforce, the gender stereotypes associated with the sector, and the suitability of men and women to work in the sector. Full details of the method and results are reported in the Supplemental Materials. Across all variables, participants perceived engineering to be a more stereotypically masculine occupational sector, compared to education and gender equality. The gender equality and education sectors were perceived to be equivalently stereotypically feminine, with respect to the gender composition of their workforces, the gender stereotypes associated with the sector, and the suitability of women to work in the sector. --- Measures The questionnaire included only the final set of items used in Studies 1 and 2 to assess the composite variables. Information about the three alternative mediator variables is provided in the Supplemental Materials. Manipulation Checks. A single item asked participants to recall the beneficiary group they read about: women, students, or workers in the construction industry. Participants could choose only one response option. Twenty-seven participants provided an incorrect response , and thus were dropped from the sample. As in Studies 1 and 3, a single item asked participants to recall the gender composition of the leadership team presented in the manipulation: more men than women, equal numbers of men and women, or more women than men. Eighteen participants provided an incorrect response to this question and thus were dropped from the sample. Collective Action Intentions. The five items from Studies 1 and 2 were used to assess the extent to which participants wanted to work with the organization . Perceived Awareness of the Problem. Following the format used in the previous studies, instructions reminded participants which organization they had read about and which gender group the majority of the leaders belonged to. Participants were asked to focus on the leaders who made up the numerical majority on the leadership team. The four items assessing perceived awareness of inequality were adapted to measure Perceived Awareness of the Problem , in order to include the range of issues being addressed by the organizations in this study. Example items include, "The leaders truly understand the problems that their organisation is seeking to solve," and "The leadership team will be able to speak for those they are trying to help." Hope. Instructions reminded participants which organization they had read about and which gender group the majority of the leaders belonged to. Participants were asked to focus on these leaders when indicating how much they felt each emotion. The items used in Studies 1 and 2 to measure Hope were used here. --- Results --- Bivariate Correlations The three focal variables were significantly associated with each other in the expected directions. Collective action intentions were positively associated with hope and perceived leaders' awareness of the problem . Perceived leaders' awareness of the problem was positively correlated with hope . --- Experimental Effects A series of 2 x 3 between-participants Analyses of Variance assessed the impact of the two factors on each measure. Significant two-way interactions were followed up by testing the simple effects of the beneficiary group factor at each level of the leadership composition factor. Table 6 presents the tests of the main effects; the interactions are reported in the text as they provide direct tests of Hypotheses 5 and 6. Collective Action Intentions. There was a significant main effect of the leadership composition manipulation and the beneficiary group manipulation . These effects were qualified by a significant two-way interaction, F = 6.29, p = .002, hp 2 = .022. Simple effects analyses indicated that in the majority male leaders condition, the simple effect of the beneficiary group factor was significant (F [2, 549] = 12.79, p < .001, hp 2 = .05; we thus conducted follow-up simple comparisons. Participants reported lower levels of collective action intentions to work with the organization seeking to benefit women , compared to the organization seeking to benefit students , p < .001, and compared to the organization seeking to benefit construction workers , p = .003. Reported levels of collective action intentions were also higher for the organization seeking to benefit students than for the organization seeking to benefit construction workers, p = .046. The presence of influential male leaders thus elicited lower levels of collective action intentions among women in the solidarity context than in both non-solidarity contexts . In the majority female leaders condition, the simple effect of the beneficiary group factor was not significant, F = 1.52, p = .219, hp 2 = .006. Thus, there was no evidence of a difference in participants' reported levels of collective action intentions to work with organizations led by majority-female leadership teams, regardless of which group would benefit: women , students , or construction workers . Perceived Awareness of the Problem. The main effects of the leadership composition manipulation and the beneficiary group manipulation were significant , and were qualified by a significant two-way interaction, F = 46.11, p < .001, hp 2 = .144. Simple effects analyses indicated that in the majority male leaders condition, there was a significant simple effect of the beneficiary group factor , which was followed up with simple comparisons. Influential male leaders were perceived to be significantly less aware of the problem in the organization seeking to benefit women compared to the organization seeking to benefit students , p < .001, and compared to the organization seeking to benefit construction workers , p < .001. Influential male leaders were also perceived to be significantly less aware of the problem in the organization seeking to benefit students than in the organization seeking to benefit construction workers, p < .001. Thus, male leaders on majority male leadership teams were perceived to be less aware of the problem in the solidarity context than in both of the nonsolidarity contexts . In the majority female leaders condition, the simple effect of the beneficiary group factor was also significant . Simple comparisons indicated that influential female leaders were seen as significantly more aware of the problem in the organization seeking to benefit women compared to the organization seeking to benefit students , p < .001, and compared to the organization seeking to benefit construction workers , p < .001. There was no difference in influential female leaders' perceived awareness of the problem between the organization seeking to benefit students and the organization seeking to benefit construction workers, p = .563. Hope. There was a significant main effect of the leadership composition manipulation, and the main effect of the beneficiary group manipulation was not significant . The two-way interaction was significant, F = 4.54, p =.011, hp 2 = .02. Simple effects analyses indicated that in the majority male leaders condition, there was a significant simple effect of the beneficiary group factor , which was followed up with simple comparisons. Participants reported feeling significantly less hopeful in response to influential male leaders in the organization seeking to benefit women compared to influential male leaders in the organization seeking to benefit students , p = .012, and compared to influential male leaders in the organization seeking to benefit construction workers , p = .009. There was no evidence of a difference in levels of reported hope about male leaders in the student beneficiaries condition compared to the construction workers beneficiaries condition, p = .907. Thus, influential male leaders elicited less hope from female observers in the solidarity context than in both the non-solidarity contexts . In the majority female leaders condition, the simple effect of the beneficiary group factor was not significant, F = 1.47, p = .231, hp 2 = .005. Thus, there was no evidence of a difference in participants' reported hope about influential female leaders, regardless of the beneficiary group: women , students , or construction workers . --- Mediation Analysis We investigated the mediators of female observers' lower collective action intentions in response to influential male leaders in the solidarity context compared to influential male leaders in non-solidarity contexts . We expected that the process demonstrated in the previous studies-perceived leaders' awareness of the problem and felt hope-would again account for more negative responses to high-status group leaders in the solidarity context. The serial mediation model specified collective action intentions as the final outcome variable, and the beneficiary group manipulation as the exogenous predictor variable, which compared the ingroup beneficiary condition to a newly created outgroup beneficiary condition that included the other two beneficiary groups . Perceived leaders' awareness of the problem and hope were specified as the mediators as in the previous studies. We tested the model using estimates based on 10,000 bootstrap samples . The total effect of the beneficiary group manipulation on collective action intentions was significant: lower 95% CI = -0.9806, upper 95% CI = -0.4284. Figure 7 presents the direct effects tested in the model. The beneficiary group manipulation had a significant effect on perceived leaders' awareness of the problem and collective action intentions: participants perceived lower leader awareness of the problem and reported lower collective action intentions in response to the male leaders in the organization benefitting the ingroup, compared to the male leaders in the organizations benefitting outgroups. Perceived awareness of the problem was associated with hope, which was associated with increased collective action intentions. The only significant indirect effect from the beneficiary group manipulation to collective action intentions included perceived awareness of the problem and hope: b = -0.09, SE = 0.03, lower 95% CI = -0.1578, upper 95% CI = -0.0353. --- Discussion This study sought to demonstrate that low-status group members' negative responses to influential high-status group leaders in a solidarity context represent a unique phenomenon that goes beyond standard accounts of group bias and role congruity. With respect to collective action intentions, perceived leaders' awareness of the problem, and feelings of hope, female observers reported more negative responses to influential male leaders in a social justice organization that benefited their gender group , compared to the influential male leaders in social justice organizations that benefited the clear outgroups of students or construction workers . Consistent with Hypothesis 5, then, ingroup bias is not sufficient to account for the especially pronounced negative responses to high-status group leaders in solidarity contexts. The difference in responses between solidarity and non-solidarity contexts held up regardless of the broader role congruity implied for high-status group leaders in each domain, in line with Hypothesis 6. Observers' responses to influential male leaders were more negative in the solidarity context compared to the non-solidarity context in which men are perceived to be a bad fit . Lastly, participants lower levels of collective action in response to high-status group leaders in solidarity contexts were mediated by their perception of leaders in the solidarity context as being less aware of the problem, and lower levels of hope. This provides further support for the specific processes we posit are at work in when low-status group members evaluate high-status group leaders in solidarity contexts. --- General Discussion Social justice efforts led by members of high-status groups can reinforce a set of intergroup dynamics that discourages participation from the very groups they seek to empower and mobilize. In Studies 1 and 2, members of low-status groups reported more negative responses to influential high-status group leaders with respect to collective action intentions, hope, and perceived leaders' awareness of inequality. In contrast, there was no evidence that members of high-status groups differentiated between influential leaders from the high-status group or low-status group, with respect to collective action intentions or hope. 7 Mediation analyses demonstrated that lower levels of collective action intentions in response to influential high-status group leaders are mediated by the perception of a specific problem presented by these leaders and lower levels of hope. Study 3 provided experimental evidence for this causal order: A manipulation of high-status group leaders' awareness of inequality increased low-status group members' levels of hope, which in turn predicted higher levels of collective action intentions. Study 4 found that low-status group members' negative responses to high-status group leaders extend beyond the preferences predicted by frameworks of ingroup bias and role congruity. Participants from the low-status group showed more negative responses to highstatus group leaders in social justice efforts benefitting the low-status ingroup , compared to high-status group leaders in social justice efforts benefitting an outgroup , even when role congruity for these leaders was low. --- Theoretical Implications By investigating observers' responses to high-status group leaders' influential presence in social justice efforts, the present studies extend a small but growing literature on the consequences of collective action and social movements. Previous work has documented the impact of collective action on activists' attitudes and future collective action intentions , and on the success of efforts to recruit new sympathizers to the movement . Our findings offer the novel insight that observers' status group membership is likely to shape their responses to social movements . Observers' feelings of hope emerged as a consistent independent predictor of their collective action intentions, above and beyond their feelings of anger . Frameworks of collective action have typically focused on anger and resentment about injustice as a predictor of social change efforts . The present studies add to a growing body of work demonstrating that positive emotions can predict collective action . Our studies also uniquely focused on collaborative collective action to work with a social justice organization, rather than the more combative action to challenge the status quo that is typically featured in empirical studies. Such contributions facilitate the development of comprehensive frameworks of collective action that reflect the range of predictors and types of action reflected in social justice efforts. Our results also contribute to the development of a more nuanced analysis of group status and perceptions of leadership. In the context of standard organizations, research indicates that leaders from low-status groups are typically perceived more negatively than are leaders from high-status groups . Other work demonstrates that the direction of perceived congruence between group status and leadership is reversed in contexts where the perspective of low-status groups is valued or where there are shifts in views regarding role congruity . Our results add to this literature by demonstrating that high-status group leaders in social justice organizations are evaluated based on considerations beyond standard preferences predicted by group bias and general role congruity. --- Directions for Future Research The specific focus of our research question highlights various opportunities for future research. First, we intentionally examined responses to leaders of hierarchical non-profit organizations because we sought to understand the impact of high-status group members' influential presence in social justice efforts. Therefore we cannot determine whether the negative responses we documented towards high-status group leaders will generalize to highstatus group members working at lower levels of the organization. It is possible that the power wielded by leaders places a particularly harsh spotlight on high-status group members working at that level. In contrast, high-status group members working in lower-level positions may be met with less cynicism because their organizational role constrains the extent to which their presumed blind-spots can undermine the social justice effort. Future work should explore this possibility. Our manipulation of leadership composition in all four studies created situations in which members of either the low-status group or high-status group made up the numerical majority of leaders. This choice reflects the reality that social justice efforts are often led by members of high-status groups , as well as our theoretical interest in understanding the impact of influential high-status group leaders' presence in these contexts. Our findings can be extended by investigating responses to highstatus group members who are in the numerical minority, or even the sole representative of their group, on the leadership team of a social justice effort that benefits the low-status group. The manipulation of leadership composition was also very conspicuous, as it explicitly directed participants' attention to the leaders' status group membership. This strategy overtly made group status salient to participants, and thus may have heightened negative responses to high-status group leaders. Observers may report more balanced responses to high-status group leaders if information about these leaders' presence were presented in a subtle manner. Future work could investigate this possibility by using more naturalistic manipulations of leadership composition, such as a list of names or photos that indicate the leaders' status group membership without directly calling attention to this characteristic. All four studies focused on a single group identity when considering responses to the leaders of a social justice effort. While this strategy afforded us greater experimental control, it is plausible that observers will consider additional characteristics when responding to influential leaders . Future work might investigate the impact of intersecting identities across categories beyond the focus of the social justice organization and additional secondary features of the leaders . Using a more naturalistic manipulation of leadership composition would allow researchers to unobtrusively manipulate these additional characteristics, thus contributing to a more comprehensive analysis of observers' responses to leaders of a social justice effort. We found consistent evidence that observers from low-status groups report negative responses to high-status group leaders' influential presence in social justice efforts that benefit low-status groups, compared to influential low-status group leaders in the same organization and compared to influential high-status group leaders in social justice efforts that benefitted clear outgroups . This effect is likely qualified by individual differences, as suggested by related lines of research. For example, low-status groups' level of suspicion about high-status groups' motivations predicts their negative evaluations of these group members' actions . Similarly, low-status group members' tendency to attribute negative outcomes and experiences to prejudice is shaped by their levels of stigma consciousness and status-based rejection sensitivity . More broadly, responses to intergroup inequality are moderated by group identification and ideologies such as social dominance orientation . Future work should consider how such individual differences might shape low-status group members' responses to the presence of high-status group members in social justice efforts that benefit their low-status ingroup. Our interest in understanding observers' negative responses to high-status group leaders dictated our decision to assess three perceived problems with high-status group involvement in social justice efforts that benefit the low-status group . These specific problems were chosen based on theory and research investigating responses to inequality, the psychology of relative advantage, and intergroup prosocial behavior. To extend the present work, additional problems might also be investigated, such as the external legitimacy or credibility of an organization led by influential high-status group members. Furthermore, as noted in the Introduction, high-status group members' participation in social justice efforts can also have positive outcomes, such as providing access to resources and networks. It is thus important to investigate the extent to which observers from low-status groups acknowledge the positive contributions of high-status group leaders' solidarity . An examination of ambivalent responses-whereby observers note the problems created by high-status group leaders while also acknowledging their positive contributions-would also extend analyses beyond the standard positive/negative dichotomy that is typically used in the literature. Future work should explore these questions further. Lastly, the present studies were conducted within the contexts of left-wing or liberal social justice efforts, which focused on improving the status of historically disadvantaged groups along the dimensions of gender and race/ethnicity . Collective action and social movements can also advocate for right-wing or conservative goals, such as maintaining or defending the existing status quo. For instance, groups that have historically been considered to be advantaged in society can perceive themselves to be victimized or disenfranchised in the context of shifting cultural and/or political environments . While the specific goals of left-wing and right-wing social movements do vary, we speculate that similar intergroup dynamics would operate when members of the high-status out-group join or lead the efforts. However, this is an open question that deserves further study. --- Conclusions Social justice efforts led by influential members of high-status groups run the risk of discouraging participation by the low-status groups they seek to empower and mobilize. Leaders often serve as a visible reminder of an organization's goals and strategic direction, both to its employees and external stakeholders. Given that observers may know little about social justice organizations beyond visible cues such as the group membership of high-profile leaders , our findings underscore the significant real-world challenges facing social justice efforts led by high-status group members. Although our results demonstrate that showcasing high-status group leaders' awareness of inequality can begin to attenuate observers' negative responses, further work is needed in this area. When low-status group members are convinced that influential high-status group leaders play a valued role in social justice efforts, the talents and resources of people from all groups can be harnessed to help effectively challenge inequality and injustice. psychology of prosocial behavior: Group processes, intergroup relations, and helping . New York: Wiley-Blackwell. Iyer, A., & Ryan, M. K. . Why do men and women challenge gender discrimination in the workplace? The role of group status and in-group identification in predicting pathways to collective action. Journal of Social Issues, 65, 791-814. doi: 10.1111/j.1540-4560.2009.01625.x Iyer, A., Schmader, T., & Lickel, B. . Why individuals protest the perceived transgressions of their country: The role of anger, shame, and guilt. Personality and Social Psychology Bulletin, 33, 572-587. doi: 10.1177 . Relationships between intergroup contact and prejudice among minority and majority status groups. Psychological Science, 16, 951-957. doi: 10.1111/j.1467-9280.2005.01643.x Van Leeuwen, E. . Restoring identity through outgroup helping: Beliefs about international aid in response to the December 2004 tsunami. European Journal of Footnotes 1. Three items in this measure describe mutable characteristics that could be held by leaders of either the advantaged or disadvantaged group . However, one item could be seen to describe an immutable characteristic that cannot be held by leaders from the advantaged gender group. Results of the main analyses on this measure are virtually identical, whether this item is included or omitted. Thus, we decided to retain all four items. 2. The correlations between the three perceptions of leaders were moderate in size. To determine that they represent three distinct constructs, we conducted a Confirmatory Factor Analysis. Results indicated that a three-factor model fit the data well; full details are reported in the Supplemental Materials. 3. We acknowledge that racial/ethnic minority groups have distinct histories and experiences within mainstream society in the United Kingdom. However, these groups also share the common experience of belonging to a low-status racial/ethnic group, and thus contend with similar issues around power and status differentials in contact experiences with members of the high-status or majority group . Following this reasoning, we decided to combine participants from all racial/ethnic minority groups into a single category in order to assess their responses to influential leaders from the high-status racial/ethnic group . 5. The simple effects of the leadership composition factor at each level of the awareness of inequality factor are reported in the Supplemental Materials. 6. We presented a different analytic strategy in the pre-registration document: A oneway ANOVA in order to investigate how the awareness of inequality manipulation and leadership composition manipulation worked in combination to influence responses to the leaders. However, the feedback we received during the peer review process persuaded us that we should conduct a 2x2 ANOVA to be consistent with the study design. We report the results of the pre-registered analysis in the Supplemental Materials. 7. We conducted separate mini meta-analyses to examine the effect of the leadership composition manipulation and the effect of participants' group membership on each of the three focal variables: collective action intentions, perceived leaders' awareness of inequality, and hope. Meta-analysis of Cohen's d or mean effect size, was calculated using an inverse variance weighting approach with fixed effects . Across the four studies, the effect of the leadership composition manipulation was significant for all three variables . Compared to the majority high-status leaders condition, the majority low-status leaders condition elicited greater collective action intentions , greater perceived leaders' awareness of inequality , and higher levels of reported hope . The effect of participants' group membership was not significant for the three variables across the first two studies . Thus, there was no evidence of a difference between participants from the low-status group and participants from the high-status group in their reported levels of collective action intentions , perceived leaders' awareness of inequality , or hope . --- Table 1 Bivariate Correlations between all Measured Variables, Full Sample --- Table 2 2 x 2 between-participants ANOVA: Test Statistics,Effect Sizes,and Descriptive Statistics for Main Effects,Full Sample Note. 1 = 7 point response scale . 2 = 7 point response scale Note. Composition = Gender composition of leadership team , Ineq. = Inequality. Bolded text indicates a significant indirect effect . --- Majority male leaders Majority female leaders

This is a repository copy of Mobilized or marginalized? Understanding low-status groups' responses to social justice efforts led by high-status groups.

Introduction HIV pre-exposure prophylaxis is a way to prevent HIV infection using antiretroviral medications. The US Food and Drug Administration has approved 2 pharmaceutical products for use as PrEP: tenofovir disoproxil fumarate combined with emtricitabine and tenofovir alafenamide combined with emtricitabine . PrEP is taken by individuals who are HIV-negative and has been proven to be clinically effective among multiple target populations [1]. Widespread use has already been associated with a significant reduction in new infections in certain local epidemics [2,3]. However, to realize the promise of this biobehavioral HIV prevention strategy, those who are most at risk must opt to take the drugs, and it is well known that some key populations are underrepresented in terms of PrEP uptake in the United States [4,5]. Documented barriers, which may be particularly acute among such populations, include lack of awareness of PrEP, lack of access to or difficulty accessing care, fewer resources to pay for out-of-pocket expenses associated with PrEP , same-sex stigma and racism in medical contexts, and immigration status [6][7][8]. Telehealth is defined by the US Health Resources and Service Administration as "the use of electronic information and telecommunications technologies to support and promote long-distance clinical health care, patient and professional health-related education, public health and health administration" [9]. This modality of care is not new [10] but is drawing attention as one way to potentially circumvent barriers to PrEP access, especially in the era of COVID-19 when access to specialty care services and care in general is compromised [11]. A recent review [12] noted that PrEP telehealth interventions are being deployed to various ends. Some efforts seek to enhance access among patients who are geographically isolated or encounter challenges locating a knowledgeable provider; others facilitate clinical consultation and/or aim to build capacity among medical providers who may be willing to prescribe PrEP but desire support [5,[13][14][15][16]. These endeavors are largely the product of partnerships between government, academic, and community stakeholders, and early results suggest that telehealth-based PrEP is feasible and acceptable and may reduce some barriers to access [12], although detailed data on the experiences of patients are scarce. Private companies have also entered the PrEP telehealth market. One such company in San Francisco, California, called Nurx, offers web-and app-based counseling, prescription, and home-delivery services for PrEP. Nurx began to offer internet-based access to birth control in 2015, expanded to include PrEP a year later, and is now operating in 29 states and the District of Columbia. Qualitative researchers at the University of California, San Francisco partnered with Nurx to explore the possibilities raised by untethering PrEP from traditional brick-and-mortar medical contexts. The initial research objective is to understand how clients make sense of their experience, with special interest in what difference, if any, the web-based format makes. --- Methods This research employed a mixed methods design, conducting an electronic chart review and semistructured telephone interviews with clients who requested PrEP through Nurx. Sampling so-called PrEP requesters was intended to enable the examination of both the experience of completing the envisioned service cycle and the barriers or circumstances that might prompt a prospective Nurx user to initiate but desist from engagement with the service. Nurx personnel used the platform's asynchronous messaging system to contact all users who had requested PrEP in California. This outreach provided a brief, initial description of study procedures and purpose and requested that users interested in learning more respond with Yes. A user's response authorized UCSF interviewers to directly correspond with potential participants through the messaging system and via email . Participants received a study information sheet via email and asked questions about the research before arranging a telephone interview. Interviewers familiarized themselves with consenting interviewees' cases through chart review. Although we do not directly report data gained through this step in this paper, it was nonetheless important, as it sensitized us to issues that might surface during the interview, such as service gaps, challenges with billing, insurance, or delivery. In some cases, chart review-which included messages exchanged between the clients and the Nurx team-provided a way to triangulate interviewees' accounts. This general procedure was repeated for PrEP requesters in Florida, New York, and Illinois. The interview guide covered previous knowledge and/or use of PrEP; learning about Nurx; interacting with the web-based platform and Nurx personnel; and receiving, taking, and desisting from PrEP, as relevant. In addition, the guide explored interviewees' motivations for requesting PrEP and sexual practices. Interviews were conducted by KAK and SDH lasted for not more than 90 minutes and were digitally recorded with permission. The interviewees received a US $50 electronic gift card. Data were gathered from April to August 2017, with interviewers jointly debriefing as interviews were conducted. The Institutional Review Board at the UCSF approved all the research procedures. Interviews were transcribed verbatim, and transcripts were uploaded to MAXQDA 12 Plus [17], a qualitative data analysis software package, to facilitate analysis. Both interviewers participated in the coding process, with transcripts divided evenly between them for first-pass thematic coding. Both deductive and inductive approaches were used to identify salient themes [18]. Deductive codes were drawn from the interview guide and the literature . Inductive codes emerged from a close reading of the text itself . After the first coding pass, interviewers jointly refined code definitions, and each interviewer reviewed the coding done by her teammate, having the option to confirm the code applications or modify them in accordance with the revised codebook. Instances in which the revised codebook did not resolve coding discrepancies were examined, and consensus was achieved through discussion. This paper draws on thematic analysis of materials from 31 PrEP requesters in California, New York, Illinois, and Florida. Exemplary quotes in the sections that follow are attributed to interviewees by number to protect their confidentiality. --- Results --- Sample and Service Flow Interviewees were predominantly male, aged 30 years or younger, racially and ethnically diverse, highly educated, and identified as gay. Nearly all had insurance, whether public, private, or through parents . --- XSL • FO --- RenderX Owing to the novelty of web-based access to PrEP, it is helpful to sketch the Nurx service flow before engaging with the thematic analysis. To do this, we drew from interviewees' descriptions to create a simplified journey map of the Nurx client experience, as it was at the time of data collection [19]. In service design, a client journey is seen to have general phases: awareness, onboarding, receiving the service, maintaining the service, and finally, if necessary, finishing [20]. In the case of Nurx, clients had to become aware of both PrEP and Nurx. Onboarding included multiple steps. Users created a profile and then initiated a PrEP request by answering a web-based questionnaire about sexual behaviors and other HIV-related risks. Those with health insurance also typically provided this information by using a mobile phone to take a photo of their insurance card and uploading that photo to the Nurx site. Personnel at Nurx reviewed this information, and a clinician then sometimes initiated a discussion with the client about the motivation for the PrEP request before sending electronic orders for HIV and other tests to a laboratory in the client's area. Laboratories relayed results back to Nurx electronically, and these were reviewed by a Nurx physician. For clients with nonreactive HIV tests, the first step in the receiving the service phase was the transmission of a prescription for PrEP to a pharmacy. However, this happened on the back end, meaning it was invisible to clients. They received a message from the physician through the Nurx site that included the results of their laboratory tests, notification that they were being prescribed Truvada, dosing guidance, and other recommendations . The 3-month prescription was usually shipped directly from the pharmacy to a client-provided address via FedEx. At the time of data collection, the clients had to sign for the package. Once the medication was delivered, the client could begin taking PrEP. Maintaining the service started with an automated message from Nurx, sent to clients several weeks ahead of prescription renewal time, informing them of the need to submit a new PrEP request through the site. The new PrEP request then generated laboratory orders, and the cycle started again. Clients who did not submit new requests or failed to complete their laboratory tests typically received additional reminders. Prescriptions would be renewed if and when clients' renal function was within the normal range and the HIV test result was negative. Only 4 participants in this study had experience with finishing. Of 4 participants, 2 men had to stop using the service because of changes in insurance coverage. Another participant stopped using PrEP entirely. The fourth participant might more appropriately be described as lapsed, as he had been unable to complete the lab work necessary for his prescription refill. Irrespective of the motivation, finishing did not require the client to do anything; they could simply stop responding to Nurx messages. Even so, several participants used the messaging system to let Nurx know that they would no longer be using the service. --- Client Experience: Satisfaction After Overcoming Skepticism Most interviewees reported being extremely satisfied with Nurx services and peppered descriptions of their experiences with adjectives such as friendly, professional, and-above all-convenient. Illustrative of the general trend, P04 summarized: For someone whose life is busy and unpredictable, this service was like a dream. This really was perfect for my needs. Such expressions of satisfaction were common, even among the 10 interviewees who, for various reasons, had not actually accessed PrEP through Nurx or were no longer using the service. As one participant who had moved to a health maintenance organization and found Nurx was no longer covered said: --- Long story short-any company that is trying to cut [insurance-and cost-related] barriers down, I really root for. So, the only sad thing [about not being able to use Nurx] is that they're going to lose an active user, now. I want their numbers to be up. [P25] The pervasive satisfaction was striking because interviewees often related having to overcome an initial hurdle as they contemplated using Nurx: wondering if the service could possibly be legit or whether it was too good to be true. This was partially driven by general skepticism about internet content and web-based interactions and was one way in which the web-based format exerted a unique influence on clients' experiences. As such, the telehealth option presents a clear contrast to brick-and-mortar clinical interactions. One interviewee explained: Getting a prescription online, it doesn't seem super legitimate at first...I just had to make sure that...people were using [the website], and...it was working for them before I put my personal information on there.. --- .Because I don't want to just hand that to some random person on the Internet. [P28] In general, interviewees recounted engaging in research intended to ascertain the service's legitimacy. This could include evaluating the Nurx site itself, accessing user reviews, conducting web searches, and reading about the service on trusted news and information sites. In a typical example, one interviewee listed multiple sources that bolstered his confidence in Nurx: --- [The company] had articles talking about it. [The website] looked clean You know, people posted their experiences about it. So, I knew...it's not some scam website that just wanted my health information or something like that. [P21] In addition to general concerns about the veracity of web content and the legitimacy of the service, the idea of web-based PrEP access gave some interviewees pause because of what they perceived as potential shortcomings of telehealth. For example, one participant was hesitant because: --- [Nurx] could very easily not get the whole clinical idea of what's going on with the patient because you're not actually laying eyes on them. And, you know, to me, that just seemed, like, really weird, because I [hadn't] really explored telehealth or anything like that at that point. So, it was just really weird to me. But I was just, like, okay, whatever. I can't get it through my doctor and I really want to get it, so let's reach out through this and try it. [P24] PrEP-specific logistical questions also surfaced. Some interviewees who had taken PrEP before learning of Nurx voiced a perception that it might be more complex than telehealth could accommodate. As one participant said: --- I was like, "This is such a cool idea." There was, I think, not skepticism, more like disbelief because it was so unlike anything that I'd seen before and I couldn't really conceptualize, "Well, how does this really work?" Because this is something that seemed too complicated. [P04] Another interviewee, who had been taking PrEP for 2 years before discovering Nurx, said his initial concern had been that: --- you have to have lab work. So I didn't understand how that would work out [online], until I actually went to their website and read the information. [P36] It should be noted that the doubts reported here arise because of the web-based format, that is, the very feature that made the service novel and attractive also raised barriers to its use. Clients who experienced this had to seek out specific information and allay particular suspicions before they felt that PrEP access through telehealth was a truly viable option. These users became sufficiently convinced that Nurx was legit to register and submit requests. The foregoing notwithstanding, Nurx's service was not flawless. Especially when pressed , clients mentioned issues in various parts of the service flow. Some participants had trouble finding the area on Nurx's website that dealt with the PrEP service , and others explicitly wished for an app . Some clients had problems at the laboratory they visited , whereas others encountered challenges with insurance . Shipping or receiving medication was a frequent source of trouble, especially the FedEx signature requirement . To provide more context on these issues, we highlight the case of P07. After signing up with Nurx, he had made a laboratory appointment for the following week and then encountered what he characterized as a little bit of a rough patch. This seemed to be a mild description of what was actually one of the most trouble-ridden service journeys in the data set. At his laboratory appointment, a data entry error meant that the staff could not locate the appointment. He made a second appointment but forgot his confirmation number; the laboratory staff said they were unable to help him without it. He made a third appointment a few days later, and that appointment was short and smooth. After subsequently receiving confirmation from Nurx that they could prescribe PrEP for him, he had problems setting up an account with the pharmacy handling the dispensation and delivery. There were issues with his insurance and confusion about how much he would have to pay for the medication. However, at most of these junctures, he mentioned messaging Nurx through the platform and receiving help. He described the challenges he had confronted as "purely, like, circumstantial. It wasn't the process by any means." When asked how the experience of using the service compared with his expectations, he said: --- It's exceeded actually. I think it's made getting this done...extremely stress-free other than the little hiccup that I had the first time. [P07] This is illustrative of one of the strongest trends in the data set: although interviewees could recount challenges with the service, their overall assessment of Nurx was positive. Many participants frequently recommended it to friends and sex partners, and several voiced solidarity with the company. For example, P21 said: I love like the whole startup thing, you know...I want to see Nurx sort of succeed. Indeed, only one interviewee's assessment could reasonably be described as lukewarm overall. Partly because of this, in the rest of this paper, we will pivot to interrogate the pervasive satisfaction that colored the narratives, understanding how Nurx services produced this and for what kinds of clients, as this can provide a window onto which barriers telehealth actually alleviates. --- Client Experience: The Production of Satisfaction At base, we consider Nurx to have produced satisfaction by achieving an acceptable balance between 2 fundamental client desires, which we have termed efficiency and humanity. Efficiency encompasses the ease of obtaining PrEP, both in terms of how simple and convenient the Nurx process itself is, and the ability to avoid disliked aspects of other ways of getting PrEP. Humanity covers the clients' wish for personalized, responsive interaction and a feeling of connection or care. We draw from interviewees' narratives to illustrate these dynamics and their interplay as well as highlight the influence of the web-based format. --- Efficiency was the way interviewees most clearly contrasted Nurx with traditional clinical contexts. The greatest gains in efficiency accrued when users who otherwise encountered obstacles locating a knowledgeable, willing, available prescriber were able to access PrEP through Nurx [21]. One interviewee, for example, described PrEP as barely a thing where he lived, adding that few health care professionals seemed aware of it. Several of this interviewee's friends had related that: their doctor doesn't know anything about it, and they're not comfortable prescribing it to them. So, they usually get referred out to the [AIDS Health Foundation] in town...So, I was actually going to go there to try to get a prescription, but they were booked out, like, six months or something for appointments. [P30] Shortly after learning of the 6-month wait, this interviewee discovered Nurx had expanded its services to his state and he signed up right away. Even users who did not face such extended wait times for appointments described the ability to start the PrEP process at their convenience as removing a significant constraint. In addition, the quasi-anonymity of submitting answers to behavioral screening questions on the web was highly valued. One participant noted that going to the doctor "for, like, sexual type, gynecologic visits" typically provoked nervousness and stated: Nurx's method of obtaining information on patient sexual practices reduced embarrassment, awkwardness, or shame, which these interviewees recounted fearing or having experienced with previous providers. We categorized this as efficient because it avoids unnecessary and unpleasant elements of obtaining PrEP in person. I The service flow often went smoothly. When it did not, however, assistance from Nurx-human intervention-was welcomed. For instance, many clients reported great concern over issues of cost and billing and a desire for personalized, real-time responses to questions about these issues. One interviewee reported that his only concern about obtaining PrEP through Nurx had been uncertainty around the cost. After receiving an initial bill of several thousand dollars for a 3-month supply of PrEP, in sticker shock, he contacted Nurx and discovered there was a problem: with how my insurance was filing it or something. So, I did have to go back and forth with them a little bit to figure that out. And the price did go down at some point, but it was still over-I think it was, like, $1500 or something was the actual amount that I had to use [the Gilead Patient Assistance Program card] to pay. --- [P30] In such cases, Nurx's ability to respond quickly and individually was key to clients' satisfaction with the service . With very few exceptions, these interactions occurred through the platform's messaging system. Indeed, when clients talked about the service Nurx provided, they mostly discussed messaging. Hence, we examined clients' perceptions of messages in detail. --- Messaging Several interviewees spoke of previously engaging in messaging with a provider through electronic portals that were a part of larger health systems, so this was not always a novel experience. One interviewee reported using such a system to email his primary care provider , asking to "get [PrEP] started." Once the interviewee learned that he would need to see his PCP in person before starting PrEP, he "decided to try and go through Nurx because [the health system's] availability for check-ups and just anything like that is really impacted" . Although messaging is not necessarily unique to telehealth, what a patient can expect to accomplish by messaging Nurx versus messaging a PCP is likely different. As previously mentioned, some interviewees noted the potential for feeling less judged through a web-based interface. When asked why this was so, one user reflected on providing answers to the sexual behavior questions: --- Well, people say that emotions or intention is a little bit [harder] to read through text message than it is face-to-face. I honestly feel it kind of applies to what's going on. [P03] Interestingly, however, messaging was described as being able to both block the communication of emotion and convey the sense that Nurx was friendly and approachable. As one client explained: The way they write their messages is very cheerful...They addressed you by name like, "Hi, G. my name is so and so and I'm here to collect your insurance information."...I guess I expected it was going to be much more robotic, like, "Please submit such and such information by this date." It wasn't like that at all. It was much more personable, like you were just texting a friend back and forth or something. --- [P13] In this sense, the messaging platform figures in clients' narratives as a tool that Nurx can deploy in the service of both efficiency and humanity . Many interviewees described the message system as similar to texting and, thus, familiar. One participant explained why this works in Nurx's favor: The thing that I know is that millennials hate to be inconvenienced. If.. --- .they don't have to actually go and talk to someone, I think they will be more inclined to see the process through. Because [also,] the process mimics that of a texting platform that they're used to being on. It mimics that of a social media platform with the way you can post photos to your physician. [P02] However, some interviewees directly addressed the use of automated messages, seeing them as distinct from those that felt like texting a friend: --- I can definitely, in the correspondences, tell if it's a form or if it's casual. Obviously, it helps if you ask, like, a bespoke question and receive a bespoke answer. Clearly, that's not coming from an AI bot. But as far as automated, like, "Hey. Thanks so much for filling out your application. We're running at full capacity right now and hope to have your results in the next day or so."...No one sat behind a computer and just casually wrote that before going to the bathroom. You know? That's read over many times by a lot of people, and they thought this was the best they could say. [P23] --- And how do you feel about that, that there are automated responses going out? [Interviewer] --- It's a service. It makes sense that it would be automated. [P23] Although this interviewee had received both form and casual responses from Nurx, here, he clearly constructs the feasibility of the service in its web-based format as dependent upon some degree of automation, that is, to remain in business, Nurx as a commercial entity had to serve a sufficient number of clients, which automation made possible. Another interviewee echoed this notion, suggesting that clients sometimes had to accept interactions that felt somewhat disconnected. For example, when specific questions arose , messages could be answered by various customer service representatives working the message queue, making him feel: In the flow, it's kind of like, I don't have a specific person to talk to...But, it's an interesting sort of trade-off. Because, at the same time, it was so easy to go through that and...I understand that as being the tradeoff. For like, the ease and simplicity, vs. like, having that like, specific person to contact and like, them being like, readily available. [P25] This client's construction of a trade-off crystallizes a more implicit notion that was common among client narratives: that efficiency and humanity are inversely related, such that as efficiency increases, humanity decreases and vice versa. The challenge for Nurx was to provide the efficiency clients sought while remaining human enough. A case in point is the experience of actually receiving the medication. Clients described home delivery as an attractive feature of the service, imagined as consummately efficient. However, it was also a touchpoint at which efficiency could falter in practice because of the FedEx signature requirement in effect during data collection . That a human, physical presence was necessary to complete a transaction touted as virtual was both a conceptual and practical obstacle for many users and detracted somewhat from the seamless way Nurx aimed to fit into its clients' busy, highly mobile lifestyles. Some users had to identify alternative delivery addresses; others consistently had to travel to FedEx locations to collect their prescriptions. It might be surprising that clients accommodated such demands, given the premium they placed on convenience and efficiency. Our analysis suggests that the concept of a trade-off may help explain. Specifically, it seems that the clients we interviewed were willing to take such actions in exchange for the convenience offered by the other phases of service flow and as long as emergent obstacles throughout the client journey were handled with sufficient humanity. Indeed, it was not only clients who had no other way to obtain PrEP that used the service; those who might otherwise have obtained PrEP from PCPs often chose to remain with Nurx as well. As summarized by one client: --- Despite all of these complications...getting access to this drug and staying on it-because it has been kind of difficult for me-...each person I've talked to [at Nurx] has, to me, seemed very genuinely concerned and sympathetic and really willing to help me out. [P15] This makes plain the importance of the balance that Nurx strikes between efficiency and humanity. --- Efficiency-Humanity Balance The foregoing notwithstanding, within the data set, clients expressed varying levels of tolerance and appreciation for efficiency and humanity. Some gave the impression that they were close to needing more of the latter than Nurx provided. One interviewee , an ardent proponent of self-monitoring and quantification, had negotiated to have the Nurx physician submit standing orders for STI screening at the laboratory where he did the required testing. This meant that he did not have to wait the customary 3 months; it is unclear how failing to obtain this from Nurx might have impacted his satisfaction with or continued use of the service. Another participant explained that: the problems that I have with [the service] aren't really, like, major issues but he recommended Nurx "be more personalized" in their communication with clients, because: --- using my name isn't going to really cut it. And very rarely do we get into, like, personal conversations unless I initiate something...yes, the anonymity helps, because you just have a little picture in a box [when chatting]. That's great. [But] that could not actually ever be a person. You could be in China, like, cutting or pasting on 27 different things like everybody expects is happening through telehealth. [P24] Thus, a lack of personalization led the participant to question whether there is a person on the other end of the chat at all or whether the person really is who they claim to be. Indeed, never physically seeing the physician seemed to have left a kind of residual uncertainty for several users. One asked: --- Is [Dr. X, a Nurx physician] a real person?...There's, like, a picture [on the site] of a really pretty lady, and it seems like she might be a stock image or something, you know? [P30] --- Yeah [...] she's a real person. [Interviewer] --- Okay. I've been wondering that. Okay. [...] [P30] Was there something else, about the way that she was interacting with you, that made you wonder about that? [Interviewer] No. It was always, like, a really personalized response. So, I knew it wasn't coming from a robot. --- Yeah. I just kind of-I think it was the photo. [P30] This skepticism about the true identity of Nurx service providers echoes the initial doubts some users expressed about the service itself, in that it is engendered by the web-based context of the interaction. However, as with these initial doubts, the discomfort this produced was not sufficient to dissuade these interviewees from using the service. In contrast to interviewees who seemed to have their needs for humanity barely met by Nurx, others appeared to have a much greater tolerance for efficiency than the service required, especially if they had existing relationships with other care providers. For these users, a sense of emotional resonance in Nurx's messages or personal connection with the PrEP provider was not that important. For example, in contrasting web-based and face-to-face clinical encounters, one participant said: [Online is] not as personable, you know, but that's honestly not a big deal to me...I like to get in there, get everything taken care of, and go, you know? We --- don't have to, you know, ham it up and just have this great, friendly relationship. [P27] These data suggest that users arrive at the Nurx clinical encounter with a set of individual preferences and needs regarding the patient-provider relationship, an orientation that circumscribes a range within which interactions will be considered acceptable. At one end, some clients seemed to wish for slightly more humanity than was built into the service's default operation; at the other end, clients were satisfied though basically uninterested in having a relationship with Nurx that went beyond the transactional. Similarly, Nurx is configured in ways that enable a range of service-user interactions and make others less likely or impossible. When the user and company ranges overlap, clients are likely to judge the resulting interactions as satisfactory, as long as their end goal is also achieved. On the basis of our data, the ideal Nurx client is a person who has access to and is relatively comfortable using the internet, did not experience or was able to overcome initial skepticism, feels their schedule is busy enough that all kinds of real-time appointments are a chore, may wish to avoid face-to-face discussions of sexual practices, and/or does not require an intensely personalized relationship with the PrEP provider. --- Discussion --- Nurx: Telehealth Considerations and Lessons Although telehealth-based PrEP provision is increasingly mentioned as a potential way to broaden access to this HIV prevention strategy [22] and a handful of innovative interventions have appeared in the literature [12], fine-grained data on the patient or user experience are scarce. The partnership between UCSF researchers and Nurx enabled the collection of qualitative data to shed light on this topic from within a commercial environment. Our analysis led us to explain client satisfaction as resulting from the balance the company struck between what we have termed efficiency and humanity . The concept of a trade-off explains how certain highly valued parts of the service helped users accept other relative inconveniences . In addressing the important questions of whether Nurx provided a way to overcome common barriers to PrEP access and what difference the web-based format made, we ground the discussion firmly in our data while answering with an eye toward telehealth more generally. In light of the massive surge in interest and use of web-based formats for medical appointments prompted by the current global COVID-19 pandemic, we hope that scholars and practitioners will find the insights and questions we raise transferable [24] and valuable. We found evidence that Nurx allowed patients to locate a knowledgeable and willing prescriber . In addition, receiving PrEP services from Nurx enabled users to avoid stigmatizing or embarrassing encounters with providers. These outcomes are important and should not be minimized, particularly as they manifested among a diverse patient population that included young men of color who have sex with men, a group disproportionately impacted by HIV [25]. That impact notwithstanding, in assessing Nurx as an intervention, we should distinguish between outcomes that are produced by, as opposed to merely facilitated by, telehealth. Connecting providers and patients despite the geographic distance between them and reducing the burden of medical appointments are inherent capabilities of telehealth. In contrast, that Nurx personnel and physicians were experienced as nonjudgmental during potentially sensitive conversations is not produced by the web-based format, in the sense of being guaranteed simply by virtue of happening on the web. While technology enabled users to interact with Nurx personnel without being face-to-face , the content and tenor of those exchanges also certainly influenced users' comfort level. Both Nurx's underlying company philosophy and their experience of providing oral contraception may have contributed to their capacity to create interactions users described as nonjudgmental and sex positive. If the providers on the other end of the messaging platform had lacked cultural competence in providing sexual health services [26], it is unlikely that users would have felt so at ease. This points to the importance of particular types of providers and messaging as components of a telehealth system [21]. As mentioned earlier, although Nurx helped patients who had previously encountered difficulties in trying to access PrEP, interviewees also recounted experiencing substantial skepticism or uncertainty about the service, especially initially. Although some of these doubts were PrEP related , others grew out of the remote or web-based modality of care and may not be specific to Nurx or PrEP. Thus, regarding telehealth generally, it is imperative to recognize that the web-based format itself may engender barriers to be overcome. This possibility tends to be absent from discussions of the promise of telehealth. Those wishing to use telehealth approaches should consider what information or strategies would prevent such doubts from arising in the first place, or at least how to allay them should they arise. Another crucial point to make, based on stories interviewees shared, was that much of what influenced their assessment of Nurx lies outside of the telehealth experience itself. Especially when confronted with novel situations, users of the service drew on previous experiences to inform their notions about acceptability and appeal. For example, most users were unaccustomed to messaging back and forth in anything close to real time with a medical provider. However, such behavior was familiar from other domains , and it seemed that not only the practice but also the meanings attached to that practice-that is, friendliness and responsiveness-became associated with Nurx. On the other hand, past real-life experiences were sometimes contrasted with what occurred via telehealth. In particular, most clients drew on face-to-face medical appointments as the implicit benchmark against which they evaluated their Nurx experience. As in-person appointments often came with what clients saw as hassles, such as lack of convenient appointment times, the embarrassment of talking about sexual health and practices, and providers who were not knowledgeable about or supportive of PrEP, inconveniences encountered with Nurx seemed minimal in comparison . When creating telehealth services, it may be helpful to explore what previous experiences users could draw on to help them interpret their new telehealth experiences in a positive light and use design to maximize these associations of relative advantage. In addition, although often not explicit in interviewees' accounts, any social determinants of health perspective [27] will acknowledge that cultural and structural factors play a role in the clients' assessment of Nurx. For instance, the degree to which clients anticipate or have experienced embarrassing or shame-inducing sexual health discussions with providers may depend on the cultural norms of the community where they live or have previously received care, which, in turn, likely undergirds how appealing it would be to avoid such encounters, and therefore how attractive clients find Nurx's model. Examples of structural influences are geographic distribution of PrEP-knowledgeable providers, what kind of insurance clients had , and whether they lived in states that had expanded Medicaid, all of which impact PrEP accessibility. To the extent that national policy influences insurance coverage , it should also be accounted for. Although it is understandable and appropriate that assessments of technology-based health interventions focus on feasibility, acceptability, process metrics, and behavioral and clinical outcomes [28][29][30][31][32], we urge scholars not to leave the wider social context in which telehealth is embedded entirely outside the frame. One further point we wish to raise for discussion is the pressing need for fine-grained inquiry into PrEP telehealth. Our research indicated features of Nurx that were key to the appeal and users' experience of the service, as well as revealed variation in terms of the overall balance interviewees desired between efficiency and humanity. In short, patients exhibited a range of needs and preferences for clinical encounters, and Nurx, like any technological intervention, was built to accommodate a particular range of these preferences. When these ranges overlapped, and the end goal was achieved, satisfaction was often the result. A serious limitation of scholarly production on nontraditional forms of PrEP access is that it rarely engages in a meaningful way with the differences in format, service design, or context of implementation among interventions. As such, we agree with Mayer et al [7] that "studies that identify the core components of effective programmatic partnerships are needed." They continue, however, to posit the goal of such work as the development of "normative guidance" and the promotion of "best practices for local PrEP implementation programs" [7]. We acknowledge best practices as potentially helpful and effective as guidelines, but as we have argued elsewhere [21], we believe that interventions, contexts, and users are mutually constitutive, which is to say that, in different contexts or with different users, even the same telehealth intervention might mean something different and hence function quite differently [33,34]. Indeed, best practices may vary so widely across different locales, key populations, or user orientations so as to preclude a singular right answer about how to do telehealth. Even within the same context, clients with different needs and preferences than the ones we interviewed could easily evaluate Nurx quite differently. For example, individuals with serious privacy concerns about the web or those who could not get over their worries that Nurx is a scam might be more reassured by and therefore more likely to opt for a service that featured real-time videoconference or telephone appointments with a medical provider. Different flavors of telehealth will appeal to different kinds of patients. Rather than assuming that we know how client-patients experience these interactions, researchers should strive to illuminate the mechanics at play, that is, we need to ask the following questions: What is the client's experience? How is this experience produced? For what kind of clients does this hold true? Digging into the specifics of the client experience in this way allows us to avoid fetishizing the modality of service delivery-after all, technology is a tool, not an end in itself. --- Limitations As with any research, this study has important limitations, mostly to do with potential selection bias among our interviewees. Most obviously, people who lack access to the internet or are unaware of PrEP and/or Nurx could not have registered as Nurx users. In addition, potential users who felt the need for an intensely hands-on experience with a medical provider around PrEP would likely not have found Nurx services attractive in the first place, and those who experienced but were not able to overcome strong skepticism about Nurx probably would not have registered as users on the site. Thus, we were unable to interview anyone representing these groups. Among the individuals we did interview, several notable categories of potential clients may be underrepresented. First, none of our interviewees identified as American Indian or Alaskan Native, Native Hawaiian and other Pacific Islander, or transgender. The latter is particularly lamentable because people of trans experience figure among those for whom telehealth has been proposed as a way to circumvent barriers to care [35]. Second, although we opted to recruit PrEP requesters in a bid to reach individuals who had negative experiences, those users may have been less likely than their satisfied counterparts to be interested in participating in this research. However, as our analysis suggests that patients with different needs may gravitate to other ways of obtaining PrEP , this simply means that other studies should pursue a similar detailed line of inquiry about patient experiences with those other methods. By comparing and contrasting our findings, we may be able to derive key characteristics or needs that would indicate how and for whom differently designed services work best. Finally, we recognize that users who lacked other ways to access PrEP may have been especially likely to report feeling positive about Nurx, even if they experienced challenges in using the service. However, this would be true of any PrEP service when a user has no other mode of access, and that other options are unavailable or unappealing is part of what produces satisfaction with real-world clinical settings. While acknowledging this, we do not believe it undermines our analysis, as it is focused primarily on understanding not how many interviewees said they were satisfied but how that satisfaction is produced. --- Conclusions Interviewees recognized the web-based nature of Nurx, with its potential for real-time communication and use of automation, as a unique platform that enabled a novel form of PrEP access. However, there was variation among clients in the efficiency-humanity balance they wanted Nurx to strike. Some clients very heavily valued efficiency and seemed to see Nurx as almost analogous to other commercial entities with whom they might have a subscription . Others demanded a more personal touch and seemed to interpret the relationship they had with Nurx-technological mediation notwithstanding-as a caring one. However, for all clients, every step in the Nurx journey was simultaneously technological, clinical, and social, informed by previous experiences in various domains , as well as cultural and structural considerations. While telehealth is not a panacea, having multiple, differently designed access options available may fit the needs of the broadest swath of potential users, thereby opening new spaces for therapeutic engagement. --- Conflicts of Interest None declared. --- Abbreviations PCP: primary care provider PrEP: pre-exposure prophylaxis STI: sexually transmitted infection UCSF: University of California, San Francisco ©Shana D Hughes, Kimberly A Koester, Edvard Engesaeth, Merissa V Hawkins, Robert M Grant. Originally published in the Journal of Medical Internet Research , 05.07.2021. This is an open-access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.

Background: HIV pre-exposure prophylaxis (PrEP) is a way to prevent HIV infection using antiretroviral medications. However, common barriers to PrEP engagement include lack of access to prescribers; discomfort seeking sexual health services; and racism, homophobia, and transphobia in medical contexts. Key populations (eg, communities of color, young men who have sex with men, and transgender women) are underrepresented in terms of PrEP uptake in the United States. Nurx is an innovative company that has offered internet-based access to PrEP since 2016. Objective: In this study, in partnership with Nurx, we aim to explore clients' experiences of digital PrEP access-including the difference made by the telehealth format-and to understand whether Nurx helped reduce barriers to PrEP. Methods: Electronic chart review and semistructured interviews were conducted with 31 PrEP requesters from California, Florida, Illinois, and New York. Interviews were recorded, transcribed, and subjected to inductive and deductive thematic analysis. Results: Some interviewees reported initial skepticism about whether a web-based PrEP service could be legitimate or feasible. Despite this, most clients were effusive about their eventual Nurx experience, and many reported that Nurx eased barriers to PrEP access through the availability of knowledgeable, willing prescribers and minimizing embarrassment and discrimination. Our analysis suggests Nurx produced satisfaction by achieving an acceptable balance between 2 client desires: efficiency and humanity. Efficiency encompasses the simplicity, speed, and convenience of obtaining PrEP, both regarding the Nurx process itself and in comparison with in-person encounters. Humanity covers clients' wish for personalized, responsive interaction and a feeling of connection or care. Nurx's messaging platform was crucial to manifesting these qualities and was largely interpreted through the familiar frame of texting. Clients conceived efficiency and humanity as inversely related in a commercial enterprise and varied in the particular balance they felt was optimal. Those who wished for slightly more humanity than the service afforded used the concept of a trade-off to explain why Nurx remained appealing.Our findings augment evidence that internet-based PrEP provision can broaden access to this HIV prevention strategy. This important finding, notwithstanding a few provisos, merits mention. Telehealth, as practiced by Nurx, was still dependent on culturally competent medical providers as system inputs, and the very technology used to overcome access barriers (ie, the internet) generated new hurdles for some clients. Furthermore, clients did not interpret Nurx in a vacuum: their past experiences and the social and structural context mattered. Finally, only granular inquiry revealed precisely how Nurx satisfied clients whose experiences and preferences fell within a particular range. Extrapolating from this, we urge scholars not to fetishize technological solutions but rather to interrogate the ways in which any intervention's design works for certain kinds of patients.

Background and Introduction: Health is complex and multifaceted, comprising domains including but not limited to physical, mental, and psychological health. While the built environment has proven to influence these domains, literature is still lacking in studying the effect of the built environment on social participation, as well as its influence on health outcomes in Singapore's local context. Given the nation's rapidly ageing population, rising chronic conditions and increased load in the provision of care, there is an urgent need for effective solutions to improve health. In particular, the built environment holds great potential, offering an opportunity for simple, tangible, and long-lasting interventions. Objective: This study aims to identify the barriers and facilitators of the built environment, both tangible and non-tangible, that limited or facilitated social participation in communal activities among seniors residing in a low-income urban community in Singapore. Methods: Photovoice go-along interviews were conducted in addition to semi-structured sit-down interviews with 25 seniors aged 65 and above residing in a low-income urbanised community, Marine Parade. Participants took the interviewer on a journey through their neighbourhood with polaroid cameras, then using the instant photographs, spoke of their thoughts and shared stories related to the aim of study. These interviews focused on elucidating the aspects of the built environment, both tangible and non-tangible, that limited or facilitated social participation and ultimately mental well-being, and physical health outcomes. All in-depth interviews were audiotaped, transcribed, and uploaded into the qualitative software NVivo, and responses were coded with theoretical and emergent themes regarding participants' perceptions of the built environment. Key conceptual categories were constantly refined until a saturation of categories informed a framework outlining the aspects of the built environment that impacted health outcomes. Findings: Collectively, 106 categories and 18 broader themes emerged. These preliminary themes reflected concerns of mobility as well as convenience, circulation, visibility, recognisability, and peacefulness of the space. Ideas were shared about how to improve the safety of pathways and routes, linkages to transportation as well as how to leverage on existing unused space to encourage social interaction. Implications on Ageing and Gerontology: Understanding how the built environment influences social participation could be key to changing health outcomes. As we grow into a Super -Aged Society, we hope to harness the full potential of the built environment towards independent and joyful ageing.

Barriers and facilitators of the built environment towards social participation, mental and physical health outcomes among seniors residing in a low-income urban community in Singapore: A qualitative study.

INTRODUCTION Sexual violence is a complex social problem that accompanies human historicity for centuries and, due to its high prevalence, is considered a public health problem . The concept of SV ranges from making gestures/comments aimed at the sexuality of another person, passing through commercialization, to the sexual act without consent or with the consent of someone who does not have the physical/emotional and/or cognitive maturity to do so . Considering that gender issues involve SV against girls and are supported by the patriarchal legacy in several Western societies , it is feasible to affirm that these serve as support for offenders to justify the perpetration of SV against them. Nevertheless, boys are also victimized in childhood/adolescence , which shows that SV against children and adolescents presents complex issues, influenced by the culture and the exercise of offenders' power over their victims, who are in a more vulnerable emotional and physical condition . The natural stage of development and the growth of children and adolescents also put them in a condition of vulnerability to SV. It should be noted that psychomotor development is conceptualized as improvement of psychological/emotional and motor skills, while physical growth or development concerns the constant and irreversible increase in individuals' . Thus, childhood and adolescence, identified as periods ranging from 0 to 9 years and from 10 to 18 years of age, respectively , are stages of life in which individuals have different human needs. That said, children and adolescents need specific care, considering that this process is influenced by nutritional, genetic and environmental aspects that can be determinants in healthy growth and development at this stage of life. Also, during this stage of the life cycle, individuals have peculiarities limiting their ability to self-protect. These include physical strength, to resist contact with the offenders, emotional maturation, to assess the consequences of sexualized acts and the difficulty of verbalizing the occurrence of this type of violence, for not understanding its sexual content . Such limitations may also be present in people with disabilities . These limitations favor the unbelief of adults and the delay in revealing what happened when SV cases are expressed by the victims, especially when the offender is a close relative . Children and adolescents are not considered as a subject of law, as they evoke national and international laws . These gaps hinder the identification of cases and the implementation of care provided to the victims' health. Recent studies reveal the correlation between childhood maltreatment and negative impacts on cognitive, psychological, and sexual development, long-term functional and physical impairment, which reflect in adolescence, adulthood, and may transcend reaching intergenerationality . Thus, the characteristics of SV suffered by children show psychosocial and cultural factors that make it difficult to reveal the abuse suffered by children and adolescents, creating obstacles to effective reporting and investigation of the facts 9) , which presupposes underreporting and underreporting of occurrences in official bodies. Therefore, it is necessary for the multidisciplinary team to be qualified to identify cases, provide comprehensive and specific care to victims , aiming at acting in an interdisciplinary network and weighing their emotional and physical development. It is noteworthy that the reception, through listening and the establishment of a relationship of trust, is a transversal step in this type of service, constituting a differential in the elucidation of the facts and in the recovery of victims of SV , its execution by the multidisciplinary team is essential. In addition to this, nursing, as a category that establishes greater contact with users of health institutions, plays an important role at the time of reception . Therefore, it is urgent to carry out studies that elucidate aspects of the physical and emotional development of children and adolescents who have suffered SV, from the multidisciplinary team' perspective. Due to this gap in the literature, the following investigative question was outlined: what is the perception of health professionals about aspects of the physical and emotional development of children and adolescents who have suffered SV? This study has its relevance based on the possibility of giving rise to a discussion about the sexual victimization of children and adolescents, enabling the creation and implementation of strategies aimed at protecting this public from this type of health problem. --- OBJECTIVE To identify the multidisciplinary health team's perception on aspects of the physical and emotional development of children and adolescents who have suffered SV. --- METHOD --- Ethical aspects The project was approved by the Research Ethics Committee, meeting the ethical standards of research with human beings, according to the Resolutions of the Brazilian National Health Council . The Informed Consent Form was prepared, which included the main researcher's objectives, data, motivations and other information about the study. The ICF was read and signed by the selected participants, and the interviews took place in reserved rooms, in order to guarantee information privacy. To protect the participants' identity, their names were replaced by numbers from 1 to 30, preceded by the letter "P", referring to the professional, according to the order in which the interviews were carried out. --- Theoretical-methodological framework This is a research based on Symbolic Interactionism , proposed by George Herbert Mead, considered the precursor of this philosophical movement. This theory focuses its attention on the mutual influence of humans on social development, suggesting that the way an individual acts provokes reactions in the other. It also highlights the relevance of communication and the influence of the symbol as forms of recognition among individuals . This study confirms that SI is a necessary requirement for care in all its dimensions, as it serves as a foundation in the care provided in health services with regard to continuity of care . Pondering about it, care for children and adolescents who are victims of SV lacks complex care, demanding interaction with victims and their families. Therefore, SI was chosen, as it makes it possible to investigate the real human behavior of multidisciplinary team members in the face of this service, resulting from the experiences lived by individuals . --- Study design This is a study with a qualitative methodological approach, based on the use of COnsolidated criteria for REporting Qualitative research . --- Methodological procedures --- Study setting The study took place in a general public teaching hospital in Bahia, Brazil. This institution offers clinical, surgical, emergency and outpatient medical-hospital services for all age groups. The approach to the field occurred through the insertion of the main author as an assistant nurse in the service, with a previous approach to care for victims of SV. Multidisciplinary team members working in Pediatric and Adolescent Inpatient Units, Intensive Care Units and Obstetric Center were approached. --- Data source Thirty professionals participated: 10 nurses, 10 nursing technicians, 03 doctors, 03 social workers, 02 nursing assistants and 02 psychologists. All were women, self-reported of gender identity cis woman and heterosexual affective-sexual orientation. 24 were of black race/color, and 14 were single. The choice of participants was intentional, including health professionals who had been working at the institution for at least one year. Those who worked in diagnostic support sectors were excluded. There were no withdrawals among the participants, nor refusals to participate in the research. --- Data collection and organization Data collection was carried out in person, between the months of June and July 2019, in the three periods . A presentation was made to the coordinators to publicize the research. The team of interviewers introduced themselves and approached the professionals in the aforementioned care units, and all who met the inclusion criteria were invited to participate in the research, to arbitrate their participation, or time was not given. Once the invitation was accepted, a second meeting was scheduled. The participants answered only once the 12 closed-ended questions referring to sociodemographic data and a questionnaire with an open-ended question that guided the interview: report cases of child and/or adolescent victims of SV that you provided care. During the interviews, other questions were presented, in a correlated way, to professionals' narrative, aiming to expand and favor the understanding of the above. The interviews were conducted by the main author, a Master's student at the Graduate Program at the School of Nursing at the Universidade Federal da Bahia, and other members of the Child and Adolescent Health Study Group , all previously trained in an extension course in approaching the theme, in the collection and treatment of qualitative data. The interviews took place in the hospital unit, were recorded on an Android cell phone and lasted from eight to 45 minutes. The end of data collection occurred due to data saturation, being discussed among the researchers. The reports were transcribed in full by CRESCER undergraduate students, corrected by volunteer nurses in the research and validated by the main author. --- Data analysis Content analysis was conducted using the model proposed by Bardin in four stages: 1) pre-analysis; 2) material exploration; 3) treatment of results; and 4) interpretation. The technique guided the data assessment and interpretation in a process that involved material text skimming, followed by repeated, detailed and in-depth readings of the transcripts. The notes of a field diary written by the interviewers also integrated the analyses, which generated codes debated between the coders and the other authors of the study . After the analysis, themes emerged that were grouped by content similarity and highlighted aspects that gave rise to two thematic categories: Physical and emotional aspects of child victim of sexual violence; Physical and emotional immaturity of adolescent victims of sexual violence. --- RESULTS --- Physical and emotional aspects of child victim of sexual violence For the participants of this study, the cases of SV that drew the most attention were those that occurred in early childhood. --- I took care of two girls, one of them was one year and four months old, her vagina was torn [...] --- It was a boy of one year and six months, wearing a cloth diaper that was full of blood in the buttocks area. [...] the two-year-old girl was raped by her stepfather. In contemporary society, the image of the young child reminds us of the angelic image of chastity and purity, and it is inconceivable that these can suffer some kind of SV. Thus, at different times, participants highlighted the innocence of child victims: The girl was three years old, a small child, a helpless child who had been raped. [...] she was a girl, infant, less than one month old, 28 days old. She was too small to suffer the way she suffered. The speeches highlight children's characteristics -innocence, defenseless, small -as if these attributes could be a shield to protect SV and not vulnerability factors to this type of aggravation. In addition, by highlighting the victims' suffering, participants demonstrate their ability to act with empathy, a necessary tool for care and in accordance with SI concepts. Also, as described by the interactionists, individuals' ability to establish bonds is due to the social relationships stimulated, replaced and accumulated throughout life, i.e., they can be transformed, reflecting the meaning that caregivers attribute to the violence experienced by children. The participants describe the physical characteristics of children who are victims of SV at an early age as "healthy", "strong", which refers to the comparison with baroque angels: The collaborators describe the victims' body structure and demonstrate that, according to their perception, the presence of adipose tissue is synonymous with beauty and refers to the stereotype of a healthy child. These images and the use of diminutive refer to the feelings about childhood expressed in the 18 th century , described by Ariès , which remain until the present day in the western world. In this way, "child being" is considered as an innocent, angelic being, with special needs and that must be protected. Participants perceived in victims of SV signs of abandonment and deficiencies: The child had many neurological sequels, the team suspected SV. The girl was five years old and had very poor speech, even with this deficit, she explained that her stepfather touched her body. She was a child of one year and six months, very depressed, malnourished, bleeding a lot in the perianal region. [...] Nutritional and vocabulary weakness, as well as neurological impairment, added to the poor physical and emotional development characteristic of age, may be responsible for the expansion of vulnerability conditions for the experience of SV by infants. From an interactionist perspective, it is plausible to say that contemporary society, while establishing laws, rights and inclusion for children/adolescents , also denies them protection. --- Physical and emotional immaturity of adolescent victim of sexual violence The professionals' testimonies are concentrated in cases of girls who were in their early adolescence phase . In more detailed reports, the adolescents' characteristics were associated with the physical/emotional development of childhood: The ten-year-old girl was so small that she still looked like a child When approaching the image of adolescents with children, there is an attempt to make a correlation between them. For that, the participants used terms in the diminutive and/or that referred to the adolescents' infantilized condition: "skinny", "skinny" and "small". These terms refer to images of children, i.e., adolescents whose body structure is fragile and, like children, need protection. Also, the interviewees highlight the maturing of adolescents' sexual organs: The pubertal development reported refers to images similar to that of children, i.e., these victims did not have a physical resemblance to adult women. Likewise, when stating that "she did not have a body", there is an emphasis on the fact that the adolescents' physical development was not compatible with adult, attractive and/or sexualized forms. Therefore, it is possible to infer that, in professionals' perception, the violation of the sexuality of these adolescents is something little accepted, which is congruent with the social conception that has its roots in a code of modern moral conduct. These conceptions are based on social interactions, relationships widely discussed by SI. On the other hand, professionals barely remembered and described the victims' physical characteristics, when they did not have an appearance similar to that of infants, as can be seen: The uncertainty regarding the precise information on cases of adolescent victims of SV reveals a weakening of the victims' image in the professionals' memory, since few remember the girls. This may indicate that cases of SV against children affected the interviewees more, while juvenile SV may be the target of the same naturalization attributed to violence perpetrated against women, since some adolescents have a body structure similar to that of women in adulthood. In this sense, the remarkable experience allows individuals to memorize details of their interaction, supporting SI premises. On the other hand, some adolescents also had childish behavior: Child mentality, i.e., adolescents who play, who wear clothes with figures from children's programs, who stay at home studying and do not engage in dating, raised behaviors compatible with immaturity. These conceptions are concepts built in modern society, where "being a child" is synonymous with playing and doing ludic activities, while, in adolescence, sexual life begins, diverting attention from school activities to relationships, such as dating. In this way, there is an attempt to bring the adolescent victim of SV closer to the puerile image of children, trying to bring them closer to a fragile and unprotected being. [...] Also, some young women attended did not have their sexuality developed: I realized, in the service, that, despite being 16 years old, she was very innocent, she was a virgin, she didn't know about sexuality like adolescents nowadays, who are smarter. [...] she was 13 years old; she was young! She didn't have the sexuality that most adolescents do. Being raped was strong for her, she freaked out in the hallway. I attended a 14-year-old adolescent, she was immature, inexperienced, had just started dating and, in my hypothesis, with the beginning of the relationship, she understood that the caresses that her great-uncle used to make her were abuse. For me, she lived the trauma all over again, which was unbearable, and she freaked out! Interactionists affirm that social constructions are imperative in the interaction of individuals. Currently, it is assumed that not having sexuality touched upon, not having a dating relationship and not having early sexual initiation distance adolescents from the sexuality theme. Such perceptions, again, emphasize the victims' naivety, in an attempt to point out this innocence as something that protects adolescents from the experience of SV and not as something that expands their vulnerability to be the sexual offenders' preferred target. Such a conception about the need to experience sexualized acts to understand what a sexual violation consists of can be harmful and hinder the development of capacity for self-protection in the victims, since children and adolescents need to know how to differentiate affection from rape and have security to report sexual offenders to people they trust. Moreover, children/adolescents do not necessarily need to have sexualized experiences to understand what a violation of their body consists of. Supporting the interactionist thinking regarding the evolution of society, the change in this understanding consists of a complex, gradual process that requires time and learning, especially with regard to the sexual victimization of children and adolescents. Thus, until everyone perceives children and adolescents as beings equally vulnerable to SV and creates effective protection strategies, discussions are required in society that should include victims and their families. --- DISCUSSION Regarding the age of victims of SV, studies indicate that cases of children in the first and second childhood are less reported/ denounced. A recent national study, with 214 women, pointed out that 19% of the first episodes of SV occur when the victim is between four and five years old, a phase in which children do not have the capacity for self-defense, but are already able to recall and describe experiences with detail and accuracy . It should be noted that sexual victimization at an early stage of life is a phenomenon that occurs in society since antiquity , persisting to the present day, although childhood is conceived, in contemporary times, as a period of purity and innocence. This conception of childhood emerges since the 18 th century, when children became the target of greater protection from society, despite continuing to be objects of mistreatment, since violence is also conceived as an instrument of education 32) . Even today, the images of candor, of an asexual and chaste being, which are historically perpetuated and translated in the social representation of children with angels, celestial beings, portrayed in the sacred images , are attributed to childhood. The changes in conceptions of "being a child" are based on social interaction in different historical moments and, currently, this is considered a special 32) and symbolic phase , since, in contemporary times, children, especially at an early age, are seen socially as innocent, fragile and angelic, therefore, that could not be tainted by SV, as the participants of this study try to illustrate. Although children are considered incapable of resisting without the care of others 16) , often, the injuries suffered, which can tarnish this image of purity, become invisible, and they may of even be denied the right to speak, which makes them targets of disbelief on the part of adults and, consequently, are not rendered to these the necessary care . This confused and distinct duality reflects the so-called self-conscious community, which is a reflection of some individuals' attitudes, as mentioned by the interactionists . The restriction of the right to speak also occurs with adolescents, although the social commotion in the face of these problems is lower when compared to those experienced by children at an early age, as can be seen in the participants' speeches. This observation is consistent with SI principles regarding the meaning attributed to adolescents in modern society that they are not as vulnerable as children, knowing how and when to defend themselves. Currently, victims of SV are entitled to specific monitoring of professionals prepared to perform qualified listening . Based on SI premises , it is believed that interactions between victims and caregivers can promote fruitful interventions, based on understanding victims' needs. In this study, participants highlight that both healthy-looking children and those with health problems were sexually victimized. A Brazilian study points to a high prevalence of psychological aggression and physical abuse against children and adolescents with disabilities, but does not mention sexual assaults . However, research with 5,917 adolescents and adults with disabilities demonstrates high chances of them suffering severe SV . Analyzing these data, it can be considered that cases of SV may be underreported, since people with disabilities may have difficulty expressing sexual rape or even recognizing it, such as children and adolescents. On the other hand, there is express incompleteness of notification forms of interpersonal violence of children regarding the completion of the field referring to the type of disability/ disorder ported . Thus, the precarious provision of data ratifies the existence of obstacles that hinder the identification of cases and their records and, therefore, may be compromising the necessary care to victims. Given that authorities are committed to expanding the range of strategies that enable health care and the defense of victims of SV, such as reporting channels, mandatory notification by professionals , guarantee of priority and comprehensive care, as required by public policies 19, , even so, difficulties arise regarding the identification of cases, care planning and implementation of coping actions related to the protection of children and adolescents victims of SV . Considering that behavioral changes can be provided through social interactions, it is necessary to expand the dissemination of SV and actions to protect children and adolescents, especially through sexual education, a fact that has been debated and gives rise to divergent positions in Brazil. Such a position is understandable, given the difficulty of social actors responsible for the care of children and adolescents in addressing the theme, as denounced by scholars . Child and adolescent SV has been worsening , becoming a concern in the current scenario of the new coronavirus pandemic, in which social distancing forces these victims to remain in contact with offenders for longer . Therefore, there is a need for the whole society to know properties that make children and adolescents vulnerable to SV so that they can collaborate with the implementation of public policies for prevention and protection, through the ability to provide information and comprehensive care to vulnerable groups, alerting the whole society. Most reports in this study refer to female victims, supporting scientific findings that assert that girls in any age group suffer more sexual violations than boys . This points to the vulnerability of women to this type of abuse in a patriarchal society, such as Brazil . Despite the creation of legal protection instruments and increased penalties for sexual offenders in the country , such measures have not yet been definitive in solving the problem of SV against women, children and adolescents in Brazil . Perhaps this is because the culture of rape is rooted in popular belief, in which the victim is often held responsible for the offenders' acts. Adolescents were described by the professionals of this study as girls who had a physical type characteristic of the beginning of pubertal development and little emotional maturity. This premise goes against the conception that juvenile victims are attractive, full of voluptuousness , having their bodies represented as objects and, consequently, are able to maintain sexualized relations, being permissible to explore them . This attitude is congruent with the stigmas imposed by society on victims of SV, which consequently make them different from the others. However, such prejudices can change when the other members of this society are under the influence of leaders, as considered by the interactionists . Thus, given all the knowledge accumulated by health professionals, they can take the place of leaders of these changes of understanding about sexual victimization. This occurs because these conceptions collaborate to blame the victims on the violence experienced, mitigating the crime committed by the offenders who sometimes claim the victims' consent to justify their acts 38) . Above all, this occurs as a result of the social representations of SV juvenile victims and the gender inequalities that attribute confidence to men's speech , since the interactionists reflect that the organization of society depends on the replica of attitudes of some citizens, which are followed by all . However, considering the difficulties that scholars and authorities have in delimiting and constructing a concept of consent for sexual acts, mechanisms were created to protect child victims of this type of crime, considering their physical/ emotional immaturity for the act and attributing unrestricted responsibility to the offenders 19,32) . Likewise, the participants of this study highlighted the victims' innocence, an opinion that is enhanced when they are children or when the adolescents appear to be physically fragile and innocent. This invisibility of adolescents, in general, may be related to the naturalization of violence perpetrated against women, a behavior resulting from structural sexism in modern societies 31) . Also, it is related to the premise of possible consent of adolescents for the sexual act, even if they are immature for such permission , i.e., it is related to the naturalization of SV against women, extended to adolescents that, due to its systemic character, can occur discreetly or even imperceptibly. In this regard, according to Mead, human beings are individuals resulting from their interaction with society and who contribute to its construction, thus, the person also reflects what their social environment is . However, it is necessary to sensitize professionals to the need to combat this naturalization. It should be noted that, when a multidisciplinary team acts based on the naturalization of violence, this attitude can interfere with its power to exercise empathic understanding, pointed out by SI as elementary for interactions , influencing the restoration of victims' self-esteem and feeling of security . A study reveals that Portuguese health professionals do not blame SV on adolescent victims' behavior and do not blame them for abusive acts perpetrated by offenders . Therefore, such attitudes can collaborate to favor the establishment of a relationship of professional-patient trust , as it supports the interaction between those involved, especially in the care of children and adolescents, allowing continuity of care, so necessary for the humanized care of victims of SV . In this regard, Mead explains that, depending on the meaning attributed to experiences of violence, individuals feel empathy or antipathy, which justifies behaviors and attitudes towards the situations experienced. As observed in this study, it is feasible to say that conditions attributed to victims, such as innocence and emotional immaturity, when converging for children and adolescents, increase their vulnerability to sexual victimization. Thus, sexual education and health education are ways to provide child and youth empowerment regarding autonomy and recognition of individual rights, providing subsidies for self-protection, such as the attitude of denouncing offenders to family members, health professionals and education . Research confirms that the application of games that simulate reality is favorable to the sexual development of children and adolescents , observing victims of child and adolescent SV with another perspective and handling the informational care concerning the needs of these individuals . This method can be implemented in health education activities, a responsibility that needs to be shared between family members and professionals, although there are difficulties . Also, sexual education can be a field of broad action for nursing, when integrated into health education, with significant individual and collective impact for victims, their families and the whole society, reverberating in the social and health spheres. --- Study limitations The study was conducted in only one public hospital, but can be replicated in another environment. Also, it presents the view only of health professionals, all women, which may have suffered influences related to the social construction of gender and academic training to which they were submitted. However, the results presented allow reflection and foster discussion about childhood and youth SV, alerting professionals to nuances of this type of care, which can be permeated by preconceived social influences. --- Contributions to health This study contributes to alert professionals and family members about aspects of physical and emotional development that increase the vulnerability of children and adolescents to SV, as they are limiting to self-protection. In addition to this, it warns about the need to create strategies that integrate preventive and protective actions for children and adolescents, a favorable field for nursing activities. --- FINAL CONSIDERATIONS In the present study, it was identified that, according to the multidisciplinary health team's perception, some children who were victims of SV were under five years old, some had a healthy aspect, and others had physical/mental disabilities. From the interviewees' perspective, adolescents were in prepubertal development, had bodies similar to children, with fragile body structure, showed no emotional/sexual maturation and were unaware of sexuality. Such characteristics may be responsible for increasing the vulnerability of children and adolescents to SV, and were perceived by the multidisciplinary health team through their experiences and conceptions formed as individuals integrated into society, as proposed by Mead. As a strategy to combat SV against children and adolescents, the use of health education and sexual education is recommended, as both serve as a foundation to empower children and adolescents, providing them with autonomy and recognition of rights. However, such intervention needs to be developed both by family members and by the multidisciplinary team and educators, allowing forms of self-protection to children and adolescents.

Objective: to identify the multidisciplinary health team's perception on aspects of the physical and emotional development of children and adolescents who have suffered sexual violence. Methods: a qualitative study, carried out with 30 health professionals in a hospital in Bahia, Brazil, between June and July 2019. Data were collected from an interview guided by a semi-structured questionnaire. The results were interpreted according to Symbolic Interactionism. Results: some sexually violated children were under five years old and healthy, others had physical/mental disabilities. Victimized adolescents were in the pre-pubertal phase, with fragile body structure, not emotionally mature and unaware of sexuality. Final considerations: children's physical and emotional aspects make children vulnerable to sexual victimization, as they facilitate the control and dominion that offenders need to have over them. As a protection strategy, it is recommended using sexual and health education linked to greater surveillance of family members, multidisciplinary team and educators.

Introduction Human behaviors sometimes show high complexity-person who behaved nicely in one moment may express hostility in another moment, sometimes appearing to be a different person altogether [21]. The accumulated wisdom of social psychology tells us how humans are susceptible to social influences and thereby behave in a context-dependent way. A person's behavior can vary from context to context because one's behavior is influenced by whom one is with . Furthermore, people are generally embedded in multiple networks in their daily lives. Boissevain [5] argues that people participate in several activity fields and each activity field can be regarded as a social subnetwork consisting of a set of people who potentially share a common relationship . Even in a relationship between two people, it is not possible to characterize a person by a single role. Imagine a small social group of people in a given region who are mutually acquainted with each other. There exist several functional subnetworks within the group that have specific roles. Many of the group's members participate in multiple subnetworks and may have different roles within each subnetwork. They are exposed to social influences sourced from different subnetworks and are faced with different obligations stemming from different social roles across these subnetworks. This diversity of social contexts can lead to the complexity of people's behaviors. A relation between people who interact with each other in different positions and roles in different networks is termed multiplex or many-stranded. This kind of overlap is more common in small and isolated societies and communities [5]. To disentangle these complex behaviors, it is essential to have methods that assess the networks embedded within the social group and the roles members have within each subnetwork. In the current study, we measured interpersonal contact between residents in a rural area over a long period of time using electronic devices. We aim to develop a methodology for capturing a comprehensive view of a social group by discovering the multiple subnetworks that exist in a group, based on the results of long-term proximity measurement data matrix factorization. The problem of finding unknown subnetworks of a community has been an age-old struggle faced by many researchers. In anthropological studies, participant observations have been used to comprehensively observe the social networks of small social groups [5,30]. However, such methods may be difficult for a study of a community of a few hundred residents. In that case, the most common survey method today is to have participants report on their own social networks in questionnaires based on recall or recognition [41]. However, the data reported by participants in these types of surveys are heavily biased and inaccurate at both the dyadic and structural levels, when compared to objective observational records, and cannot be a substitute for observational data [3,4,11,28]. To conduct more reliable research, we need a method for objectively observing interpersonal contacts. One commonly used method of objectively assessing social networks, is through Email and SNS communication logs [13,24,26]. However, communication often happens outside such electronic media, and face-to-face verbal and nonverbal communication also plays a strong role in maintaining social relationships. In many instances , electronic communication logs may be insufficient for revealing social networks, and a method of measuring face-to-face interpersonal interactions is needed. One recent solution is to use wearable devices. These allow the assessment of face-to-face interactions with high temporal resolution over long periods of time [8,12,44]. For example, the MIT Reality Mining Project uses a device named "sociometric badges" to record interpersonal interactions between employees in a work environment over time [40]. "Sociometric badges" are devices capable of collecting multi-channel log data, including physical proximity. The strength of personto-person proximity measured by this wearable device has been demonstrated to be linked to the subjective quality of communication [32,42]. Using wearable devices for social network research thus allows for long-term observation, which is also useful for extracting more stable social networks than short-term observation [15]. Social network studies using wearable devices have primarily targeted interpersonal relationships in the work environment, such as within business organizations [32,44]. However, till date, no social network survey using wearable devices has been conducted on elderly people living in rural areas. The potential value of such a wearable device is more likely to be realized in a field where research participants can move freely, rather than in a temporally and spatially confined area . As such, this method was especially appropriate for our study, which tracked the daily interpersonal contact history of participants residing within a rural community. Participants carried a small smartphone from morning to evening, without restrictions on time and area, so that we can collect data of diverse social interactions, not limited to specific types of interactions . Though data obtained from wearable devices can provide clues to reveal the structure of the community, analysis is still a dogmatic premise. Network clustering is a conventional method for revealing the structure of a social group from the records of social ties, by estimating the sub-groups to which a group's members belong from the topography of the social connection data. Various methodologies of network clustering have been proposed [14], that ultimately serve to assign members to one or more sub-groups. However, these network clustering methods are based on cross-sectional data of social ties at a single time-point, and therefore cannot account for temporal changes in connections obtained longitudinally by wearable devices. In this study, we adopted a different approach to extract social networks, which is based on cluster analysis, due to necessity of integrating high resolution data logging of interpersonal interactions. These were measured over a long period of time with wearable devices. Similar to the contact-tracing applications used for COVID-19 management in some countries, the wearable devices model human interaction by communicating wirelessly with other devices within an immediate vicinity and store the information as a proximity log. The log data contain richer social information than general social network data, yielding information, for example, on who was with whom and for how long, and who was present and absent at the same time on a given occasion. To take advantage of these benefits , our study adopts an approach that resembles factor analyses. The subnetworks within a community are considered as latent common factors that cannot be directly observed, and interpersonal contact history is considered as an observation of the latent network. We also consider the social network of the entire community to be the superposition of these latent subnetworks. That is, the problem of finding unknown subnetworks of a community can be replaced by extracting common factors from interpersonal contact history. By applying factor extraction method to the entire interpersonal contact histories of a community, it is possible to find potential common factors, or subnetworks. It is necessary to consider that the factors of these histories are non-negative, i.e., equal to 0, if no contact occurs and positive if there is contact. We, therefore, applied nonnegative matrix factorization to interpersonal contact history in this study. NMF is one of the methods to decompose matrices that have 0 or positive values [9,25], and it has been widely applied to identify structures of various types of data, such as images, documents, genes, as well as time series data of acoustic signals. Thus, NMF is optimal for understanding the structure of social networks from interpersonal contact histories. This study attempts to discover unknown subnetworks within a community by decomposing the interpersonal contact history data into weighted rows of ties and network activity sequences using NMF. The social network created from the weighted row of ties corresponds to a set of participants who are connected to each other at the same time , and the network activity sequence corresponds to the activity levels of the subnetwork. One of the advantages of this method is that the latent networks can be extracted without depending on specific events or locations. We collected proximity logs from wearable devices for 7 months. The study was conducted in a local community in Kyoto Prefecture, Japan. Data of physical proximity among community residents were analyzed by NMF to extract multiple latent subnetworks. To examine the validity of our new methodology, temporal characteristics of these latent subnetworks were examined through comparisons with qualitative data we collected through interviews with the community leader . We also administered a questionnaire survey to the same residents to measure their subjective health and attitude toward the community . Several studies have suggested that pro-community attitude of residents is important for a community as it promotes the community's crime control [35,37] and disaster prevention [16,31]. Through the correlations between these variables and the positions of residents in latent subnetworks , we examined whether our new method can extract relevant indices in predicting important variables in the fields of social science and public health, and which subnetworks are relevant to these variables. Our study site was a local Japanese agricultural community. Though human communities have complex overlapping multiple networks, agricultural communities are generally outstanding in this aspect. One characteristic of such communities is that people work near their homes . This proximity of workplaces and residences results in co-existence of several different types of social networks in a single community, where a resident may participate in multiple networks while playing different social roles across networks. For example, one resident, Mr. A, a rice-crop farmer, would participate in a network for agricultural infrastructure maintenance with other farmers in the community. He may also participate in another network for the neighborhood watch, and perhaps an association for homeowners, and a network for private social gatherings among friends. These different networks can co-exist in the same community in a complex intertwined way. The strength of Mr. A's connection with Ms. B can be different across different networks . As our purpose was to develop a methodology to extract latent networks from a complex accumulation of multiple interconnected networks, a local agricultural community provided an excellent context to test our idea. Another helpful characteristic of a localized agricultural community for our purpose is in its physical distance from other communities. In urbanized areas, communities and groups are generally located close to each other. In local areas, communities are relatively disconnected and traffic between them is less frequent. As a result, social relationships are relatively exclusive of individuals outside the community. Therefore, significant relationships of a resident are generally complete within their community. This completeness of networks is desirable when we try to capture influences of networks on one's life outcomes. If an individual has many important connections with outsiders, investigating the network structures within a community does little to help us with identifying important networks for that individual. --- Methods --- Study site and participants The study site was a local community located in a northern rural area of Kyoto Prefecture, Japan . The geographical area of the community is approximately 5.0 km 2 [36], and the total population of the community is 840, consisting of 318 households [10], for a population density of 168 persons/km 2 . The percentage of elderly people is 30.5% [10]. This community is a part of a farming area, where 11.5% of the households are engaged in farming [29,36]. To illustrate the extent to which this community is sparsely populated and small, we compared it with another community: in downtown Kyoto City , the population density is 14,756 persons/km 2 , the percentage of elderly people is 24.7%, and the percentage of individuals engaging in farming is only 0.06% [36]. Thus, when compared with this urbanized area, our studied community is less populated, more aged, and more devoted to farming. Within the accessible area for the studied community's residents, there is only one elementary school and one junior high school, and no senior high school or university. Therefore, residents had to leave the community at least temporarily if they intended to pursue higher education. As typical of rural Japanese communities, this community has several seasonal events. This includes a New Year's Day celebration at a local shrine, a summer evening festival, an autumn festival, and other community-based rituals . Prior to the study, we conducted two pilot tests to examine the communication infrastructure at the study site. We also checked the functionality of the devices and applications, and their usability for participants . The first pilot test was conducted with 10 participants. Following this, we upgraded the application and created a user manual to improve usability for participants. The second test was conducted with 18 participants. After the two pilot tests , we started a campaign to recruit participants for the main study. Following the advice of the community leader, we distributed flyers that targeted several local groups, such as a sports team, the neighborhood watch, a social group of elderly people, and so on. The community Journal of Computational Social Science 5:1069-1094 leader helped us approach diverse range of groups so that our study would cover a broad range of people in the community. We also recruited participants at a community event where many locals attended. The study was explained to the residents as one that investigated people's behavior and health. Participants would carry two small devices with them and would receive the activity tracker as compensation for their participation when the data collection was completed. They were also informed that health guidance by a social worker and a doctor would be provided to participants who requested it, based on the daily step and sleep data obtained from the activity trackers and a self-report health questionnaire. Those who agreed to participate in the study received the devices, a consent form, the self-report health questionnaire , and the user manual of the devices. The consent form and the health questionnaire were returned to the project team via post. For some participants, a DVD of a short video clip that explained how to use the devices was provided. In total, it took approximately 1 month for a sufficient number of devices to reach the community . We recruited a total of 90 community residents, who received one device pair each. From this pool, we analyzed data from 58 participants, whose log data confirmed that they had been carrying the devices continuously for more than 2 months. Table 1) shows characteristics of the participants. The study was approved by the Institutional Review Board at Kyoto University. All participants gave their informed consent. The participants provided their residential addresses to receive feedback by mail. Such identifying information was, however, accessible only to one project member who could not access the data obtained by wearable devices. --- Study period While the study commenced on November 1, 2017, it took 1 month for a sufficient number of devices to reach the community, so for this analysis, the study period was set from December 1, 2017 to June 30, 2018. --- Procedure --- Devices Each participant carried two devices with them: a wristwatch-type activity tracker and smartphone, of which the latter was the primary device for the current study. The smartphone was equipped with Android 4.2.2 and Bluetooth 4.0 , and the maximum range of Bluetooth communication was 10 m . A custom-made application was installed on this terminal. This application recorded the MAC addresses of nearby Bluetooth devices and the time of detection, and the Bluetooth antenna was refreshed every 5 min. This made it possible to record the MAC addresses of other devices that were nearby at 5-min intervals while the participants were out. With this application, the recorded MAC addresses and time data were uploaded to an online storage once a day via the 3G line. The activity meter data were also uploaded via the 3G network. Participants were instructed to always carry the terminal and instructed to connect the device to a charger when they returned home. The terminal was set to restart automatically when disconnected from the charger. To help the participants use the device properly , we occasionally contacted them. In December 2017 , we contacted participants whose data had not been uploaded for three consecutive days or more, and repeated explanations on how to use the device. In February 2018 , members of the project team visited the community and gave each participant an interim report of daily step data, sleep data, and data of self-report health questionnaire along with health guidance by a social worker and a doctor. On this occasion, maintenance of the devices was carried out when necessary. We also gave the residents a brief lecture on health. Participants who could not meet the project team at this occasion received the interim report via mail. In April 2018, we provided spare devices to the community leader, in case there were some participants who needed them. --- Self-report questionnaire on community-related attitudes In the month of May, 2018 , the participants received a paper-and-pencil questionnaire via mail . The questionnaire included demographic items and two sets of self-report items to measure participants' attitudes, as well as their perceptions on their community life. The first set consisted of items to assess several aspects of a participant's positive attitude toward the community, such as community attachment, cooperative behavior toward the community, and trust toward community members . The items were from a series of large-scale social surveys that some of the current authors had conducted . 1 The second set was designed to measure participants' openness, or attitudes toward new ideas and new people coming from outside of the community . 1 The series of large-scale social surveys were conducted in more than 500 communities mainly located in the western part of Japan. All surveys used paper-and-pencil questionnaires, and most of them were delivered via mail to residents in the target communities. The communities were diverse, including urban, farming, and fishing communities, scattered across different geographical areas of Japan. The surveys were conducted as part of a multiple-purpose project. Therefore, each survey needed to measure many concepts, and hence could not have many items for each one. As such, the project team developed items to measure several different aspects of participants' community life by a small number of items. The items we used in the current study came from the surveys of this project. Some of them were created by modifying items from standardized scales . We chose items to assess several aspects of a participant's positive attitude toward the community and examined their combinations . We took the same approach for the openness measure as well. --- 3 Journal of Computational Social Science 5:1069-1094 These items were also from the same series of large-scale social surveys. For both sets of items, response options were on 5-point scales, with options ranging from 1 to 5 . In addition, the questionnaire also included an item to measure subjective health [23] and happiness [1]. For these two items, response options were on 11-point scales, with options ranging from 0 to 10 . The questionnaire was completed anonymously and then returned to the project team via mail directly or through the community leader . --- Qualitative data After the primary data collection, the project team visited the community to collect qualitative data on community activities that occurred during the study period. We interviewed the community leader, the leader of the aged club, and a community hall staff. --- Results --- Factorization of proximity log data During the survey period, the average number of times that one participant's device detected another's was 9.26 times/day, in which the device scanning was performed every 5 min. In this study, we assumed that when one device detected another device nearby, there was social contact between the owners of those devices. For all combinations of participants, the levels of social contacts every 30 min were scored based on the total number of device detections in the 30 min , resulting in 10,176 epochs for the survey interval . The number of combinations of all the participants were 58 C 2 , as N = 58. Let be I × J social contact matrix, where the element y ij represents the num- ber of social contacts in a combination of two participants at epoch i. We assume that the social contacts reflect the sum of the activities of K latent social networks with different configurations . This study aims to find the latent networks in the target community by decomposing the social contact matrix into the product of the basis matrix representing the time series of network activities and the coefficient matrix corresponding to the levels of connections between two participants: The dimensions of the factorized matrices and are I × K and K × J respectively . The element h ik of I × K basis matrix can be regarded as the activity of the latent network k at epoch i. Thus, the matrix shows the time series changes of the activity levels of the latent networks. The element u kj of K × J coefficient matrix describes the degree of connectivity between each participant in latent network k. All the elements of social contacts matrix are non-negative values, by requirement. Moreover, the basis matrix and the coefficient matrix should consist of non-negative elements, because it is natural to think that latent social networks have additive effects on social contact rather than subtractive effects. Thus, in this study, we utilized non-negative matrix factorization to find latent networks in the community. NMF attempts to find an approximate factorization for ̂ ≃ that minimizes the distance D between ̂ and . In this study, we consider NMF in which the distance D is measured by Euclidean distance between the matrices. The function D EU to be minimized is given by where ∥ ⋅ ∥ F denotes the Frobenius norm, and y ij ≃ ij = ∑ K k=1 h jk u jk is subject to the constraints of h i , u j ≥ 0 , where 0 ≤ i ≤ I, 0 ≤ k ≤ K, 0 ≤ j ≤ J . All compu- tations were done within R using the package NMF [19]. The optimal number of ranks K was determined to be five based on the cophenetic correlation coefficient and the residual sum of squares, as well as interpretability . To elucidate the characteristics of each factor, we examine the basis matrix and the coefficient matrix . Each column of the basis matrix corresponds to a time series of each factor, which represents the activity level of the network associated with each factor . Meanwhile, the coefficient matrix represents the strength of participants' connections with each other associated with a particular factor. We reconstructed each factor and its corresponding network from the coefficient matrix as an undirected graph . Almost all the participants in our surveyed community were acquainted with each other. In evaluating the characteristics of individuals in such a small community, it is necessary to consider the importance of neighboring individuals. Therefore, we used eigenvector centrality to measure centrality [6]. Eigenvector centrality measures the importance of a node by considering the importance of its neighbors. It assigns a relative score to every node in the network based on the assumption that a connection to a highscoring node will contribute more to that node's score than an equivalent connection to a low-scoring node. Centrality score of each participant was calculated for each subnetwork; they were log-transformed to be used in later analyses . The temporal patterns of change in the activity level of each potential ≃ . D EU =∥ - ∥ 2 F = ∑ ij y ij - ij 2 , subnetwork, corresponding to the basis matrix , is shown in Figs. 7 and8. Figure 7 covers the entire study period for all the subnetworks. On the other hand, Fig. 8 shows time activity levels of three subnetworks that had unique patterns for more focused periods of time. High activity levels reflect the high proportion of members of each subnetwork participating in the activity. As shown in Fig. 7, many epochs of factors 1 and 5 have high activity levels, suggesting that these factors are associated with public events. On the other hand, factors 2 and 4 show low activity levels, indicating that they are mainly related to personal contacts. The first factor showed higher levels of activities in April . According to interviews with the community leader, collaborative community activities related to rice farming, such as sowing rice seeds, transplanting rice seedlings, rice field maintenance, and weeding, occurred during the period of high activity in April. This work was mainly performed by members of a farming group in the community, and their activity schedule corresponded to the time series pattern of interpersonal contacts in Factor 1 . This suggests that Factor 1 is a component related to the collaborative work of agriculture. Further, Fig. 6 shows that Factor 1 was linked to the network of relatively older participants, which is consistent with the fact that the members of the farming group were relatively old. Figure 6 also suggests that ties were not equally distributed among participants-some were densely connected while others were not connected in this subnetwork. This consistent pattern was also found from the distribution of centrality scores , which shows a negatively skewed distribution. Thus, in this latent subnetwork, there was a divide among the participants in terms of the degree of connectedness with others. This is consistent with our interpretation that this factor Journal of Computational Social Science 5:1069-1094 is related to agricultural activities , in that activities related to this subnetwork required some residents' cooperation. The second factor appeared to reflect diurnal activity, and we interpret it as representative of daily interactions with family. Figure 6 shows that the network associated with Factor 2 was composed of a small number of combinations of the participants. Similarly, Fig. S1 shows that a large part of participants were located at the median of this distribution, suggesting that many participants were connected with some other participants to the similar degree. These characteristics of this factor are consistent with our interpretation that Factor 2 was related to interpersonal contacts among family members living together. The third factor was also associated with everyday activities . A notable characteristic of this factor is that it had low-level but long-period activities, unlike the spike-like patterns shown in Factor 1. Generally, the activities started in the morning and ceased around the evening , and often showed shortperiod reductions around noon. The level of activities was generally low, suggesting that only a few people were there at one time. Though there were some days that had higher levels of activities , the interview data suggested that community activities occurred at the community hall on these days . These patterns suggest that Factor 3 reflected interpersonal contacts among people who visited the community hall for different reasons. From the interview, the community hall functioned as a gathering point for the residents, with several different kinds of activities taking place in the hall . On weekdays, one staff Fig. 6 The social network corresponding to each factor. Each dot represents a participant. The darker the color, the higher the age All Factors Fig. 7 The time series of social activity levels corresponding to each factor. The bottom graph combines the graphs of all factors. The vertical axis represents the coefficients of the basis matrix . The higher the value, the more contact among the members in the subnetwork at the epoch 1 3 Journal of Computational Social Science 5:1069-1094 member was continually stationed in the hall even without any scheduled special activities, and assisted residents who visited the hall. Therefore, we concluded that the community hall served as "hub" that connected residents directly or indirectly, for various reasons . Given that the residents differed in the frequency of visiting the community hall, the number of chances to interact with other residents at the community hall was also different. The fourth factor was hard to interpret. Higher activity levels were observed on January 8 and June 2, 2018 , but it was unclear what kind of contact these were; we could not discern any related activities from the interviews conducted. Moreover, lower activity levels were observed during the months of April and May. The centrality distribution score was not skewed, unlike Factors 1 and 5. There was no clear peak in the distribution. At this point, we are hesitant to interpret this factor. The subnetwork of Factor 4 shows that there are several Fig. 8 Excerpt from the base matrix , which represents the change in activity level over the month for the networks corresponding to the first, third and fifth factors. The higher the value, the more contacts among the members in the subnetwork at the epoch. Each label indicates the point when various activities took place, as revealed by the interviews. Factor 1 was associated with rice farming activities. Factor 3 was associated with activities at the community hall. Factor 5 was found to be related to various activities including community promotion activities small groups that are not strongly connected to each other. It suggests that there were several gatherings whose activity level decreased due to unknown factors during the same period in April and May. The activity levels of Factor 1 indicate that farming activities were more active during April and May. This agreement implies that farming activities might have suppressed the activities of the subnetwork of Factor 4. The fifth factor is associated with the network which saw involvement from many community members across a wide range of ages , though age was positively associated with centrality score in this factor . Figure S1 suggests that despite the presence of some participants with low centrality scores, the majority of participants had similarly high scores for centrality. High activity levels coincided with community festivals, meetings of local nonprofit groups, bazaars, community gardening activities, and settlement promotion activities . This suggested that Factor 5 was a component closely connected to community promotion activities, which was organized by the community promotion committee and attended by many residents. Another major activity was seen on February 7 in all the five factors . This was the day the project team visited the community to conduct interviews with the participants and provided healthcare information. Table 4 shows descriptive statistics of the centrality scores of the latent networks as well as the self-report scales. Table 5 shows the correlations between them. All the centrality scores of latent networks were positively correlated with each other. Among them, the strongest correlation was found between Factor 1 and Factor 5 , suggesting that residents who were located at the center of the network for farming activities were also located at the center of the network for community promotion activities. --- Reliability of self-report scales The internal consistency of items for measuring positive attitude toward the community is shown in Table 2. We used a principal component analysis to assess the internal consistency of the nine items for measuring pro-community attitude and the five items for openness, separately. For pro-community attitude, two items had low factor loadings and thus were excluded. The remaining seven items showed sufficiently high Cronbach's coefficient alpha and McDonald's coefficient omega . "Pro-community attitude" was computed by averaging seven items. See footnote 2 for validation checks of 2 Two surveys from the aforementioned large-scale project can provide data to examine the validity of this composite measure of our pro-community attitude. One survey collected data from 6409 individuals from 533 communities and used the same six items for pro-community attitude as the current study . The other survey collected data from 1066 individuals from 91 communities and used the same seven items for pro-community attitude as the current study. These two datasets showed that the internal reliability of this composite measure was high . In addition, these two datasets showed that the composite measure of procommunity attitude had reasonable correlations with other variables. For example, the pro-community attitude score was positively correlated with how long a participant had lived in the community . Participation in collective activities in the community 5:1069-1094 this composite measure. The item "I participate in community activities ," which was not included in the pro-community attitude score, was also used in the analyses below as a separate item measuring "participation in community activities." For openness, one item with low factor loading was excluded and the remaining four items showed acceptable Cronbach's coefficient alpha and McDonald's coefficient omega . These were averaged to provide a measure of "openness". See footnote 3 for validation checks of this composite measure. --- Correlations between centrality scores in latent networks and self-report scales Among the self-report scales, pro-community attitude was positively correlated with participation in community activities . Pro-community attitude was also positively correlated with happiness but only marginally. Participation in community activities also had weak positive correlation with openness and subjective health. Openness and subjective health were positively correlated with each other. Finally, subjective health was positively correlated with happiness. Our primary focus was toward understanding the relationship between pro-community attitude and centrality scores of latent networks. Centrality in the Factor 1 network was positively associated with pro-community attitude. Factor 2, which presumably reflected network among family members, did not have any significant correlation with the self-report scales. Centrality in the Factor 3 network was positively associated with subjective health. Centrality in the Factor 4 network, an unknown network, was also positively correlated with pro-community attitude. Centrality in the Factor 5 network did not have any correlation with the selfreport scales. Participation in community activities did not correlate with any centrality scores of latent networks. We will discuss this in "Discussion". 3 The same two surveys can also provide data to examine the validity of this composite measure of openness. Survey 1 used the same four items for openness as the current study. Survey 2 used the same three items for openness as the current study . These two datasets showed that the internal reliability of this composite measure was not necessarily high as in the current study . Yet, the composite measure of openness had reasonable correlations in these datasets. The surveys had two items from a scale of innovation-promotive behavior [38]. The two items were originally designed to measure a behavioral tendency to make proposals for a company . The composite measure of openness was positively correlated with this active behavioral tendency . maintenance work on public facilities, disaster-prevention group activities; see [39]) was also positively correlated with the pro-community attitude score . These findings are supportive of our approach combining the seven items to create a measure of pro-community attitude. --- Footnote 2 --- Discussion With the development of IoT technology, the process of obtaining detailed data on people's spatial proximity over time has become increasingly accessible. Social network surveys using such IoT technologies have higher ecological validity than surveys using paper questions or interviews, because they capture the real day-to-day behavior displayed by survey participants and is free of response and recall biases. However, in communities with complex interpersonal relationships, where multiple social networks overlap, it remains a challenge to discover interpretable social networks from large amounts of long-term, digitally recorded data. This may be partially due to the insufficient use of temporal information in social network studies using wearable devices. In this study, we aimed to extract latent social subnetworks in a local community by factorizing the time series log data matrix of the spatio-temporal proximity using NMF, and to evaluate the interpretability of the extracted latent network. We conducted a 7-month study using wearable devices in a farming community in Japan. This dataset provides rich information on changes of interpersonal contacts-not only micro-scale changes , but also macro-scale changes. Seasonal differences are especially important given that several activities in farming communities are season dependent. In addition, our dataset itself is an important contribution, as it consists of a wide range of age groups in a isolated population, and are generally harder to access for researchers beyond university students or crowdsourced workers who are commonly used in this discipline. We extracted five latent subnetworks from proximity logs. The proximity logs were decomposed into a basis matrix and a weight matrix . We found that the extracted subnetworks showed reasonable and interpretable temporal patterns, suggesting the validity of our method. For example, Factor 1 showed a time series pattern of social contacts that tracked the activities of the farming group. Factor 1 showed high levels of activities around early April, where farming activities are generally busy. At a more fine-grained scale, high levels of activities were observed exactly on days when the farming group's activities were carried out. The other subnetworks had unique characteristics, and some of them showed interpretable patterns . We also found that the extracted subnetworks provided useful information in predicting important variables in the fields of social science and public health . We measured these variables by the selfreport scales and examined which subnetworks are relevant to these variables. As a result, we found that the centrality score of Factor 1 was positively associated with scores of pro-community attitude. That is, individuals at the center of the farming-related social network were more likely to be involved in reciprocal/cooperative relationships in the community than other individuals. This finding is consistent with previous studies showing that farming is connected to several collective activities in communities such as collective works to maintain shared facilities . Interestingly, the farming-related network was not the only one sustaining reciprocal/cooperative relationships in the community. The centrality score of Factor 4 was also associated with pro-community attitude, suggesting that this unknown subnetwork is essential for cooperative relationships in the community. Yet, the time series pattern of this subnetwork did not correspond to the dates of community activity that we learned from the interview with the community leader and others. One interpretation could be that Factor 4 perhaps reflects more casual gatherings of informal groups and random encounters, which may play important roles to maintain cooperative relationships [33]. From another perspective, it may be that a person's high frequency of random contacts with others reflects how deeply their daily activities are intertwined with those of other residents . Either way, this unexpected association suggests that our method helps reveal social networks that are hidden yet play important roles in communities. However, not all the subnetworks were related to pro-community attitude. For example, the centrality score of Factor 2 was not correlated with pro-community attitude. In addition, unexpectedly, the self-report measure of participation in community activities did not correlate with any of the centrality scores including Factor 1 . If the farming group plays central roles for reciprocal/cooperative relationships in the community , those with high centrality in Factor 1 should show the greater tendency to participate in community activities than others. One possibility has to do with a ceiling effect for the item of participation in community activities. The item might assess light commitment to the community rather than more heavy commitment . In fact, the median of this item was relatively high compared to pro-community attitude . If that was the case, even though members of the farming group committed more deeply to participation in community activities than other residents, the current self-report item for participation in community activities failed to capture such a difference. This implies that we need an item asking how deeply one engages in community activities if researchers are interested in individual differences in the involvement in agricultural communities, where participation in community activities is generally high compared to other types of communities . Self-rated health was positively correlated with Factor 3. Factor 3 presumably reflected a wide range of interpersonal contacts occurring at the community hall. The community hall played the role of a "hub" in this community. Different groups in the community visited the hall either regularly, or on an ad-hoc basis. For example, the elderly group had regular meetings at the hall. A group of women regularly gathered at the hall to conduct exercise sessions, and several people gathered there for occasional drinking sessions. The hall might be a place to connect different groups and diverse residents from the community. Then, Factor 3, which traced interpersonal contacts at the community hall, might be a network covering a wider range of social relationships in the community than the other subnetworks. If so, it is understandable that a person who cannot engage in even such a network may have a health issue that prevents them from commuting to the hall. Taken together, the findings suggest that our method can extract a social network that helps us identify individuals who may be of poor health. --- Implications First, the current study showed that spatio-temporal proximity data over a long time recorded by wearable devices is useful to detect meaningful structure of social networks. This is important given that self-report methods of interpersonal contacts have a non-negligible limitation that it cannot capture unnoticed/unmemorable interpersonal contacts [3,4,28]. Such contacts can still be a significant source of social influence on human minds [20] and thus need to be quantified. Using wearable devices for a long period can be a solution for this methodological issue. In fact, the Factor 4 subnetwork, which was seemingly representative of hidden social networks, was linked to pro-community attitude. Second, the current study provides a novel method to extract complex and multilayered social network structures. By factorizing proximity data, we were able to extract multiple latent subnetworks. This method is especially useful when researchers try to unravel complex interwoven ties among community members. People are often embedded in multiple social networks simultaneously , and these networks are somewhat overlapping . By capturing such different types of networks simultaneously, we can examine what kind of networks can be a channel for the transmission of various types of social influences . Third, the current study provides a new perspective on social networks. Traditionally, studies have developed methods to classify individuals into clusters and assess networks among them. In the current study, we proposed a new method to classify ties into latent subnetworks. This approach is based on a perspective to view proximities as observations that reflect latent structures behind of them. By viewing interpersonal contacts in this way, the multi-layered nature of social relationships in the real world can be targeted in empirical investigations. Under this view, it is only natural that individuals are embedded in multiple subnetworks and are sometimes forced to play different social roles across various contexts. As different social roles sometimes place conflicting obligations on an individual, the multi-layered nature of social networks is an important research theme regarding social stress that people face in their daily lives. --- Limitations and future directions First, our sample size was not large, and our efforts to recruit participants did not necessarily cover the entire community. We largely relied on the community leader's direct and indirect connections to recruit participants. This limitation comes from our decision to conduct the study in a farming community, in which social networks likely overlap and are interwoven complexly. As smartphones were not so common in such a community at the time of the survey, we needed to distribute the devices and asked the participants to bring them every day. Given that the devices were not necessarily familiar to the participants, this was not an easy request to accept. Therefore, we had to rely on the community leader, as the most influential person in the community, to recruit participants. Future studies with a larger sample and wider range of participants in a target community are needed. Alternatively, given the difficulty to collect network data over a long period of time, each study can be replicated at a smaller level with a relatively smaller sample size. By accumulating data from such studies, researchers can perform a meta-analysis, which would help overcome any problem caused by a small sample size and help conduct more finegrained analyses as well . If accumulated data come from different types of communities, meta-analysis 1 3 Journal of Computational Social Science 5:1069-1094 would also enable the examination of the generalizability of the findings across different contexts. The current study can be a part of such accumulation. Second, on a related note, we could not ensure that participants brought the devices with them every time they left the house. When not physically with their devices, any encounters with other participants would be missed. One solution could have been to use participants' own smartphones, but with the low rates of smartphone ownership, this was another limitation of our decision to conduct the study in a farming community. For this study, a farming community provided an ideal circumstance for our research question , but future studies in more urbanized settings would be useful to examine the applicability of our method to different settings. Third, like most factor analyses, deciding on the number of factors was a challenge. In our case, the number of factors was determined not only by referring to the cophenetic coefficients and the residual sum of squares, but also by considering the interpretability of qualitative data obtained through the interview. Yet, it is still difficult to narrow down interpretations . In future research, the utilization of GPS location data may provide useful information for determining the appropriate number of factors. This study only measures physical proximity data and does not measure communication via PC or smartphone. Therefore, we were unable to examine whether or not there exists a subnetwork that is based on electronic communication, though the existence of such a subnetwork is possible. In future research, it should be important to examine the differences and interactions between social networks based on electronic communication and networks based on face-to-face communication. Finally, to measure participants' health, we relied on self-reported questionnaires. As there would be an issue of reference group effect [22], the results should be interpreted with caution . Future research should consider using biological markers to assess participants' health. --- Conclusion Human society often comprises several, multi-layered, complex social networks. To understand interpersonal behavior, we must first disentangle such complexity and extract interpretable subnetworks. To this end, the current study proposed a new method using NMF. This method successfully extracted five subnetworks from a 7-month survey of a farming community of Japan, that used wearable devices to track instances of social interaction. The extracted subnetworks helped predict individual differences within the community along the levels of pro-community attitude and health. The study contributes to the literature by adding a new method and a new perspective to comprehend face-to-face social interactions and structures of latent social networks that explain these interactions. --- --- --- --- Consent to participate All the participants provided their informed consent. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. --- Authors and Affiliations Masashi Komori 1 • Kosuke Takemura 2 • Yukihisa Minoura 3 • Atsuhiko Uchida 4 • Rino Iida 4 • Aya Seike 4 • Yukiko Uchida 4

As individuals are susceptible to social influences from those to whom they are connected, structures of social networks have been an important research subject in social sciences. However, quantifying these structures in real life has been comparatively more difficult. One reason is data collection methods-how to assess elusive social contacts (e.g., unintended brief contacts in a coffee room); however, recent studies have overcome this difficulty using wearable devices. Another reason relates to the multi-layered nature of social relations-individuals are often embedded in multiple networks that are overlapping and complicatedly interwoven. A novel method to disentangle such complexity is needed. Here, we propose a new method to detect multiple latent subnetworks behind interpersonal contacts. We collected data of proximities among residents in a Japanese farming community for 7 months using wearable devices which detect other devices nearby via Bluetooth communication. We performed non-negative matrix factorization (NMF) on the proximity log sequences and extracted five latent subnetworks. One of the subnetworks represented social relations regarding farming activities, and another subnetwork captured the patterns of social contacts taking place in a community hall, which played the role of a "hub" of diverse residents within the community. We also found that the eigenvector centrality score in the farming-related network was positively associated with self-reported pro-community attitude, while the centrality score regarding the community hall was associated with increased self-reported health.

Introduction Education contributes to the development of an individual's analytical mind and reasoning ability, which aids in the formation of confidence, self-esteem, and self-respect . In today's society, an uneducated child is disadvantaged in terms of income, health, and opportunity. In A complex array of factors contributes to the problem of educational barriers for women. Social norms and cultural practices perpetuate discriminatory attitudes, devaluing the significance of girls' education and reinforcing traditional gender roles. Economic disparities and limited resources force families to prioritize the education of male children, resulting in lower enrollment and higher attrition rates for female students. Inadequate infrastructure, such as a lack of gender-responsive and safe schools, impedes access and contributes to higher absenteeism and dropout rates among female pupils. These barriers to female education have significant repercussions on socio-economic development. Women's limited access to education hinders their ability to make informed decisions regarding their health, resulting in increased maternal and infant mortality rates. In addition, women's lack of educational opportunities perpetuates income disparity and restricts their participation in the formal labor force, hindering economic development. The solution to the issue of barriers to female education requires a comprehensive strategy that addresses the underlying cultural, economic, and infrastructural obstacles. Effective interventions are required to challenge gender norms, promote gender equality, enhance school infrastructure, and provide targeted support for girls and women to access and complete their education. By understanding the complexities and consequences of these barriers, policymakers, educators, and stakeholders can develop evidence-based strategies and policies to create inclusive and equitable education systems that ensure every girl has the opportunity to learn, flourish, and contribute to the socio-economic development of nations. --- Women Empowerment in the Bible Jesus, the great teacher and role model, acknowledged the inherent equality of all people and consistently demonstrated the worth and dignity of women as individuals. Jesus regarded their fellowship, prayers, service, financial support, testimony, and witness as valuable. He thoughtfully appreciated women, taught women, and catered to women . He illustrated this in three ways: First, he addressed women directly in public, which was rare at the time and the woman caught in adultery , He held women fully responsible for their wrongdoing. Jesus broke key conventions in how he saw women and their duties, presenting a new perspective on women's rights in society. --- Economic and Social Benefits of Education According to Paul , education, even at the elementary level, provides living and practical skills and enables one to access critical resources to alleviate poverty. Those in certain regions of the globe who have not been exposed to modernity should be aware that women's education benefits the economy. According to recent World Bank and World Economic Forum reports, women play a crucial economic role as they become more productive citizens through education. Educated women contribute to the labor force's size, quality, and efficiency. They can obtain better-paying occupations, allowing them to support their families with daily necessities, health care, and education . According to the World Bank and the World Economic Forum, investments in female education can yield a growth premium in Gross Domestic Product trends and closing the gender employment disparity can increase per capita income . --- Barriers to Female Education Females encounter numerous obstacles when attempting to acquire an education. Trends in inequality, traditional beliefs, poverty, early marriage, and adolescent pregnancy are examples of these obstacles. Each of these is examined in detail below. --- Inequality Trends Women play a crucial role in every society, shouldering the heaviest burdens while being the least able to appreciate the fruits of their labor. This is evident in the remarks made by former President of the United States of America, Bill Clinton, at the annual meeting of the Clinton Global Initiative. He stated, "Women perform 66% of the world's work and produce 50% of the food yet earn only 10% of the income and own 1% of the property." Whether the issue is enhancing education in the developing world, combating global climate change, or addressing virtually any other challenge, empowering women is essential . This is especially true in African societies, where women are traditionally regarded as subhuman and attributed to inferior intelligence. Girls presently represent 48% of primary school enrollment, while boys represent 52%, as reported by the World Bank. In the last few decades, the gender disparity has narrowed, but girls still account for 55% of all out-of-school children, or 122 girls for every 100 boys out of school. Even greater disparities exist in most developing nations . In Yemen, there are 270 girls for every 100 boys, while in India, there are 426 girls for every 100 boys. Women have been subjected to severe discrimination due to a lack of legal rights and limited independence from their spouses. Despite recent progress, girls face lifelong disadvantages and exclusion in education systems. In 2013, an estimated 31 million primary school-aged females and 32 million lower secondary school-aged girls were out of school . Sub-Saharan Africa has the lowest proportion of countries with gender parity. South and West Asia have the largest gender gap in their out-of-school population, with 80% of their out-of-school girls unlikely to ever attend school compared to only 16% of their outof-school boys. According to present trends, it is anticipated that by 2015, 69% of countries will have achieved gender parity in primary education, and 48% will have achieved gender parity in lower secondary education . In employment, males are typically deemed to earn more than females. According to a study conducted by Bardhan and Klasen in Egypt, university-educated men and women in Egypt earn different amounts. Women earn less money than men. According to them, this disparity was caused by endowment distinctions between the sexes. They argued that, despite the importance of educational opportunity equality, it is insufficient for attaining pay equality . Mati's study on the 'Effects of Household Chores on the Girl Child's Education' in Kenya revealed that 66.7% of respondents agreed with the phrase "a woman's place is in the kitchen," indicating where the value of women and girls is relegated. If a female child is not interested in school, 46.7% of respondents concurred that she should assist with household chores. Despite the level of education, 46.7% of the community surveyed believed that a married woman is more valuable to society than an adolescent. According to Mati's findings, 36.7% favored secondary education as the highest level of education females should attain, while 26.7% favored university education. Mati found that 86.7% of the community supported the notion of girls staying at home instead of attending school to assist with and learn household chores because, regardless of their level of education, they would wind up in the kitchen. Although this mentality is still prevalent in some nations, the international community is trying to eradicate it. Women are encouraged to enroll in and pursue courses from various programs to empower themselves. Currently, women hold administrative positions of prominence and are held in high regard. --- Traditional Beliefs Traditional beliefs are strong feelings, customs, or ways of doing things that have existed for a long time among a specific group of people and have been handed down from generation to generation without being challenged. They consist of ethical or moral beliefs about how people should live and interact with others. Here is an example of a prenatal moral conviction held by Lindsey regarding boys and girls: Gender roles are acquired throughout infancy and adulthood. The family, schools, institutions, media, tradition, and culture all contribute to reinforcing and discouraging certain behaviors in boys and girls. In a given social context, the gender roles of boys and girls may be flexible or inflexible, similar or distinct, and complementary or contradictory . Consequently, there are gender disparities and gaps that are not simply male-female gaps because they are not based on biology. From the moment, a girl infant is wrapped in a pink blanket and a boy infant in a blue one, gender role development begins. The colors of pink and blue are among the first indicators used by society to distinguish female from male. As this infant grows, other cultural artifacts will assure that this distinction remains intact. Girls will be given dolls to diaper and tiny stoves on which to cook pretend meals. Boys will construct building with miniature tools and wage war with toys and tank. The incredible power of gender socialization is largely responsible for such behaviors. Pink and blue begin this lifelong process . There are gaps between boys and girls due to their distinct roles and social locations. This implies that the relationships between men and women are social and, therefore, not fixed. Bayisenge indicates that most of the time, sociocultural approaches and customs govern the position of girls and women in society. The traditional role of a woman is within the household. Due to the expectation that girls will eventually marry and become housewives; mothers consider what their daughters learn at home to be at least as important as what they learn at school. As girls mature and acquire household skills and behaviors that prepare them for their future roles as wives and mothers, tradition dictates that they remain close to their mothers. These sociocultural perspectives portray girls' formal education as a departure from established communal norms and practices. Formal education is frequently viewed as unsuitable for females and, as a result, is not cherished or valued. Cultural attitudes are so ingrained that even school-aged girls believe they need only to learn how to cook. Some African traditions believe that girls are born for males, and cooking and housekeeping skills are sufficient for women . --- Poverty Girls' educational opportunities are hampered by their poverty status. Children of school-going age who are not enrolled in school are associated with poverty at the national and household levels . In states where fewer children of either gender attend school, shortages of schools and low school quality continue to be significant issues. At the domestic level, the direct cost of educating a child is the most cited reason why children are not sent to school. In many countries, particularly those in Africa, education is ostensibly free, but parents must pay informal fees for compulsory items such as uniforms, books, pens, extra lessons, exam fees, and funds to support school buildings. In other places, the lack of functioning public schools forces parents to send their children to private schools that, despite being technically affordable, are unaffordable for the poorest families, who risk becoming destitute to provide their children with a better life through education. In addition, the impact of starvation on children in education systems as a cause of poverty is grossly under-reported. When a child is severely malnourished, it affects brain development, resulting in slow brain development and poor school performance. By age 5, approximately 171 million children in developing countries are stunted due to malnutrition. Stunting impairs a child's cognitive abilities, concentration, and focus on school. Thus, stunted children are 19% less likely to be able to read by the age of eight . --- Early Marriage and Teenage Pregnancy Early marriage has multiple negative effects on both young women and the society in which they live. It is a violation of human rights and females' rights in particular . Early marriage has profound physical, intellectual, psychological, and emotional effects on both girls and boys, limiting their educational and employment opportunities and chances for personal development . This study emphasizes girls more because this issue affects girls in far greater quantities and with greater intensity. UNICEF contends that girls and society bear the costs of child marriage. As a result of adolescent marriage and pregnancies, society will bear an increasing number of burdens, including population pressure, healthcare costs, and lost opportunities for human development . Early entry into marriage and pregnancy impedes adolescent girls' access to and continuation of education. In many cultures, school attendance is frequently incompatible with the responsibilities and expectations of marriage and motherhood . Economic progress cannot be achieved if the girl child is not given the opportunity to receive an education. Early marriage unquestionably denies school-age children the right to an education that would promote their personal growth, provide the foundation for adulthood, and contribute to the future comfort of their family and community. Women should be granted education because they facilitate and assure the fulfilment of other human rights. For several impoverished, desperately low-income families, the prospective benefits of educating daughters are viewed as distant, and their education is not an investment reward. Families believe that a girl's education will benefit her husband's family, not her own. In addition, some parents believe that girls do not need an education to fulfil their duties as wives and mothers, that education undermines cultural practices, and that it teaches girls to reject tradition. Otoo-Oyortey & Pobi stated in this regard: For over 50 years, governments worldwide have reaffirmed everyone has the right to an education. The Millennium Development Goals were most recently adopted by the 191 United Nations members in 2000. However, according to UNICEF , 62 million girls are among the 115 million children still denied this right. Marriage need not imply that a girl's or boy's education ends, but in many societies, it does because of the views of parents, schools, and spouses. Young wives frequently have older husbands who want them to uphold tradition, stay home, and care for the home and children. Due to customs and the power of their husbands, married women are not permitted to attend school. Even if they were allowed to return, a girl might find it too challenging to balance her responsibilities as a wife and mother with attending school due to the rules, timetables, and physical atmosphere of the school. Girls' sense of security and self-confidence can be further undermined by bullying and harassment from teachers, students, and other parents, leading them to stop attending school . The community and future generations are negatively impacted when girls leave school to be married. The World Bank , reports that the early marriages lead to an economic loss of the society since economic productivity of females who drop out of school to be married is curtailed. Evidence suggests that children of young, uneducated moms are less likely to start their education, perform well in school, or complete their education past the required number of years. Their daughters are more likely to drop out of school, get married early, and start the cycle of poverty again . --- Early Marriage as a Developmental Challenge The Sustainable Development Goals directly oppose early marriage . The first six goals, which are to end extreme poverty and hunger, achieve universal education, advance gender equality and empower women, decrease child mortality, enhance maternal health, and combat HIV/AIDS, malaria, and other diseases, are in danger of being attained . Examining the connection between age at marriage and development reveals that later marriage is necessary for At the age of about 14 years, my father sent me to my uncle so that he could let his wife train me for marriage. He believes that if I continued to go to school, I would be a spoilt girl and no man would agree to marry me. Being spoilt girl meant that I would be too wise to marry back in his village where he could get my dowry . achieving desired development-related goals. Getting married later may help people finish school, get training for a job, and have the knowledge and skills necessary for being a successful adult . Child marriage and poverty are often related in different nations. It disproportionately impacts the population's poorest members and perpetuates poverty cycles. When low-income families decide to have many children while having little money to support them, the poverty trap transforms into a demographic trap. The increased costs that society bears due to early marriage include population pressure, health care expenses, and missed possibilities for human development . Additionally, early marriage undermines international initiatives to alleviate poverty in emerging nations. The widespread practice of child marriage makes it harder for families to escape poverty in the developing world, undermining serious international commitment to combating poverty, HIV/AIDS, and development challenges . One way to combat poverty and developmental issues is through educating girls. With education, females are free to select their futures rather than having them predetermined by their parents and legal guardians. According to Salsavira et al. , adult wages and postponed marriage have a strong correlation. Increased educational attainment is crucial for women's economic future and their capacity to participate in and contribute to the global economy, but this is unattainable for girls who are married off at a young age. Women who marry young are more prone to prioritize surviving and dealing with life's challenges over other aspects of their development, including formal education, preparation for employment, work experience, and personal development . Therefore, communities trying to improve education levels and end the cycle of poverty may find it difficult to overcome early marriage. --- Limitations of Impact In some circumstances, the impact of women's education on development is less significant. In locations with significant poverty levels, the socio-economic impact on families and the advantages of investing in women are much smaller . Additionally, in some instances, women's education is of considerably poorer quality than that of men, which lessens its efficacy. The so-called hidden curriculum in schools, which reinforces particular ideals, can accompany this issue. The emphasis on boys' superiority can lead educated women to forgo marketable training in favor of typically female vocations that pay less poorly, with negative economic and societal repercussions . Investing in sons is more useful to parents in civilizations where women are married off and depart the family while men remain and care for their parents. Additionally, some African society believes investing in a boy's education is more advantageous than a girl's because a boy who leaves school early is more likely to benefit the family than a girl who leaves school at the same grade level . This encourages families who only intend to enroll their kids in primary school to prioritize investing in the education of their sons above that of their daughters. Socially, gender roles may hamper women's education's potential to advance gender equality. This is especially true in cultures where education for women is only valued as a way to produce more desirable wives. While there is little debate on the pure economic advantages, there is some disagreement regarding how to assess the social benefits and some variation between research . --- Method This section describes the procedures of gathering and analysing data for this study that gears towards investigating barriers to female education and its impact on slow socio-economic development of the family. It discusses the methods, designs and procedures that were used to conduct the inquiry. --- Research Design This study utilized a descriptive survey design. According to Mugenda and Mugenda , a descriptive survey design is appropriate for providing an accurate, descriptive analysis of the characteristics of the sample from the population. Consequently, the study design was significant because the data acquired from the large group included a description of their attitudes and actual experiences regarding the impact of the educated girl child on the socio-economic development of the family as perceived by female students and staff. It was anticipated that the design would increase knowledge of barriers to female education and their impact on the socio-economic development of the family. --- Study Population The research population for this study totaled 447 female lecturers and female students combined at Africa International University for the 2016/2017 academic year. Africa International University has females who serve in various roles in the institution. Thus, the population of this study includes presently enrolled female students and female lecturers at Africa International University. The students and lecturers in this study were taken from various countries. Mugenda and Mugenda defined sampling as a process or technique of choosing a sub-group from a population to participate in a study. It is the process of selecting a number of individuals for a study in such a way that the individuals selected represent the large group from which they were selected. The researchers used the published list of students' names from each program from the records office. The researchers gathered a sample size of 19% of the total research population, where 85 out of the 447 were selected. This enabled the researchers to gather sufficient data on the impact of educated females on the development of the family. The stratified sampling method was used to obtain this sample from the study's total population. This was done considering marital status, family size, responsibility, age, and residential status. Based upon the above-described sample selection criterion, the researchers selected 20 participants from the lecturers and 65 from the female students, as indicated in Table 1. --- Sample Size and Sampling Procedure --- Data Collection Questionnaires were administered to the respondents. The researchers provided a consent form which the respondents signed before filing the questionnaire. The questionnaire aimed to enquire the participants' level of education, employment status, their thought on the benefits of female education, and how they can rate the barriers to female education. This consent form demonstrated the respondents' willingness to participate in the research. It also reaffirmed their understanding of the research to be carried out by the researchers, an opportunity to ask questions, and an opportunity to withdraw from the research. --- Results The first part of the research focused on the socio-economic characteristics of the respondents. Tables 2 and3 looked at their education and employment status inclusively, considering that the 20 lecturers were already employed while some students were employed before they left for studies at AIU. 2, all respondents were educated, ranging from senior secondary school level to doctorate level. Notwithstanding, the bachelor level had the highest frequency of respondents , making up 34.1% of the respondents. This level was followed by the masters's level with 24 respondents . Then diploma respondents were 10 . There were 9 respondents with doctorate level degrees representing 10.6%, senior secondary school leaving students were 8, which is 9.4%, and the least was the certificate students, who were 5 forming 5.9% of the respondents. The level of education is crucial because it helps assess the impact of educational attainment on women and determine if there are employment opportunities for educated women. --- The Highest School Grade/Level Completed by Respondents --- Employment Status --- Source: Field Data Table 4 shows the results gathered on the socio-economic benefits of respondents' parents' families because of the education of their girl children. This table is in two folds. First, it examines the extent the respondents make their parents' family enjoy those outlined benefits. If they are, the second aspect rates to what extent they were making it happen. In the first aspect, 95.3% of the 85 respondents acknowledged that they support their families financially by paying their health bills, frequently communicating with them, frequently visiting them, supporting their husbands in caring for the home and having children doing well in school. The other 4.7% said they have not been supportive in the above-mentioned areas. This percentage, however, coincided with those who said they had not been working before. Regarding their male child support, 57.6 of respondents indicate that they have children who have completed schooling and are supporting them, while 42.4% reported having children who have completed schooling but are not supporting them. In terms of female child support, 71.8% of respondents indicated that they have female children who have completed schooling and are supporting them, compared to 28.2% of respondents reporting that they have female children who have completed schooling but are not supporting them. --- 33 Concerning the extent of the support for each of the above-mentioned benefits, 4.7% reported not financially supporting their parents, 12.9% indicated slightly supporting, 50.6% moderately supported, and 31.8% strongly supporting their parents financially. With regards to supporting parents in paying medical bills, 4.7% do not support in that direction at all, 17.6% do so slightly, 42.4% moderately, and 35.3% strongly support. In terms of communication, 12.9% slightly had frequent communication with their parents, 47.1% had moderately, and 35.3% strongly stayed in touch with their parents. The benefit of parents being visited frequently, 15.3% did so slightly, while 23.5% moderately and 56.5% strongly. Considering children's performance in school 14.1% have children slightly doing well, 49.4% moderately, while 31.8% have children strongly doing well in school. With regards to support to husbands in running the home, 12.9% support slightly, while 40% support moderately and 42.4% support strongly. In view of support received from children, 42.4% are not supported by their male children though they have completed schooling and are now employed, 38.8% are slightly supported, 17.6% moderately, and 1.2% are strongly supported by their educated boy children. Finally, 28.2% of the respondents do not receive support from female children mainly because they do not have such children, while 11.8% received support but slightly, 31.8% moderately, and 28.2% strongly. Barriers to female education were considered, and the leading barrier was Parental Discrimination. In Table 5, it can be deduced that 63 considered this as a strong impediment to the education of the girl child, 22 considered it as a moderate impediment, while none of the respondents took it as a slight of non-impediment at all. In addition, 59 of the respondents considered early marriage pressures moderate, while 0 considered it either slight or not an issue at all. Poverty is considered by 62 as a very strong barrier to girl child education, while 23 considered it as moderate. None of the respondents considered it as either not a barrier or a slight barrier. "In availability of school Facilities," no respondent considered it a barrier, while only 1.2% considered it a slight barrier. Others considered it as a moderate , and strong barrier. In response to Tradition -Female Genital Mutilation as a barrier to female education, 61 indicated it is a strong barrier, while 24 considered it as moderate. However, 0 saw it as either not a barrier at all or as a slight barrier. Finally, 62 indicated that "Teacher discrimination" is a strong barrier, while 23 stated it as a moderate barrier. Zero percent stated it as either not a barrier at all or a slight barrier. --- Conclusion The purpose of the study was to investigate the impact of the educated girl child on the socio-economic development of families at Africa International University, Nairobi, Kenya. The education of the girl child is not considered significant in the African family setting. However, the girl child can become more supportive of the socio-economic development of the family. Therefore, this study investigated barriers to female education and its impact on the family's slow socio-economic development, especially in African societies. Parental discrimination, early marriage pressure, poverty, unavailability of school facilities, tradition, Female Genital Mutilation and teacher discrimination were the major barriers identified. It was established that there is a direct correlation

Despite a global effort to ensure equal access to education, girls and women face difficulties accessing quality education, especially in African countries. Economic disparities, cultural norms, discriminatory attitudes, and early marriages contribute to the inability of females to access education equally as their male counterparts. Human Capital Theory holds that humans can increase their productivity with education and skills. The impact of the barriers to education among the girl-child goes beyond individual well-being because it negatively affects income growth and leads to the low socio-economic status of the family. There is an urgency to address the barriers to female education to promote sustainable economic development at the family and national levels. This study adopted a descriptive survey design. The target population was 447 female lecturers and students. Out of which, 51 were female lectures, and 396 were female students at Africa International University (AIU). With the stratified sampling technique, the sample size used in the study was 85 respondents, comprised of 20 lecturers and 65 students, which is 19% of the total population. Primary data was collected using an open-ended and closed questionnaire. The study revealed that 95.3% of the respondents are employed, and they support their families financially. The study further indicates that 74.1% believe parental discrimination is a key barrier to female education, while 72.9% believe poverty is a barrier to female education. Therefore, the study recommends a need for parental and community sensitization on eliminating barriers to female education.

Introduction Since the first case of the novel coronavirus was detected in Wuhan, the capital city of Hubei province in China, on December 31, 2019, the outbreak quickly became a global crisis and greatly impacted many people's lives. Under the massive quarantines instituted worldwide due to the rapid spread of the virus, social media platforms have become one of the most essential information channels to the world and between users for all-around, realtime, non-physical communication. There are 3.5 billion active social media users, equal to approximately 45% of the global population, and visiting social networking websites is one the most popular Internet activities with the highest user engagement . The exponential growth of social media use elevated the presence of public-centered media environments . Social media platforms can thus be the best places to learn about people's interests and concerns about the outbreak of the new epidemic because information sharing and the diffusion of information about the novel coronavirus occurs simultaneously within social media networks. Among the popular social media platforms, Twitter facilitates interactions between online users, and with minimal access restrictions to the platform, any users or accounts can become opinion leaders or influential by functioning as an information or communication hub in the network . Twitter has become a "model organism" for research due to its extensive use and relatively open data policy . However, existing studies have gaps in analyzing information sharing about the new coronavirus on Twitter. For example, in the COVID-19 Twitter networks, some studies found a large number of non-or less credible information sources, including uniform resource locators , while other studies reported evidence of a distinguished presence of public authorities and experts instead of low-quality information promoted by retweet activities on Twitter . Moreover, a great amount of fake news and misinformation circulated regarding this new epidemic . Alex Hern, technology editor of The Guardian, said that "Twitter has become a hotbed of inaccurate and dangerous advice, while others -such as Pinterest and WeChat -restrict users' ability to communicate about the outbreak at all" . Including the 2009 H1N1 pandemic, Twitter has often been used for public health surveillance and crisis management . Despite the frequent application of Twitter as PHS due to its popularity and accessibility, there is little research viewing Twitter as an information channel during a pandemic, despite its heavy use for information sharing in disaster settings. Drawing from the concept of PHS, we coined the term "public health information surveillance" and defined a Twitter network as an information channel functioning as a PHIS application. Most COVID-19 studies, including those cited above that yielded different results, analyzed a dataset without considering the deepened circumstances of the epidemic. In this study, we examined the phases of the epidemic via three chronologically different Twitter networks: the local epidemic phase in January 2020, the global epidemic phase in February 2020, and the pandemic phase in March 2020. Therefore, by investigating the new coronavirus Twitter networks as information channels performing as PHIS during a global public health crisis and further examining the conflicting findings from the early COVID-19 research studies, we examined the characteristics of information sharing among three Twitter networks with temporal gaps. This stagebased approach was particularly imperative because it uncovered new findings from those reported in existing COVID-19 research. The policy suggestions based on our findings can facilitate public health policy and public health information policy. This study also contributes to the applicability of Twitter as PHIS. --- Theoretical framework The development of information communication technology has enabled Internet-based disease surveillance that uses digital data to "nowcast" and forecast an epidemic . According to the World Health Organization , PHS is "the continuous, systematic collection, analysis, and interpretation of health-related data needed for the planning, implementation, and evaluation of public health practice" . PHS can help researchers discover trends in disease prevalence and emerging public health issues and determine intervention points. PHS has undergone many developments over the years, as shown in Fig. 1 . However, Google Flu Trends failed to predict the H1N1 pandemic in 2009 and overestimated the number of flu cases during the 2012-2013 flu season, predicting more than double the number of cases reported by the Center for Disease Control and Prevention . This provoked criticism from scholars, which resulted in removal of the site in 2015 . Although the multiple erroneous predictions of Google Flu Trends sparked awareness of the biases ingrained in digital surveillance, Internet-based search tools and social media have expanded the scope of PHS by providing real-time data. Since the 2009 H1N1 pandemic, public health organizations have increasingly been using social media for the distribution of health information, preparation for emergency situations, and communication related to disease management . Social media functions as an effective information carrier for facilitating quick communication with the public in addition to offering possible benefits for infectious disease management and surveillance . Aiello and his colleagues claimed that the components and characteristics Fig. 1 Major events in digital PHS of social media users' posts can offer researchers opportunities to extract further information through natural-language processing and image analysis. The most popular PHS application of Twitter data includes modeling flu infection rates through tweet frequencies . Moreover, the geolocational information of tweets allows researchers to model the spread of diseases through human geographic mobility, possibly allowing their models to achieve greater precision . While Twitter has been by far the most-used social media site in the context of PHS since 2006, there is a lack of research on its use as an information channel for the global community during a pandemic. Moreover, given the urgency of efforts to stem the COVID-19 pandemic, the international community has been striving to find ways to enhance PHS. We introduce the concept of PHIS while responding to the urgency and filling the gap in the PHS research. Examining the nature and type of information distributed and shared during the epidemic can be as important as controlling the spread of the virus because disinformation or misinformation could exacerbate negative consequences, particularly in the digital age . Twitter, as an information carrier, allows its users to include information that motivates them to seek and share content from diverse information sources. The literature has primarily discussed three types of information carriers: channels, sources, and messages. Channels have been described as "an information transmission system" or "the means by which the message gets from the source to the receiver" . They are often cast as restraints, as in the pipeline metaphor; that is, they keep messages within themselves . A channel is the largest aggregate among the concept of information carriers, which is the stem of an individual's decision tree regarding the trajectory of information-seeking and sharing . In this study, we viewed Twitter as an information channel. A source produces a specific node or venue for information. "A source is an individual or an institution that originates a message" or "[a] work, etc., supplying information or evidence as to some fact, event, or series of these. Also, a person supplying information, an informant, a spokesman" . Information sources embedded in Twitter may include news articles, scholarly content, personal content, links to websites, and images that share similar or different amounts of reliability, credibility, attributes, trustworthiness, or dynamics . A message is composed of words, symbols, or indications that convey a certain type of content emerging from a certain source within a certain channel . Berlo defined messages as "the expression of ideas , expressed in a particular way , through the use of code" . Therefore, we expected the content of the sources embedded in the tweets sampled in this study to contain messages that carried specific implications regarding the novel coronavirus outbreak. Social media content generally includes many sentiment-incorporated sentences. Analyzing sentiments requires mining attitudes and emotions, such as happiness, sadness, or anger, which lie in the realm of personal impressions rather than facts, in addition to reflecting an individual's perspective on a specific topic . On Twitter, sentiment analysis research has been popular and was largely conducted at the word level due to Twitter's strict wordlimit policy. In addition, we examined scientific information sources and messages shared on Twitter via Altmetric measurements because the attributes of scientific sources and messages shared on Twitter could provide helpful insights on information sharing when the deadly virus was rapidly expanding worldwide in early 2020. Altmetrics aggregates the impact of scientific research studies included in social and mass media, which supplement the impact factor score . Therefore, by taking a stage-based approach, we aimed to examine information sharing on Twitter related to the novel coronavirus outbreak by applying the theoretical framework of information carriers . According to Johnson and Case , "Information carriers are the primary repositories of information available to individuals within their information fields" . In this study, we defined Twitter as an information channel for the novel coronavirus outbreak and defined embedded media, influencers, influential tweets, scholarly content, and popular websites in the Twitter networks as information sources in the channel. We defined information, ideas, emotions, signals, and implications as messages included in the information sources. This study aimed to answer the four research questions below: --- Data collection Twitter data were collected on January 24 , February 14 , and March 14 . Each dataset was retrieved with an application programming interface using the data import function of NodeXL, a type of social network analysis and visualization software . We restricted the datasets to a maximum of around 18,000 tweets per retrieval and examined a total of 74,466 relationships on Twitter. In the Twitter network retrieved on January 24, 2020 , 18,075 vertices created tweets including "coronavirus" or replied to or mentioned tweets containing the term. In the Twitter network retrieved on February 14, 2020 , 17,720 vertices were created using tweets that included the term "coronavirus" or were a reply to or mention of tweets containing the term. In the Twitter network retrieved on March 13, 2020 , 23,889 vertices were created tweets that included the term "coronavirus" or were a reply to or mention of tweets containing the term. Extra tweets included in the datasets were retrieved from prior periods, generally one week to about ten days earlier, which enlarged the complete sample period covered by the data. We considered the chosen dates as important for several reasons. The WHO declared a Public Health Emergency of International Concern on January 30, 2020. Thus, the Twitter network retrieved on January 24 can provide insight into the characteristics of information sharing on Twitter before COVID-19 became a global concern. On February 11, 2020, Situation Report 22 published by the WHO reported 43,103 confirmed cases worldwide, and 99% of the cases came from China. On the same day, the WHO officially named the novel coronavirus "COVID-19," an amalgamation of "coronavirus," "disease," and "2019" . Thus, the Twitter network retrieved on February 14 can provide insight into the characteristics of information sharing on Twitter after COVID-19 became viewed as a threat to global public health. On March 11, 2020, a total of 118,319 incidents had been reported globally from 115 nations, and the WHO declared it a pandemic. Therefore, the Twitter network retrieved on March 13, 2020 can provide insight into the characteristics of information sharing among Twitter users shortly after COVID-19 was declared a pandemic in which a novel virus spread globally due to the public's lack of immunity . To obtain scientific information sources and messages for the periods examined in the study, we retrieved Altmetric data from altmetric.com. We retrieved scholarly journal articles that were shared on Twitter via a PubMed search query associated with public health sciences research at altmetric.com on April 5, 2020. We applied the four keywords "COVID-19," "Coronavirus," "Corona virus," and "2019-nCov" for the three data-collection points to match the articles with non-scientific information sources and messages. --- Data analysis To answer the proposed research questions, we examined and compared differences and similarities between the overall attributes of the entire networks, characteristics of the embedded sources, and features of the messages included in the information sources in the retrieved Twitter networks. We applied SNA to examine the coronavirus Twitter network as an information channel under the PHIS concept. SNA can illustrate relationships and disclose influencers via centrality measures. Wetherell et al., 1994) explained SNA as follows: Social network analysis conceptualizes social structure as a network with ties connecting members and channeling resources, focuses on the characteristics of ties rather than on the characteristics of the individual members, and views communities as "personal communities," that is, as networks of individual relations that people foster, maintain, and use in the course of their lives. We took a typological approach to evaluate the structural features of each network, and, to identify top influencers of the three networks, we used betweenness centrality, a measure of how frequently a Twitter user is located in the shortest path between two other users and how the user links clusters by bridging gaps in the network . A Twitter account with high betweenness centrality plays a significant role in linking the major groups and many participants; otherwise, many tweets would be fragmented or isolated in the network. Furthermore, high betweenness centrality indicates a high level of influence and connectivity . We performed SNA using NodeXL . Four types of relationships were found on Twitter: retweets, replies-to, mentions, and tweets that were selflooped edges or were neither replied to nor mentioned by any Twitter users. Unique edges were defined as unique relationships on the network. N1, N2, and N3 included 22,086, 25,178, and 12,414 edges, respectively. Each dataset was processed by calculating metrics such as indegree, outdegree, and betweenness centrality for N1, N2, and N3. After removing 8287 duplicated edges, we examined 51,391 unique connections in this study. For network visualization, the Clauset-Newman-Moore algorithm was applied to create clusters. The algorithm calculates the dominant clusters in the network by categorizing vertices into the best-fitting group based on patterns of interconnectedness; this grouping algorithm defines a few main groups and several smaller groups overall . All three networks were illustrated by applying the Harel-Koren Fast Multiscale layout algorithm to the data. The most frequently mentioned URLs, domains, hashtags, words, word pairs, and sentiment in N1, N2, and N3 were calculated. --- Findings • RQ1: What are the characteristics of the Twitter networks as a communication channel regarding the COVID-19 outbreak? People sharing the same information field also share a context that provides the foundation of information for further interaction . The context of this study is the global novel coronavirus outbreak. Twitter is an effective and easy-to-access information channel that most people can join with few restrictions if they have Internet access. To answer the first research question, we analyzed N1, N2, and N3 in terms of communication channels regarding the COVID-19 outbreak while focusing on the major influencers in each network. Figures 2,3 and 4 provide a network visualization and the most frequently mentioned words in the N1, N2, and N3 networks, respectively. The network visualization of N1, N2, and N3 demonstrates the characteristics of "community clusters" . However, under close examination, it is apparent that each network presented typological differences. The inter-group connections are relatively sparse in N1 compared to the other two networks, which implies that the novel coronavirus emerged suddenly as a global topic on Twitter. As the virus has spread internationally, inter-group borders have become blurry. Figure 3 shows a seeming peak in people's focus on the coronavirus as the entire configuration looks like a tight crowd. Figure 4 shows that, while the largest group remains, a few medium-sized groups have gradually dissolved into smaller clusters. This is Fig. 2 Top 20 groups in N1 with the top keywords in each mainly because the coronavirus has become a worldwide pandemic with varying levels of interest to different regions and populations. Some researchers also claimed that specific information sources and topics triggered several conversations while nurturing their own audiences and/or communities. This phenomenon can encourage the appearance of various perspectives on a topic or event depending on its importance or applicability to diverse audiences, uncovering various opinions on a topic among social media users . N1, N2, and N3 largely possess the attributes of "community clusters." For example, all of them had several isolates or self-loop vertices, as in G2 of N1, G1 of N2, and G2 of N3. All three of the networks are composed of multiple medium-or small-sized groups instead of a single dominant, centralized group. Despite the common topic of the COVID-19 outbreak, the Twitter users formed multiple small-and medium-sized conversation groups based on their interests and the relevance of the event with relatively little impact from influencers compared to other conversational archetypes on Twitter, such as polarized crowd networks and broadcast networks . The results indicate that actors with high betweenness centrality carry a large amount of traffic with the shortest paths, which makes them significant actors in linking conversations in a weakly connected network . They facilitate connections between Twitter networks by reducing or removing information discrepancies within the network. Therefore, vertices with high betweenness centrality are often considered "top influencers" in the network compared to those who simply have many followers. As shown in Table 1, international news media outlets, such as @cnn, @ajnews, and @skynews, presented high betweenness centrality in N1. As the primary issue-relevant institution of COVID-19, @WHO was a substantial gap-bridger since the beginning of the epidemic. In a Twitter network, laypersons often achieve high betweenness centrality when they create or retweet tweets that are popular within a specific network. For example, @siwuol is a man who has around 6000 followers but retweeted a tweet that graphically described the prevention tactics for the novel coronavirus, which was retweeted 186,200 times and liked 129,400 times. This single retweet critically positioned him in the N1 network. Despite having few followers, other ordinary users, including @988partick, @old_ plot1996, and @smilenio2 in N1; @ischinar in N2; and @hyejoohobi, @slashtrashqueen, @elkhalifag, @jasminebri_anna, and @miguelrmzcorro in N3, played the role of "information connectors" due to their locational merit in each network. In N2, @WHO was the top influencer in the entire network. Among the other influencers, two-@jenniferatntd, a writer, and @kamalaharris, a U.S. senator-were public figures. Interestingly, @cnn was the only news outlet among the top ten influencers in N2. Twitter.com suspended the accounts of @ischinar, @howroute, and @livecrisisnews while notifying users that "Twitter suspends accounts which violate the Twitter Rules" . As its Twitter handle suggested, @viriyabot, including @spectatorindex, in N2 was a bot account. In N3, after the coronavirus outbreak was declared a pandemic by the WHO, several prominent political leaders emerged as top influencers, including the U.S. president , the prime minister of Spain , and the president of the Republic of Colombia . Table 2 summarizes the top ten Twitter handles in terms of betweenness centrality in N1, N2, and N3 by account type. The bar graph in Fig. 5 presents the composition of the top influencers by their account types. • RQ2: What are the characteristics of the information sources included in the Twitter networks regarding the COVID-19 outbreak? We extracted the top domains and URLs included in the tweets in the N1, N2, and N3 networks for use as representative information sources. Among the top 30 domains presented in Table 3, except for two social media platforms, Twitter.com and YouTube.com, and one government website, whitehouse.gov, all the top domains were news outlets. Theguardian.com, nytimes.com, and reuters.com appeared in all three networks as popular information sources. This indicates that Twitter users were seeking information regarding the COVID-19 outbreak largely through reliable news sources, and the most popular news sources for those who shared information regarding the outbreak were the Internet versions of The Guardian, The New York Times, and Reuters. Due to the global scale of the event, geographically diverse national news outlets were also identified, such as scmp.com , albertonnews.com, and elnactional.cat. Interestingly, breitbart.com, an American news outlet known for its far-right politics and conspiracy theories, was ranked as one of the top information sources in N2. As in Table 3, Twitter users appeared to be largely interested in news stories and information reported by major national and international news outlets. In terms of the URLs, at the beginning of the outbreak, the most frequently shared URLs were those of Chinese websites, such as scmp.com and http:// nbl. whiov. ac. cn . After COVID-19 became a global concern, resulting in severe disinformation and backlash against Asians, thebreitbart.com and NYpost.com, which is an American news outlet known for sensationalism and its conservative bias , was ranked multiple times as one of the most frequently shared information sources in N2. In the N3 network, which was assessed after the COVID-19 outbreak was declared a pandemic and the number of confirmed cases and death toll had risen in the U.S., seven out of ten most frequently shared URLs were U.S.-based websites, such as https:// www. white house. gov/ briefi ngs-state ments/ presi dent-donald-j-trump-mobil izedfull-resou rces-feder al-gover nment-respo nd-coron avirus/, https:// www. nytim es. com/ 2020/ 03/ 14/ techn ology/ coron avirus-purell-wipes-amazon-selle rs. html and https:// www. foxne ws. com/ world/ hindu-group-india-cow-urine-coron avirus. • RQ3: What are the characteristics of the messages included in the information sources in the Twitter networks and tweets regarding the COVID-19 outbreak? We extracted the top words, top word pairs, and top hashtags from the data to examine the characteristics of messages included in the tweets in N1, N2, and N3 . The top words and hashtags presented messages in line with the information sources. For example, the top words, top word pairs, and top hashtags in N1 showed that many tweets in N1 referred to the novel coronavirus as "breaking news" while mentioning its origin, Wuhan, China. When discussing topics similar to those in N1, in N2, Twitter users mentioned public health and death tolls. The official term for the virus, COVID-19, was the top hashtag as well as one of the most frequently mentioned words in N2. As in the cases of the top domains and top URLs, a conservative orientation, reflected in the inclusion of catchphrases like "MAGA" and "americafirst" among the top hashtags, was observed. In N3, many more Twitter users used the term "COVID19" in their tweets and retweets than in N2. N3 also showed increased awareness of the status of the coronavirus outbreak as a pandemic and the suspension of the UEFA Champions League's season . One of the most frequently shared URLs in N3 was a link to a document titled "Mental Health and Psychological Considerations During COVID-19 Outbreak" . The popularity of this information source appeared to mirror the messages implied in some of the top word pairs, such as "human, rights" and "providing, human," as the document consisted of a series of messages that "can be used in communications to support mental and psychosocial well-being in different target groups during the outbreak" . Table 5 illustrates the results of the word-level sentiment analysis of N1, N2, and N3. Overall, negative sentiments were stronger than positive ones in all three networks. N1 displayed the biggest gap, 1.8%, between positive and negative sentiments. This negativity might have been caused by fears and worries about the new coronavirus outbreak, even though the outbreak was not yet a global concern. Although positive sentiment increased during the examination periods, N2 contained the largest number of negative sentiments, 3.5%, as well as the most violent or angry words , as shown in Table 7. The rapid spread of the virus worldwide or the backlash detected in N2 might have caused this reaction. However, as the sentiments in N3 demonstrated, after the COVID-19 outbreak was declared a pandemic, Twitter users used more positive, as well as fewer negative, words than in N1 and N2 with a gap of only 0.6% between the two sentiments. This might have occurred partially because people began to encourage each other after the panic at the beginning of the outbreak. Tables 6 and7 present the top ten positive and negative words and all the angry or violent words in the negative category in N1, N2, and N3, respectively. These largely shared positive and negative words can reveal the trajectory of messages shared among the users in the three networks. For example, while "safe," shared 2,796 times, was the dominant concern, the users might not have taken the coronavirus outbreak event seriously, as indicated by the high use of the word "joke" in N1. As the outbreak incident became a global concern, the users in N2 promoted supportive messages by including words like "support," "positive," and "encourage" in their tweets, and this trend continued to N3 by the sharing of words such as "positive," "support," and "thank" when the virus was declared a pandemic. In terms of negative messages, "kill" and "death" were widely shared in N2, revealing the view of this event as a "crisis." The case as pandemic was highly regarded as a "crisis" and was expressed as an emotionally annoying event, as indicated by the inclusion of the words "fuck or fucking" in the messages in N3. "Racism" appeared in the top 20 words of N2 and N3. However, "virus" and "outbreak" were shared the most in all three networks, suggesting that Twitter functioned as an information sharing channel of the COVID-19 event. • RQ4: What are the characteristics of scientific information sources and messages regarding the COVID-19 outbreak shared on Twitter? Tables 8, 9 and 10 present the research studies that obtained the highest Altmetric Attention Scores from the three data-retrieval points, January 24, 2020; February 14, 2020; and March 13, 2020. The AAS is an indicator of how frequently a research study has been mentioned on several types of online platforms, such as news sites, blogs, Twitter, Facebook, Wikipedia, and LinkedIn . The AAS is a weighted score calculated via a pre-established algorithm, and it represents the attention that a study has attracted online rather than its quality. We collected the top ten research studies, as measured by the AAS, for each Twitter dataretrieval point because we could expect that the AASs of the studies were likely largely determined by the attention they received from Twitter users. As shown in Table 8, many scientific studies were produced and shared frequently on Twitter since the first confirmed incident of the novel coronavirus at the end of December 2019. These studies mainly examined the medical aspects of the newly identified coronavirus and were particularly focused on investigations of the incidents in Wuhan, China, where the virus was first reported . Almost half of the studies that were shared were past studies about coronavirus infections in bats . As shown in Table 9, as the epidemic progressed and became a global event, new findings from scientific studies attracted Twitter users' attention. For example, some of the most widely shared studies were about examining the infection route , developing a possible vaccine for the novel coronavirus , and confirmed incidents in countries besides China, such as the U.S. and Germany . Multiple studies in Table 10 referred to the novel coronavirus as "COVID-19" , and scientific studies that examined the clinical characteristics, including the respiratory syndrome, of the virus and risk factors for deaths resulting from 4,5,and 7) attracted a lot of attention on Twitter as the coronavirus outbreak became a pandemic. While four studies published before the outbreak are included , New England Journal of Medicine , and the Journal of the American Medical Association . Many studies were conducted by joint teams of researchers from medical and academic institutions. --- Conclusions and discussion In this study, we identified the COVID-19 Twitter network as a PHIS application and defined it as an information channel that included information sources containing embedded messages. We reported findings from comparative analyses of three "coronavirus" Twitter networks with temporal gaps by examining the characteristics of scientific and nonscientific information sources and messages. Overall, the three Twitter networks demonstrated the characteristics of "community cluster networks" . This result indicates that diverse small and medium-sized groups discussing coronavirus relevant topics were created as a communication channel on Twitter. In the local epidemic phase in N1, trustworthy information and news sources were identified as the top influencers and top domains and URLs, which confirmed the findings of some previous COVID-19 studies ). On the contrary, during the global epidemic phase in N2, discontinued and suspicious Twitter accounts, such as @ischinar, @howroute, and @livecrisisnews, and bot accounts, such as @viriyabot , @sharonhoole , and @spectatorindex , were examined as top connectors. These bot accounts are automated Twitter accounts controlled by bot software. The scores in the parentheses indicate Botometer scores out of 5, and a higher score implies more bot-like activities . These results corresponded to the findings from some early COVID-19 studies that investigated infodemic on social media during the COVID-19 epidemic . In addition, U.S.-based far-right or conservative news media, such as breitbart.com and NYpost.com, were also widely shared domains in N2. The word-level sentiment was the most negative in N2. The findings may suggest that, when a local epidemic became a global crisis, those who wished to spread manipulative information saw an opportunity to take advantage of the critical circumstances by spreading fake news, malicious information, and hate to certain groups. We suggest that social media companies and governments take prompt actions to prevent an infodemic before a regional epidemic becomes a global crisis. In N3, the public were actively communicated to the top political leaders, including @ realdonaldtrump, @sanchezcastejon, and @ivanduque, while sharing information sources and messages they posted on Twitter. These high-profile international politicians were determined as top influencers of the pandemic network in N3, which included a government website, such as whitehouse.gov, as one of the primary information sources. Carried messages examined through top word pairs, such as "government due" and "human rights," were relevant to governance under the pandemic situation as shown in Table 4. Interestingly, @siwuol, which was neither a popular media outlet nor a political or public figure, presented the highest betweenness centrality in the entire network. In fact, several top influencers who played important roles as information super spreaders and connectors within the networks were ordinary individuals. This finding suggests that any Twitter user can become a significant influencer by linking multiple clusters and boosting information sharing across the network. Interestingly, no celebrities were recognized as top influencers, although celebrities are often identified as top influencers in various Twitter networks due to their popularity and numerous followers . This unusual absence of celebrities from the list of top influencers during the public health crisis could be an avenue for future studies. The characteristics of the information sources analyzed through the top domains and URLs and the implied messages investigated via the top words and sentiments within the Twitter content suggests that the elements of information carriers were coherently interlocked, forming an organism of information. Our study results indicate that public health organizations should make more efforts to communicate with the public during a global epidemic. It seemed troublesome that the WHO was the only public health organization that appeared as one of the top influencers in the N1 and N2 stages, and no such organizations were examined in the pandemic phase. As demonstrated in Fig. 5, bot-like accounts and suspended accounts heavily affected the Twitter network while outperforming other types of influencers in the global epidemic phase. Thus, promoting credible information sources and instantaneous updates by public health agencies, as well as the active presence of political leadership, could alleviate the harms resulting from an infodemic created by fake news, misinformation, disinformation, and bot-made content on Twitter. Based on Park and Chung and the findings of our study, we strongly recommend building a transnational PHIS platform to effectively deal with the pandemic crisis. There is a dearth of research on examining scientific information sharing behavior via social media during a global epidemic crisis though Bornmann et al., attempted to investigate Twitter as social-spatial sensors in disseminating scientific information on certain disease, such as HIV, malaria, and tuberculosis. In this regard, the implication of current research is on a par with recent other studies. This study discovered that the largely shared scientific information was about transmission dynamics and clinical features of the disease and case studies of the new coronavirus patients especially infected in Wuhan, China. These findings suggest that clinical scientists quickly initiated investigations on the new coronavirus to inform the world by diminishing an informational deficit with respect to the new infectious disease. Moreover, the trended scientific information per different stage affirms the use of Twitter as PHIS during the pandemic. The results of this study could help public health organizations as well as the governments design information strategies, which often requires prompt decision-making to manage urgent needs during a pandemic crisis. --- Author contributions All authors contributed to the study conception and design. Material preparation and data collection were performed by Han Woo Park and analysis were performed by Miyoung Chong. The first draft of the manuscript was written by Miyoung Chong and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding The authors did not receive support from any organization for the submitted work. --- --- Conflicts of interest The authors have no relevant financial or non-financial interests to disclose.

In this study, we defined a Twitter network as an information channel that includes information sources containing embedded messages. We conducted stage-based comparative analyses of Twitter networks during three periods: the beginning of the COVID-19 epidemic, the period when the epidemic was becoming a global phenomenon, and the beginning of the pandemic. We also analyzed the characteristics of scientific information sources and content on Twitter during the sample period. At the beginning of the epidemic, Twitter users largely shared trustworthy news information sources about the novel coronavirus. Widely shared scientific information focused on clinical investigations and case studies of the new coronavirus as the disease became a pandemic while non-scientific information sources and messages illustrated the social and political aspects of the global outbreak, often including emotional elements. Multiple suspicious, bot-like Twitter accounts were identified as a great connector of the COVID-19 Twitterverse, particularly in the beginning of the global crisis. Our findings suggest that the information carriers, which are information channels, sources, and messages were coherently interlocked, forming an information organism. The study results can help public health organizations design communication strategies, which often require prompt decision-making to manage urgent needs under the circumstances of an epidemic.

Introduction Feminism indicates a collective movement towards equality in all political, economic and social rights for both men and women. The feminist ideology has a stance that "women should be treated equally and are not to be made inferior to men". Women are taught to be fragile and soft-spoken while men act as their guardians, the feminist ideology eradicates this mind set. It is a set of ideologies and theories, that seek to create a society where equality should come in all aspects for both genders. Feminism has a long history and it mainly focuses on women's struggles and experiences in their daily lives . The ideology of feminism is considered a fashionable academic subject, studied in many disciplines, along with in-depth details of the gender roles and responsibilities and the woman's conditions in society. The Western world connects this ideology with modernphilosophy, psychoanalysis, femininity, motherhood, sexuality, economic and social privileges as well as cultural appropriation . The study explores feminist ideology in perspective of online streaming platforms as the online entertainment industry has come with a major growth. In the past few years, convenient access to internet services has made it possible for everyone to grab the online content. There are numerous online entertainment platforms i.e Amazon, Netflix, YouTube, and Torrent downloads, Zee5 online and many more. The online video content consists of web-series, episode, short films, documentaries, entertaining videos and much more. The online streaming platforms are appealing the attention of the audiences like never before. They are fighting the content of television which was motionless and thriving forecasts of the rise of good quality content . The study focuses on web-series, it is a series of short videos which get released over a certain periods of time in order to tell story to viewers. It is designed like a television daily soap but it has been watched on online platform. The concept of web-series came from the Western countries but now, the trend of watching online content has become popular worldwide. With growing importance of online streaming, it is not only creating opportunities for content producers but also attracting young generation for its colorful and creative online production that not only grab attention of viewers . In Pakistan, the popularity of internet and access to online content emerges by each passing day. The statistics show 61.34 million internet users in Pakistan in 2021 and which has been increased by 11 million from the last year 2020. The social media users in Pakistan have been increased by 9.0 million as it hits 46.0 million in the year 2021. The social media and digital trends in Pakistan are pure instances of the internet usage and the dire demand of online content . From the past few decades, the emergence of online content with its innovative format of storytelling not only grabs the attention of audience but also creates a trend of binge watching across the globe. People have switched from conventional media to new media and online entertainment content sets new standards in respect to diversified and variety of thematic productions in media industry. Web-series are a kind of TV bites or mini episodes that followed a conventional TV episodes tradition and merged it with online streaming entertainment . With a vast growth of digital media platforms, web-series is emerged as an ideal platform for women depiction with a wide range of genres, narratives and content without any censorship. Surprisingly, online media also represents gender roles with same patriarchal approach followed by mainstream electronic media as it shows women in supporting or miserable roles. On the other hand, online media always depicts men in a dominated and superior personality. This conventional depiction of men and women can still be seen on digital media platforms . Supporting the above argument regarding feminism, the study correlates with the concept of individualization process that addresses a procedure towards self-actualization. It has also been observed that individualization is considered as the achievement of self-realization which strengthens the current research work as it highlights the significance of selfrealization among women and also create an opportunity for them to create a fair balance in personal and professional lives . There is a great rise in online feminist activism which supports the issues concerning women and their brutal experiences with violence and harassment on screen and off screen. Social media platforms have always encouraged online debates for the rights of women but online misogyny affects women's rights and shatters their confidence to work in society independently . Social media supports different feminist movements like #Metoo or #NiUnaMenos to accept some degree of feminism as a part of their online appearance . Social-media platforms such as Facebook, What Sapp, and Instagram are undoubtedly prominent communication networks for people worldwide . Exploring Feministic Dogmas in Pakistani Web Series: A Qualitative Content Analysis on "Churails" This is the reason, the present study focuses on the representation of misogynistic attitude of men in the web-series and analyzes how it affects women in their personal and professional lives. The study also observespatriarchal behavior in the representation of gender roles which definitely appears as a misogynistic attitude. This online hate through multiple digital platforms can come with serious consequences against women in their real lives . Digital media has come up with an amazing space and introduced a vast range of online production platforms. Today, Pakistani media is expanding while merging with new media and enforcing the social issues regarding women representation on these digital platforms . The study highlights first ever web-series of Pakistan, which has definitely hit the global platform by representing women issues in Pakistan and promoting the importance of women rights. Being the bench mark for the beginning of a new media production, Asim Abbasi came up with the first ever Zee5exclusive and Zindgi original Pakistani drama web-series "Churails" launched on August 11 th , 2020. The web-series stars the lead actresses named Sara Khan, Zubaida, Jugnu Chaudhary and Batool Jan . The study explores feministic beliefs that follow few categories i.eself-realization, women characterization and misogynistic approach to represent gender roles... Women related problems have always been discussed and portrayed in different ways through dramas, movies and web-series with an aim to provide equal rights to women and to empower them. Some web-series intend to bring drastic changes in women's conditions and positions where women can better understand their self-worth and portray themselves economically and socially stable. --- Objectives  To analyze gender roles through the elements of women characterization and misogynistic approach with the use of dialogues in Pakistani web-series  To explore whether the female characters express their self-worth through the use of dialogues in Pakistani web-series --- Significance of Study The study particularly analyzes the dialogues and the contextual meaning behind the words and identifies the relationship between words and meanings. The research helps to identify the existence of women issues and misogynistic attitude of society towards women in the context of Pakistani society. It also strengthens feminist ideology by highlighting the aspect of individuality in terms of self-realization as it aims to provide women an understanding to stand against violence and misogynistic approach towards them and focuses to empower women by realizing them the importance of self-worth. The research study benefits to Pakistani women to know their rights and make them cognizant of their individuality in society. It also facilitates future researchers in order to explore and investigate women perception regarding self-realization and its effectiveness in developing societies like Pakistan. The current study also shows a great contribution in recognizing the women independence and individuality in respect to personal and professional development as it would not only be beneficial for themselves but for the social and economic conditions of country as well. In a country like Pakistan, it is significant to understand the dire need for women development and gender equality in all sectors to grow globally. --- Review of Literature Liberal feminism has always found an intense interest in the social context and structure of society. It believes to study the real life experiences of women living within the strong patriarchal beliefs in society. In form of liberal feminism, the theory has introduced a variant to move beyond the forms of individualism and concepts that are considered significant in the name of liberalism. Although gender inequality is prevailing in societies and there is a dire need to take steps with a feasible approach to bring the desired outcome with plausible options . The digital era has produced a vast trend of binge watching in the audience and promoted variety of online streaming channels. The internet has seen as convergent service medium, that is increasingly carrying a prime role for additional media. The study highlights that the culture of binge watching is expanding not only in Pakistan but across the globe. Undoubtedly, it is a liquid modern world and continuous technological advancements show that new media has emerged with its multiple platforms to distribute entertaining content online . In relation with the current research study, many articles on web series and YouTube content examines a complex relationship in the entertainment content of a YouTube channel and web-series along with networked publics and politics. The reviewed research papers contribute to understand the emotional communications between users in the comment sections of online platforms. It shows that how users are connected to the online platforms i.e YouTube and other networking sites for web series and engage themselves in public debate and political deliberation . The online streaming platforms are captivating the attention of the audiences and producing high quality content for its users. Online streaming platforms have set new standards in terms of creating fresh and relatable content. Web-series in different countries are representing socially contracture image of women that is covered with all stereotypical norms. The reviewed research also investigates how the content structure of web series continues to impact gender representation on digital platforms . In relevance with the present research, some studies on portrayal of female identity in digital media addresses that emergence of digital media platforms like Netflix, Amazon Prime, and other digital streaming, web-series is also viewed as an ideal platform in highlighting and portraying female stories Internet is comparatively free from censorship and content creators can play with complete independence in all genres and narratives. Like the current study, it also focuses on the feministic ideology in web-series and highlights the representation of feminism along with all stereotypes and stigmas . Another reviewed article on representation of gender roles in media clearly assess the web-series in Indian society's context where media has portrayed a stereotypical image for women and shown a submissive role of women in society. In relevance to the current study, it also highlights the patriarchal structure that is strengthening the misogynistic attitude and misrepresentation of women in new media . On contrary, a reviewed research paper addresses that the perspective of director on representation of gender also affects the narratives massively. The reviewed study also shares the role of web-series in the narrative construction regarding multiple attributes of gender, politics, sex and religion. . Several studies on patriarchal ideologies and Female Depiction in Pakistani Cinema also relate with the current study as films are an astounding entertainment medium and an eccentric storytelling expedient. It is a prominent, relentless and a faultless synchronization of filmic communication. It ratifies also reality straightforwardness or incongruity and total fib. Movies are the further most rampant medium of show business. Many Pakistani movies like Bol, Dukhtar, Bachana and Khudakeliye, address feminist standpoint and patriarchal ideologies. After analyzing the selected films the canvassers find that the theme of the films is very courageous, penetrating and very important in today's situation. The studies indicate that film directors and producers of the selected movies have projected feminism very strappingly and also confronted patriarchal ideologies in our society . According to a reviewed literature, there is contribution to explore qualitative study with Pakistani women to discover how Muslim women studying English in higher education milieus in Pakistan engross with feminist thought. The reviewed study emphasizes to capture the relationship between these women's 'secular' education and their religious social uniqueness as young, urban, middle class working women in a Pakistani higher education context. In particular, the study sets out to explore how Pakistani women at higher education institutions interact with and use advanced' procedures of knowledge's , particularly those subjugated by western frameworks of intellectual thought and reasoning, in the context of their own potentially different social lives and self-identities as Muslim women. Young women academics in accumulation to negotiating with the Western notions of Women's movement also simultaneously challenge the local patriarchal hegemonies and unadventurous religious sermons in their social context . Furthermore, a reviewed research paper on patriarchal appropriation of sacred texts used in movies and films demonstrates how Pakistani patriarchy military exercises sacred texts to construct reassuring illusions for women, this paper uses the theory of Islamic Feminism to unknot the politics of religious elucidation and the rambling influence of Islamic fundamentalism. It notes that the protagonist's innermost conflict is a transgression act and coerces her into distracting socio-religious precincts. The reviewed study states that it is significant to bridge the harbor between Islam and feminism, in order to witness the effectiveness of women's rights movements in the Islamic world . The present study relates with the article focuses on representation of feminism and explains the marginalization of women in Pakistani society. The study shows that the women in Pakistan have been suffering in every walk of their lives. They are not allowed to marry with their consent, they are duty-bound to serve their husbands, and are being trounced in the name of honor and so on. This suggests that men are enthusiastic about their power to subdue women. . Therefore, the research paper has examined and decomposed the predicament of Pakistani women in the novel, Black Bird in a White Cage . Pakistani media is promoting western term of feminism by introducing the first feminism based web series Churails. Four women run an office calling themselves Churails. At the point when one of them evaporates, their investigation drives them to something a lot greater than themselves. The purpose of this study is to evaluate the perceptive of Pakistani youth on feminism, also the effect of feminism on youth, and to explore the digital media content on feminism in Pakistan. An online survey is conducted through social media to which 150 respondents of age 16-30 year old responded. This study highlights that the youngsters of Pakistan are fully aware of the term Feminism and support gender equality. In which the 41% of youth agreed with the concept of gender equality and feminism. This study shows that the wrongly portrayed feminism by Pakistani Media is causing discomfort in gender which will affect the feminist movements in future. This study also shows that web series like Churails promotes vulgarity among the youngsters . The web series has been seen in breaking the stereotypical portrayal of women and the depiction of women characters in online streaming, whereas it has also strengthened the stereotypes by highlighting the hatred attitude of men towards women. In relevance to the current research, many recent studies have discussed the representation of women in the web-series and also portrayed the stereotypical image of women constructed in society. Studies have found out that new digital platforms, particularly online streaming intend to 128 Exploring Feministic Dogmas in Pakistani Web Series: A Qualitative Content Analysis on "Churails" focus on socially constructed gender representation in society and also highlights the theory of feminism to see the perspective of women regarding feminism and how webseries portray feminists. --- Theoretical Framework Liberal feminism emerged from the strong philosophical notions invested in liberty and extended these conceptions to women by promoting the concept of individuality and selfworth. It focuses on equal rights of women and their participation and contributions towards economy. Liberal feminists have identified deeply gendered nature of these conceptions that fails to uphold women's rights in society . Liberal feminism abbreviates women empowerment, equal opportunities and women's rights which clearly relates this theoretical concept with the current study . Liberal feminists also support that the resolution to gender inequality is allowing women to have the equal opportunities with men in the developed society. Because of the gender discrimination in society, women are left out in the process of transformation, therefore they are continuously dropping their status and it is the dire need to ensure the progressive and biasfree environment for both the genders . In relation with the current study, it highlights feminism and the importance of individuality which emphasizes equal rights. The web-series portray different classes of women and highlights the phenomenon that women issues are existing everywhere either upper or lower classes, women are suffering discrimination and males are considered as dominant personalities. The web-series promotes the concept of individuality and shows that women are capable of taking their own decisions and can protect themselves from the dangers of society. The web series sheds light on the unfortunate circumstances that are still present in our society such as domestic violence, early marriages, objectification, physical abuse, rape crimes and mental torture. Domestic violence can also be observed as a major context of online misogyny, which has been represented on the digital media platform that strengthens the misogynistic approach in web-series, Dragiewicz, et al., . Through the lens of liberal feminism, feminist ideology has been categorized in selfrealization, male chauvinism, and women characterization. Self-realization is relative to the theory as the foundation of liberal feminism emphasizes on women's potential and self-worth, making her understands her position in society as an individual. The study focuses on equal roles of women socially, politically and economically. It also highlights the misogynistic approach where male dominance comes with a belief that man is powerful and authoritative and suppresses women's identity in all fields of life. Women subordination in society has become a traditional norm but feminism stands against these conventional and compromised societal norms which do not give rights to women. --- Research Methodology The study is cross sectional and exploratory as it collects data from Pakistani web-series by conducting qualitative content analysis for all ten episodes to analyze the representation of gender roles through interpretation of dialogues. The research applies a complete enumeration which ensures to study every unit regarding gender roles in the web-series. The qualitative content analysis facilitates to interpret the contextual meaning of the dialogues regarding gender roles used in the web-series. The study categorizes the feminist ideology under three types i.e. self-realization, women characterization and misogynistic approach. The qualitative content analysis examines the contextual meanings and interpret the dialogues of all the episodes of web series, and assists to build a foundation of the feministic beliefs through the theoretical approach of liberal feminism. --- Themes Self-realization Realizing one's own individuality and worth along with all potential and abilities to perform in life Women Characterization Concept of the male gender being superior to women and having ease of access for opportunities in society as compared to women who are treated as weak and unworthy, men are accommodated more culturally and socially --- Misogynistic Approach It strengthens the concept of male superiority over women and treats them passively. Misogyny promotes hate against women and dominance in society --- Research Questions RQ1. How do the dialogues represent gender roles in Pakistani web-series? RQ2. How do the female characters express their self-worth through the use of dialogues in the Pakistani web series? --- Findings and Data Analysis The study analyzes the feministic beliefs in terms of three categories i.e. self-realization, women characterization and misogynistic approach. It particularly explores the dialogues of the Pakistani web series "Churails" in order to find out the meaning behind the dialogues that are used by the played characters in series. It also highlights the representation of women in different aspects and portrays the patriarchal approach of society that not only exists in lower class but in elite class as well. The dialogues along with the categories in which they fall are shown below: --- Self-Realization Dialogue 1: Apnay jazbaat samajnay hon toh therapist ki madadlo, apne haqooq seekhnay hon tou NGO k paas jao aur seekhne samjhne k baad jab yeh ehsaas ho k ye sara system he mustakil lay raha hai tou hamare darwazay py dastak dou. English Translation: When you want to understand your emotions, go to the therapist, when you want to know your rights, go to the NGO and after learning and understanding everything when you know that the system is continuously damaging you over, knock on our door. --- Interpretation Sara discloses and clarifies the reality of all the "helpful institutions" that are made for women who are stuck in horrible situations in life and in reality; these institutions serve no purpose for the helpless women. Sara, being from a strong economic and political background, is suffering from mental distortion from her in laws and her husband as he catches him having several affairs. She's narrating all this when being interrogated by the police of how things went down. She is present in the jail cell due to the story cover up for Zubaida and Sheila's kidnapping investigation and wanting to find out about the reality of KK's advertising agency. Her firm belief in equality and realization made her create the safety net that is churails as women deserve help in the time of need. Dialogue 2: Lagao hath, lagao phir dekhna mai kya karti hun --- English Translation: Try to touch me and you will see what I can do --- Interpretation Protesters gathered outside Halal clothing's and start protesting against the churails as they claim they have provoked their women and challenged their masculinity. A man from the protest tries to attack pinky but fails to do so, pinky reacts with anger and warns him not to touch her again or she will retaliate with anger. --- Dialogue 3: Aur dubara agar mujay bachon kay samnay hath lagaya toh I'll kill you! English Translation: And if you ever touch me like that again in front of the kids, I will kill you! --- Interpretation A confrontational scene between Sara and her husband, who forcibly pins her against the car in front of their children, when she decides to leave the house of her in laws and she warns her husband to never do that again in front of her children because she does not want them to see or experience domestic violence in their lives between their parents. Children experience fighting between their parents which adversely effects on their minds. Domestic violence is --- Women Characterization Dialogue 1: Array hum toh buss ak selfie lay rahay thay. Ghar chalnay ko thori bol raha hun? Wasey ap sharmili nae hain. Balkey, sharmili ka ulatt hain. Phassi hui jeans, kya baat hai sath chalna hai? English Translation: We were just taking a selfie. I'm not asking you to come home today. Although, you are not shy, in fact you are opposite of it. Fitted jeans!do you wanna come home with me? --- Interpretation The above dialogue from the web-series "Churails" describes that how men in society mistreats women by eve teasing and commenting them on their dresses. The dialogues has been taken from a scene where Churails are entering in café and watching that a group of men harassing few girls and they continuously distress them by their weird and derogatory comments on their dressing and bodies. It shows that men characterize women public property if they are at public place, and they consider it their right to comment or harass them whenever they want to, just for the sake of their entertainment. For men, a woman can be seen as a sexual object and nothing more. The dialogue of this scene clearly dictates that how male use this privilege of power, dominance, and authority. While watching such indecent and inappropriate language of those men, one of the girls "Zubaida" shows courage and starts beating those men to let them learn a lesson and not to harass women. Dialogue 2: Mai kab say keh raha hun kay yeh teesri wali ki naaf nazar aani chahiye. Yeh kafi naihai. English Translation: I have been saying for so long that her belly button should be shown, it is not enough. --- Interpretation Interpreting the beginning sequence from the web series Churails, where each episode has a snippet which includes metaphors regarding women, the men in this scene, are particularly focusing item girls and asking them to make their bodies visible and considering them as sexualized objects. It is strengthening this aspect of men misogynistic attitude towards women as they do not respect women and suppress them by promoting women as a source of commodity. Dialogue 3: Yeh ap ky liye bheja hai boss nay, garami kum karnay kay liye. English Translation: Boss has sent these girls for you, to entertain you. --- Interpretation The above dialogue has been taken from the scene where Zubaida has been kidnapped by few men. This group of women "Churails" enters the kidnapping place with a trick to save Zubaida. Shams pretends himself as van driver and other girls also pretends as dancing girls in see-through and visible dresses along with a huge amount of alcohol. Shams offers these women to kidnappers by telling him, Boss has sent these women to please them. This dialogue is a clear depiction that women are always used as a sex symbol and this shows that women are just seen as a way of pleasure and amusement by dancing, fulfilling men's sexual desires. --- Misogynistic Approach Dialogue1: Babe, boys will be boys. Come on sab is umar mei ye karte hain larkay, exactly ye nahi but boys are just curious about human body. English Translation: Larkay toh larkay hotay hain. Come on everyone at his age does it, not exactly this but boys are just curious about the human body. --- Interpretation This dialogue is from the scene where Sara is out for grocery shopping with her kids and her elder son harass a woman. Sara feels embarrassment and apologizes that woman. oldest son and while she goes over to apologize to the school's principle, her son gets caught teasing an older woman so when she comes home and asks her husband to take the situation seriously, as Sara does not want her son to be like her husband, he shrugs it off as "boys will be boys" ideology, normalizing his son's action and persists on the idea that he would have done the same as all men do, idealizing such behavior. --- Dialogue 2: Mujay tumhay ijazaat he nai deni chahiye thi English Translation: I should not have given you permission. --- Interpretation The above dialogue has been taken from the scene where Sara and her husband are arguing with each other over the people's reaction towards their "Halal Clothing". This issue got hype when a former client of "Halal Clothing" murdered her own husband after knowing about his infidelity. Sara's husband starts regretting that he should not give her permission for this business as she was running her hidden agency under the name of clothing business. This murder triggers the public and particularly men came on roads and protested against Sara and her group of women. The dialogue shows that a woman cannot take her own decision as she has to look for a man's permission. Dialogue 3: Inho nay humari mardangi kou uksaya hai aur humari auraton ko lalkara hai. English Translation: They have provoked our masculinity and challenged our women --- Interpretation This above dialogue has taken from the scene where the group of aggressive mobs is gathered outside "Halal Clothing" and insulting all women working. The dialogue clearly interprets the attitude of men and their behavior towards women as they do not want their wives and sisters to raise voice for their rights. This group of girls named "Churails" creates awareness among girls and women to stand against violence and raise voice for their own rights. This dialogue from this scene clearly shows the stereotypical mindset of men who are always ready blaming women and always consider women individuality as a question to their masculinity. --- Discussion The present study discusses feministic ideology to analyze the representation of gender roles through self-realization, woman characterization and misogynistic approach. The research study conducts qualitative content analysis to interpret the dialogues used by both the genders. Moreover, liberal feminism supports the study as it addresses individualization which coincides with one of the categories of feministic ideology i.e self-realization. This theory not only supports the concept of individuality among women but also stresses significance of self-worth among women. Exploring Feministic Dogmas in Pakistani Web Series: A Qualitative Content Analysis on "Churails" The theoretical contribution of liberal feminism also facilitates the study in the formulation of research questions regarding self-worth as the current research tries to analyze the portrayal of self-realization among women through web-series. The study elaborates that woman empowerment and equality does not only exist in elite class but lower class does also have this understanding of individuality. The study examines that women have been shown realizing their self-worth and understanding their own potential to make decisions and to get rid of all the sufferings they face against gender discrimination and violent crimes. Complementing to the interpretations of the dialogues, women characterization has also been seen as the significant category where study highlights male thinking patterns and the way they characterize women by judging their clothes, choices and living styles. The web series shows objectification of women in different circumstances and portray women only as an object for men to fantasize about. In relevance to the reviewed study on Indian webseries, it has been observed that the discussions of the current research clearly shows that media portrays a stereotypical image for women and highlights the gender stereotypes society which strengthens the misogynistic attitude of men against women and also misrepresent women by objectifying them as commodity . Web-series shed light on all the social sufferings of women and represent several horrific women-suppressed stories which not only highlights domestic violence, sexual abuse, child abuse, rape cases, abduction but women trafficking as well. Street harassment can also be seen in the web-series, which is considered as an unspeakable harm which underplays victim's experiences . By contributing to the idea of individuality in terms of self-realization, the study discloses that the web-series show a transition of a good woman to a rebellious one due to the terrible suffering they face routinely. In order to maintain a social balance in gender roles, it is a dire need to eliminate the idea of suppressing women and degrading their capabilities. Besides that, misogynistic approach has also been witnessed during interpretation of dialogues and it identifies several male-dominant aspects that occur in a society like Pakistan. The web series contains misogynistic attitudes towards women where men suppress women and consider her a sub-standard creature by showing his strong and dominant personality. Considering the belief of ruling over the lives of women, the episodes of the web series show the domestic violence, child abuse, sexual abuse, work place harassment, and women trafficking not only in lower class but in the people of elite class as well. The study majorly witnesses the physical and sexual abuse in the web series which clearly describes that males' hate against women. The study discuses that gender roles are being portrayed in web-series as they are constructed on societal patterns. Women are always targeted because they are seen as fragile and weak characters, unable to fight back. Men are seen as the decision makers of the society due to the superiority and dominant nature. The society has created these gender images which are not only destructing the roles and responsibilities of both the genders as individuals but as social partners as well. --- Conclusions The study concludes by highlighting the importance of self-realization and the misrepresentation of gender roles in the web-series. The current research shows that gender roles has been portrayed in a stereotypical way where males are seen as misogynists and their misogynistic attitude against women has been shown in almost all of the episodes of web-series. Although, the characters of women emerge as a sex symbol and commodity. . On contrary, the study also explores that web-series put forwards this idea of self-realization which is timely and relevant nowadays. Liberal feminism focuses on individuality, meaning that it makes women realize their self-worth and potential and it also helps them to fight against misogynistic attitudes and stand up for their rights. For a progressive world, both of the genders have to play pivotal role in contributing towards social and economic conditions of the world. The study concludes that misogynistic approach strengthens the concept of male chauvinism in web-series. The web series sheds light on a patriarchal approach of society, where men are considered dominant and superior and women are shown submissive and suppressed and always dependent. Male dominance and privilege in Pakistani society is the deep-rooted ideology that is being imprinted in the minds of men by society. In relevance to the current study, there is a dire need to understand the significance of both the gender roles in society. --- Recommendations Web series has been the new trend of consuming online content globally. The social issues present in the society still need to be addressed, the era of online consumption has created a platform for social awareness more than before. In Pakistan, "Churails" is the first web series that launched in the country which deal with social issues of the women and encourages women to fight for their rights. The study has some recommendations for future research in relevance to this subject.  The study recommends to analyze the women representation through semiotics in order to identify the signs and symbols used for women  It also proposes comparative analysis of Pakistani and Indian web-series in respect to gender inequalities  The study puts forward the idea to evaluate the differentiation between women life styles of elite class and poor class  It also suggests to explore Pakistani media narratives regarding gender roles in society.

Highlighting the feministic approach, the study analyzes gender roles along with three major categories i.e., self-realization, women characterization and misogynistic approach. The study examines first ever Pakistani web-series in order to explore the representation of gender roles through dialogues that is being used to construct an image of socially constructed responsibilities of gender in Pakistani society. By employing liberal feminism as the theoretical framework, the study stresses the idea of individuality and self-realization by providing women equal opportunities in all spheres of life. The concept of self-realization specifically lightens the phenomenon of individuality as it is significant for women to know their value and self-worth in society as an individual. While analyzing the dialogues of web-series, the study conducts qualitative content analysis to interpret the contextual meanings of dialogues which shows that Pakistani web-series represent females as a source of objectification whereas male characters in the series shows hate against females. On contrary, the study also discloses that Pakistani web-series are emphasizing the significance of selfrealization as many dialogues shows that females express their self-worth by raising their voices against crimes and fight for their own rights. .

Introduction Cyberbullying is a new way for young people to express their dissatisfaction, often in a violent manner, as it is becoming easier to gain access to information and communication technologies. It is defined as an offensive, premeditated act performed on a victim who is unable to defend themselves via electronic communication technology on a regular and ongoing basis by an organisation or individual [1]. Globally, the common occurrence of cyber-victimization ranges from 5% to 59% [2,3], and cyber-offenders range from 6% to 46% [3]. A survey conducted by UNICEF on cyberbullying involving more than 170,000 adolescents and young adults in 30 countries in 2019 reported that one third of youth were identified as victims of cyberbullying [4]. Several studies from different countries reported that cyberbullying was found to have a negative impact on mental health such as depression [5][6][7][8], which might lead to suicidal behaviour in both the aggressors and victims [9]. The link between cyberbullying and suicidal behaviour can be explained using the interpersonal theory of suicide [10], which proposed that individuals' suicidal ideation or desire for suicide heightened when experiencing uncontainable feelings of two potent interpersonal suicide risk factors, thwarted belongingness, and perceived burdensomeness. The presence of suicidal desire and the capability for it then leads to lethal or near-lethal suicidal acts [10]. Thwarted belongingness is a sense of loneliness and the absence of a reciprocally caring relationship, whereas perceived burdensomeness refers to the belief that the self is so flawed as to be a liability to others, and affectively laden cognitions of self-hatred. Two studies examined the implications of cyberbullying victimization and suicidal behaviour with depression as the mediator using the interpersonal theory of suicide [11,12], in which cyberbullying victimization was identified as profound interpersonal stress that could lead to adolescent loneliness, social disengagement [13], and low self-esteem [14], all of which are key components in the interpersonal theory of suicide [11,12]. As a result, the increased degree of thwarted belongingness and perception of burdensomeness are linked to an increased risk of suicide among teenagers who have been cyber-victimized when mediated by depression [11,12]. We have not yet found any studies applying the theory to explain the link between cyberbullying perpetrators and suicide. Even though the studies exploring cyberbullying and suicidal behaviour using the interpersonal theory of suicide are limited, a number of studies proved that cyberbullying perpetrators or victims were strongly associated with suicidal behaviour among adolescents [8,9,15,16]. Cyberbullying victims or perpetrators had double the risk of suicidal behaviour than those not involved in cyberbullying [15]. Despite the positive association shown between cyberbullying and suicidal behaviour, cyberbullying does not directly cause suicide [17]. Cyberbullying was associated with psychiatric illnesses such as depression [5][6][7][8], anxiety [18], and stress [19], which may predispose to suicidal behaviour among adolescents [9]. Suicide occurs all over the globe in low-to-high-income countries. Approximately more than 70% of suicide incidences occurred in low-and middle-income countries [20]. Suicide is the fourth leading cause of death among individuals aged 15 to 29 [21] and third leading cause of death among female adolescents aged 15 to 19 years old [21], making it a global public health concern. Suicidal behaviour among adolescents is influenced by multiple factors including individual factors , family factors , cyberbullying and socio-environmental factors [22,23]. Prior studies reported that apart from bullying, suicidal behaviour among adolescents has also been linked to other factors including having perceived academic pressure [23], no close friends [24], gender [25][26][27], poor household income [27][28][29], urban households [30], history of abuse [31][32][33], lack of parental support [34,35], lack of peer support [36,37], history of substance abuse [25,29,33] and unstable family environment [38][39][40]. These studies reported that adolescents who had an unstable family environment [38][39][40], poor household income [27][28][29], history of abuse [31][32][33], and lack of social support from family [34,35] and friends [36,37] had higher odds of suicidal behaviour. Adolescents who had depression were four times more at risk of suicidal behaviour than those who did not have depression [34]. Several research studies on identifying risk factors of suicidal behaviour among adolescents reported that perceived academic pressure [23], being female [25,26] and having a social media account [41] were positively associated with suicidal behaviour among adolescents. To our knowledge, no existing study has yet to observe any association between the type of school and suicidal behaviour among adolescents. However, suicidal behaviour is positively associated with stress among adolescents [33]. This can be linked to the type and level of stress experienced by adolescents attending different types of school. A local study on stress among secondary school adolescents in Kelantan, Malaysia [42], reported that students who attended religious schools were exposed to a higher level of academicrelated stressors, which caused moderate-to-high stress compared with other schools. This was followed by boarding schools and then national schools as the students put high importance on achieving excellent academic performance. Hence, academic stressors may trigger suicidal behaviour [43]. In addition, a local national survey on adolescent health reported that adolescents who lived in urban areas had a higher incidence of suicidal behaviour compared to those who lived in rural areas [30]. The Malaysia National Cybersafe Schools Survey 2013 found that a quarter of school children have frequently encountered online bullying through Facebook, blogs and instant messaging [44]. Recent statistics in Malaysia have shown that 25% of adolescents in Malaysia have been scarred as victims by moderate-to-serious online bullying, whereas as many as 54% have expressed a propensity to be a cyberbully [45]. The COVID-19 pandemic has been connected to mental health difficulties as the consequences of physical distance and stay-at-home orders. According to the 2019 National Health and Morbidity Survey , 424,000 children had mental health issues during the pandemic [46]. The Ministry of Health reported 465 attempted suicide cases between January and June 2020, but statistics from The Befrienders Kuala Lumpur revealed more individuals feeling distress and suicidal in July, August, and September 2020 than the previous three months [47]. Living in the new norm in which most school activities, particularly learning, teaching, and socializing, are conducted entirely online has resulted in adolescents becoming more vulnerable to cyberbullying as they are spending more time online [8], thus increasing the risk of suicidal behaviour. Three out of every ten young people in Malaysia are victims of cyberbullying, which occurs most frequently on social media platforms [4]. According to these statistics, cyberbullying has become a substantial threat to the well-being of Malaysian adolescents, and suitable actions are required to prevent additional harm from online aggression. Research on cyberbullying is relatively new; hence, much is still unknown in Asian countries including Malaysia regarding the consequences of cyberbullying [48,49] on mental health, particularly its link to suicidal behaviour. Thus, the aim of this study is to examine the prevalence of cyberbullying and suicidal behaviour as well as the association between the two phenomena among adolescents in Peninsular Malaysia. --- Materials and Methods --- Study Design and Participants A cross-sectional study was conducted in 11 states involving 17 districts and 24 government secondary schools in Peninsular Malaysia. The subjects were selected via multi-stage stratified cluster sampling. Stratification was performed according to types of school , localities and zones . Simple random sampling was used to select one state from each zone and one district from each state within each stratum. From each district, one school was selected by simple random sampling. In each selected school, all eligible students were included as shown in Figure 1. The inclusion criteria encompassed students aged between 13 and 17 years old who were Malaysian citizens, literate, and must be able to understand the Malay language . Those with any conditions that prevented understanding or completion of the questionnaires such as severe intellectual disability, pervasive developmental disorder and uncontrolled hyperactivity or inattention were excluded from the study. --- Sample Size Calculation The sample size was determined using the single proportion formula [50] using prevalence of 52.2% based on an estimated prevalence of cyberbullying [51], a design fect of two to allow for a clustering effect, a precision of 0.04 and a drop-out rate of 2 Based on this sample size calculation, a total of 1440 respondents were required for study. --- Questionnaire and Measures The questionnaire for data collection included a study proforma, the Malay ver of the Cyberbullying Scale and the Malay version of Patient Health Quest naire-9 . --- Suicidal Behaviour The socio-demographic profile and suicidal behaviour of the respondents were quired using a study proforma. Suicidal behaviour was assessed by asking the follow questions : Suicidal ideation was examined by asking "Have you ever seriously conside trying to commit suicide in the past 12 months?". Suicide planning was examined by asking "Have you ever made a plan in the 12 months about how you would attempt suicide?". Suicide attempt was examined by asking "In the last 12 months, have you tried committing suicide?". Suicidal behaviour was the dependent/outcome variable in this study with the bin outcome of 'yes' and 'no'. Variables assessing suicidal ideation and plan had a bin "yes" or "no" response. Suicidal ideation was defined as any "yes" response to the q tion of "Have you ever seriously considered trying to commit suicide in the pas months?". Suicide plan was defined as any "yes" response to the question of "Have --- Sample Size Calculation The sample size was determined using the single proportion formula [50] using the prevalence of 52.2% based on an estimated prevalence of cyberbullying [51], a design effect of two to allow for a clustering effect, a precision of 0.04 and a drop-out rate of 20%. Based on this sample size calculation, a total of 1440 respondents were required for the study. --- Questionnaire and Measures The questionnaire for data collection included a study proforma, the Malay version of the Cyberbullying Scale and the Malay version of Patient Health Questionnaire-9 . --- Suicidal Behaviour The socio-demographic profile and suicidal behaviour of the respondents were acquired using a study proforma. Suicidal behaviour was assessed by asking the following questions : Suicidal ideation was examined by asking "Have you ever seriously considered trying to commit suicide in the past 12 months?". Suicide planning was examined by asking "Have you ever made a plan in the past 12 months about how you would attempt suicide?". Suicide attempt was examined by asking "In the last 12 months, have you ever tried committing suicide?". Suicidal behaviour was the dependent/outcome variable in this study with the binary outcome of 'yes' and 'no'. Variables assessing suicidal ideation and plan had a binary "yes" or "no" response. Suicidal ideation was defined as any "yes" response to the question of "Have you ever seriously considered trying to commit suicide in the past 12 months?". Suicide plan was defined as any "yes" response to the question of "Have you ever made a plan in the past 12 months about how you would attempt suicide?". For suicide attempt, the answer was divided into five options based on frequency: none, once, two-three times, four-five times, and six and above. For suicide attempt, those who answered 'none' was considered as 'no' response and those who answered once and above were considered as a 'yes' response. Therefore, suicidal behaviour was defined as a "yes" response to any suicidal behaviour item, either ideation or plan or attempt. The responses were categorized as yes or no . --- Cyberbullying Victimization and Perpetrators Cyberbullying was assessed using CBS-M, which consisted of two general questions and fourteen items measured with five Likert-type scales ranging from '0 = never' to '4 = all the time'. The first two questions were objective-based questions asking for the medium used by students to cyberbully others and where individuals were cyberbullied by others. Summing the individual raw scores for items 3 through 16 yielded a total score with higher values indicating more frequent experiences of being a cyberbullying victim [53,54]. CBS-M was validated by [55] among secondary school adolescents in one of the districts in Malaysia. It had good psychometric properties and good internal consistency . Cyberbullying victimization and perpetrators were independent variables of interest in this study. Cyberbullying victimization was defined as those who responded to any item in "Have other children used any of the following items to bully you?' in CBS-M. Cyberbullying perpetrators were defined as those who responded to any item in "Have you used any of the following items to bully other children?' in CBS-M. Respondents who responded to any items in both cyberbullying victimization and perpetrators were classified as cyberbullying victimization perpetrators. The frequency of cyberbullying victimization was categorized into two categories according to the median score of the CBS-M. Respondents who scored ≤ 2 were classified as less frequent and respondents who scored > 2 were classified as more frequent victims of cyberbullying. --- Depression Depression was assessed using the Malay version of PHQ-9. It contained self-report measurements for nine questions based on the nine DSM-IV criteria for the diagnosis of major depression. The nine items in PHQ-9 focused on symptoms experienced during the two weeks prior to answering the questionnaire. Each item came with four answer options , which sums up to a total score between the range of 0 to 27. Respondents with scores of 10 and above were categorized as having depression [56,57]. The Malay version of PHQ-9 used to determine depression in this study was validated with good internal reliability . The Malay version of PHQ-9 had a sensitivity of 87% and a specificity of 82% [56]. Respondents with scores of 10 and above were categorized as having depression [56,57]. --- Sociodemographic, Depression Status and Other Factors Associated with Suicidal Behaviour Other independent variables considered to be potential confounders included the parental marital status, household income, school locality, type of school, age, gender social support from family members, perceived social support from friends, history of abuse , perceived academic pressure, social media account ownership, parents fighting in front of adolescents, and depression. --- Data Collection The data were collected from May until September 2021 using a self-administered anonymous online questionnaire. The schools were temporarily closed from all educational activities during the data-collection period because of the movement control order due to the COVID-19 pandemic. After obtaining consent from parents and students, the online questionnaire was distributed to the selected students through the school counsellors, who were given instructions by the primary researcher to explain the details of the research to both the parents and students. Any student's inquiry regarding the questionnaire was forwarded to the primary researcher for clarification through the respective school counsellors. --- Statistical Analysis Data entry and analysis used IBM SPSS Statistic for Windows, Version 26 [58]. Descriptive statistics for categorical variables were summarized as frequencies and percentages, while each continuous variable was summarized in the form of the mean and standard deviation . The prevalence of cyberbullying and suicidal behaviour among secondary school adolescents in Peninsular Malaysia were presented in the form of frequency and percentage. Simple and multiple binary logistic regression were used to examine the association between independent variables and suicidal behaviour as a binary outcome. The strength of association was estimated using the crude and adjusted odds ratio with a 95% confidence interval . Independent variables associated with the outcome at the p < 0.25 level in the simple logistic regression were included in multiple logistic regression models using the forward stepwise selection method by likelihood-ratio test, the backward stepwise selection method by likelihood-ratio test, and manual selection methods. The significance level was based on p < 0.05. --- Results --- Sociodemographic and Other Characteristics of the Respondents A total of 1290 respondents participated in the study, leading to an overall response rate of 89.6%. The number of girls was twice that of boys. The respondents were predominantly Malay and Muslim . The mean age of the respondents was 14.48 . A majority of the respondents' parents were married and living together , and more than half of the respondents came from low-income households . Most respondents perceived having a happy family and social support from family members and friends . Among the respondents, 91.9% were social media users and regular Internet users . Only 15.7% of the respondents had a history of abuse, and 0.6% of them had a history of substance abuse. On the other hand, 74.6% experienced academic pressure from their family, and 32.7% had shown symptoms of depression . --- Prevalence of Cyberbullying and Suicidal Behaviour among Secondary School Adolescents in Peninsular Malaysia From Table 2, 13.7% of adolescents reported to be cyberbullying victims, and 3.8% were cyberbullying perpetrators. In addition, 2.4% of adolescents reported to be both cyberbullying victims and perpetrators. Table 2 also shows that 17.1% exhibited suicidal behaviour, in which having suicidal thoughts has the highest percentage compared to suicide plan and suicide attempt . --- Medium Used among Cyberbullying Victims and Perpetrators and Frequency of Cyberbullying Victimisation The common medium used among cyberbullying victimization and perpetrators were instant messaging and text messaging . Table 4 shows that the median score of cyberbullying victimization frequency was 2.0 , whereby more than half of the students had a total score of less than or equal to 2 . --- Association between Cyberbullying and Suicidal Behaviour among Adolescents in Peninsular Malaysia In simple logistic regression, cyberbullying victims and perpetrators were significantly associated with suicidal behaviour. Other variables also significantly associated with suicidal behaviour were age, gender, school locality, perceived social support from family members, perceived social support from friends, history of abuse, parents engaged in fight in front of children, frequency of cyberbullying victimization, and depression status. Multiple logistic regression analysis reported that cyberbullying victimization, age, gender, having parents who engaged in fight in front of their children, depression, history of abuse, and being perceived as having social support from family members and friends were significantly associated with suicidal behaviour . Cyberbullying victims were reported to have a higher likelihood of suicidal behaviour than those who were not. Table 5 reported that girls had higher odds of suicidal behaviour than boys. Adolescents who had a history of abuse, whose parents engaged in fights in front of them, who had depression, and who perceived having no social support from family members and friends had a higher likelihood of suicidal behaviour than otherwise. Every one-year increase in age reduced the odds of suicidal behaviour by 19.0% . --- Discussion From this study, we observed that 13.7% of adolescents were victims of cyberbullying and 3.8% of them reported being perpetrators of cyberbullying. Our findings differ from those of prior studies in Malaysia and other countries mainly because of the different questionnaires used, adolescent age groups, and the time when the studies were conducted. The prevalence of cyberbullying victimization among adolescents was 14.6% in South Korea [59], 7.2% in Australia [60], 8.8% in Spain [61] and 29.7% in Hong Kong, Taiwan, and Mainland China [62]. The prevalence of cyberbullying perpetrators among adolescents was 6.3% in South Korea [59], 3.5% in Australia [60], 3.1% in Spain [61], and 16.7% in Hong Kong, Taiwan, and Mainland China [62]. In Malaysia, [51] reported as high as 52.2% of adolescents in the state of Negeri Sembilan being involved in cyberbullying victimization while 31.6% of adolescents were victims of cyberbullying and 20.9% were perpetrators in the state of Penang [63]. The prevalence of adolescents who reported that they were both cyberbullying victims and perpetrators was consistent with previous research in the United States [64] and Sweden [65]. However, it was lower compared to other studies in South Korea [59], China [66] and Iran [67]. Cyberbullying prevalence varies across the states as some states, particularly the more urban ones, may have greater access to technology, thus leading to a higher prevalence of cyberbullying. Our study's prevalence of cyberbullying victimisation and perpetrators fits within the vast range of previous prevalence estimates of 5-72% and 3.1-46.3%, respectively [3,51,[59][60][61][62][63][68][69][70][71][72][73][74], which indicate a serious need to pay more attention to the cyberbullying problem. Our study was conducted during the COVID-19 pandemic and MCO when all schools in Malaysia were temporarily closed; hence, students used online learning to continue their home-based learning, which posed a high risk for cyberbullying due to the increase in Internet and social media usage [75][76][77]. The increased adolescent presence in cyberspaces is consistent with our finding, showing that 91.9% of the adolescents were social media users and regular Internet users, which might have contributed to the prevalence of cyberbullying among adolescents. In addition, the most common medium used by cyberbullying victims and perpetrators in our study were instant messaging and text messages, which were consistent with other studies [15,68,78,79]. Suicide prevalence estimates differ substantially between countries and studies [80]. According to our findings, 17.1% of adolescents reported suicidal behaviour with 11.9% having suicidal thoughts, 10.2% having suicide plans, and 8.4% who had attempted suicide. These findings were similar to those reported in Germany [81], Europe, Asia and the Western Pacific Region [82] and meta-analysis performed among adolescents between 1989 and 2018 [83]. Furthermore, suicide prevalence in our study was lower than that reported in Mozambique [36] and the Americas region [82,84]. However, the suicide prevalence in our study was higher than that reported in earlier Malaysian research [52]. The differences in the study findings could be explained by variances in suicidal behaviour measurement as well as changes in the context of the time and study location. For example, Refs. [82,83] reported an incidence within a span of 365 days in the European and Western Pacific region while [84] reported the occurrence in their lifespan in the United States. We discovered that adolescents who had been victims of cyberbullying were more prone to engage in suicidal behaviour than those who had not been victims of cyberbullying. This discovery was in line with previous research [7][8][9]15,23,[85][86][87]. Cyberbullying victimisation has been shown to have an association with an increased risk of depression and anxiety symptoms, which may predispose to suicidal behaviour [7,8,86,87]. In our research, 32.7% of adolescents reported that they were depressed, and we also discovered a positive association between depression and suicidal behaviour. However, there was no significant association of suicidal behaviour with both cyberbullying perpetrators and the frequency of cyberbullying victimization in our study. Thus, our results were in contrast with other studies showing a positive association of adolescent suicidal behaviour with cyberbullying perpetrators and with the frequency of cyberbullying victimization [9,15,88]. This finding could be explained by the limited number of cyberbullying perpetrators identified in our study . Adolescents with depression were more likely than those without depression to engage in suicidal behaviour. This finding was consistent with other studies [25,33,34,85,89,90]. Suicide risk increased by 47-74% in people with mental illnesses, and 50-65% of suicide occurrences were associated with depression [90]. Academic pressure was recognised as one of the top stressors among secondary school adolescents across various types of schools where they were pressured to maintain a strong academic performance in school [42,91,92]. More than half of the adolescents in our study reported that they were under academic stress. As a result, adolescents are under a great deal of stress, which in certain situations can lead to depression, making a person feel hopeless and helpless, and leading them to consider suicide as the only way to end their suffering [92]. Almost half of the adolescents who responded to our survey came from full-boarding schools, where students were expected to excel academically while being away from home, thus leading to high academic stress [93]. In our study, other factors were also reported to be significantly associated with suicidal behaviour included age, gender, having parents who engaged in fights in front of their children, history of abuse, and perceived social support from family members and friends. We revealed that increasing age protected adolescents from suicidal behaviour. This was substantiated by [24], who found that younger adolescents were at a higher risk of suicidal behaviour. Younger suicide victims were observed to have poor problem-solving skills [90]. Girls were shown to be more likely than boys to engage in suicide behaviour, which was consistent with findings from other studies [25][26][27]94,95]. Disparities in mental and behavioural difficulties may explain some of the differences in suicidal behaviour risk between boys and girls [96]. Adolescents whose parents fought in front of them were more likely to demonstrate suicidal behaviour than those whose parents did not. Having parents who fight in front of their children indicates a potentially unstable family situation, which is a risk factor for suicide behaviour in adolescents [38][39][40]. This result was confirmed by other reports showing a significant association between the family environment with suicidal behaviour among adolescents [38][39][40]97]. We also discovered that adolescents who perceived having no social support from family members or friends were more likely to engage in suicidal behaviour than others. Such a finding was consistent with other studies [24,34,82,94,98]. Adolescents are less likely to discuss their difficulties due to the lack of social support from family and peers, making them feel more distressed [98,99]. Teenagers were under a significant level of stress as a result of the COVID-19 pandemic, such as school cancellation and social isolation [100]. During the pandemic, the lack of social support was recognised as one of the risk factors for suicidality in youth [101,102]. The physical and social separation required to prevent COVID-19 from spreading further might have caused adolescents to feel a loss of support, thus increasing their risk of suicide. In addition, the lack of direct social interactions with peers has increased loneliness and social isolation, both of which can increase the likelihood of suicidal behaviour [102][103][104]. Our research also reported that adolescents who admitted that they had been abused were more likely to engage in suicidal behaviour than those who indicated otherwise. This finding is consistent with previous studies showing a significant association between a history of abuse and suicidal behaviour among adolescents [31][32][33][105][106][107]. Movement restriction caused by the pandemic lockdowns led to a surge in recorded incidences of child abuse, neglect, and exploitation [108,109]. Cases of child sexual assault also increased as a result of the simultaneous lockdown measures [110]. --- Strengths of the Study There are a few strengths to be highlighted in this study. As we included adolescents from all zones, the sample in our study may represent Malay adolescents in Peninsular Malaysia. To our knowledge, this is the second local study looking at the association between cyberbullying and suicidal behaviour among school-going adolescents, and our sample size was larger than the first local study. The study was conducted in Malay, which is the local population's native language, to ensure greater language accessibility and understanding. All study instruments had a Malay version that has been validated with good reliability and validity results. The use of a self-administered questionnaire using web-based mode in our study was cost effective. In addition, the survey could be administered concurrently within a larger scope, meaning less time spent on administration [111]. Using an online and self-administered survey also ensures high willingness of respondents to disclose sensitive information due to the assurance of anonymity [112][113][114], which is vital because our study touches on sensitive issues such as suicidal behaviour, history of abuse, and interpersonal issues. Furthermore, since the data collection took 5 months which occurred within the period of the movement control order due to the COVID-19 pandemic, there was no variation in MCO restriction policies from May to September 2021. Therefore, the variation in responses could be minimized. We were also able to quantify the level of association between the explanatory variable and the outcome variable using multiple logistic regression while controlling for a number of confounders [115]. --- Limitations Several limitations were encountered in this study and should be considered in the interpretation of the findings. A cross-sectional design does not establish a temporal link between the exposure and the outcome [116]. The majority of respondents in our survey were Malays as they make up the largest ethnic group in the Malaysian population. This could limit the applicability of the findings to other ethnic groups. Adolescents with disabilities, particularly those with intellectual disabilities, were not included in our study. Suicidal behaviour among adolescents may be influenced by those having a disability [117]. Therefore, future research involving a larger number of other ethnic groups and adolescents with disability is recommended. In addition, qualitative research on identifying other suicide factors such as cultural factors or socio-cultural changes in the youth environment due to the availability of the Internet and other communication tools such as smartphones or a reaction to the threats and restrictions related to the pandemic is required. Furthermore, the usage of the word 'suicide' in the suicide questions might have drawn students' attention and provoked affirmative answers. Nonetheless, these questions were routinely used to assess suicidal behavior among adolescents in the national local survey in Malaysia [52]. A history of chronic illness and mental illness were self-reported; thus, it could be underor over-reported. As data collection was conducted online and self-reported, we were not able to query or verify the information submitted. Due to this reason, we were also not able to collect data from those living in rural areas with limited internet access, thus preventing them from participating in our study. One possible strategy to improve the validity and reliability of the self-report data is to follow up with the respondents using qualitative methods with the involvement of mental health experts being recommend in future research. --- Conclusions In conclusion, this study contributed to the knowledge on cyberbullying and suicidal behaviour among adolescents in Malaysia, particularly during the COVID-19 pandemic. It revealed an astonishing number pertaining to the prevalence of cyberbullying and suicidal behaviour among adolescents in Peninsular Malaysia. Cyberbullying victims were found to be a risk factor for suicide. Other risk factors included young age, history of abuse, being a girl, living in an unstable family, depression, and the perception of no social support from family and friends. Our findings warrant the need to strengthen public awareness on cyberbullying and its impact on mental health, as well as more effective strategies focusing on the early identification and intervention of adolescents with a higher risk for suicide. --- Data Availability Statement: The data presented in this study are available from the corresponding author on reasonable request. The data are not publicly available due to ethical concerns. ---

Social media engagement has contributed to the rise of cyberbullying, which has recently triggered tragic suicides among adolescents. The objective of this cross-sectional study is to determine the prevalence of cyberbullying, suicidal behaviour, and their association among adolescents in Peninsular Malaysia. The study was conducted among 1290 secondary school adolescents aged between 13 and 17 years old in Peninsular Malaysia using a self-administered and anonymous online questionnaire. Our findings reveal that the prevalence of cyberbullying victimization and perpetrator is 13.7% and 3.8%, respectively. The prevalence of suicidal behaviour among adolescents is 17.1%, in which 11.9% had suicidal thoughts, 10.2% had a suicide plan, and 8.4% had made a suicide attempt. Multiple logistic regression shows that adolescents experiencing cyberbullying victimization is a significant risk factor (p < 0.001) for suicidal behaviour after adjusting for other confounders. An alarming number of adolescents in Peninsular Malaysia found to be involved in cyberbullying and its significant association with suicidal behaviour warrant the need to strengthen current interventions. Since the study was conducted during the COVID-19 pandemic, our findings make an important contribution in reporting current trends in cyberbullying and suicide among adolescents, especially when they are becoming ever-more present in cyberspaces.

Introduction Our thoughts, emotions, and behaviours are influenced by those around us-in conversations at home, in shared activities such as going out for dinner with friends, and while working on projects with colleagues. In fact, humans spend considerable time in social settings [1]. Research finds that the number of others around us is key to understanding social phenomena such as free riding, trust, or helping behaviour. For example, people are less likely to free ride in smaller groups [2], more prone to trust others in smaller groups [3], and they are more inclined to feel responsible to help others in need when part of a small group [4], compared to larger group sizes. Impressive strides have been made to expand theory on human group size, addressing interdependence at different levels of group size and activity and how larger social group sizes may have selected for larger primate brains-Dunbar´s social brain hypothesis [7]. These theoretical frameworks focus on group sizes of about 2, 5, 30, 300 and network sizes of 5, 15, 50, 150, respectively. Despite the wide range in group sizes discussed in theoretical work, dyads have received the most research attention in practice. Three examples illustrate this point. First, many economic games are two-player games or have only recently been extended to N-person games . An Exception is, for instance, the Public Goods Game, but these are not as widely studied as the aforementioned two-player games. Second, social exchange has been primarily investigated as a dyadic phenomenon [9]. This is not surprising as social exchange involves reciprocity . When exchange relations involve more people than two, typically it is no longer social exchange but economic exchange [10] in which a distribution rule or money plays an important role [11]. Third, in terms of early human development, major developmental theories focus on the dyad [12]. These include attachment theory, separation/individuation theory, and social learning theory. There is also a predominance of the dyad in friendship and romantic relationship research [13]. However, despite the fact that the dyad has received most research attention from scholars examining human interaction there is surprisingly little empirical research on what size of group people actually look for and encounter in everyday life. The focus on dyads in human interaction research is undoubtedly, in part, due to the methodological difficulties that arise when investigating larger group sizes, but it raises the question whether the dyad is also overrepresented in everyday life. Do we observe the dyad emerging as the predominant group size across a wide variety of social situations more than would be expected by chance? We have noted that systematic research on group size is scarce, even though it appears an important variable in research on, for example, free riding [2] or helping behaviour [4]. Still, there have been incidental observations in older research that also support the primacy of the dyad. For instance, nearly seven decades ago, research found that the dyad constituted 71% of informal groups and work groups [14,15]. Similarly, the dyad proved to be the most common group size at observations during free play at a nursery school, tourist parties visiting a national monument, and the size of dinner parties in a restaurant or conversation lounge [16,17]. More recently, a program of research by Dunbar and colleagues [18][19][20][21] has found support for Cohen's [16] hypothesis that, for conversation groups specifically, there is an upper limit of four persons who can engage in active conversation, primarily due to cognitive constraints. It also seems plausible that conversations can be coordinated better within a group of up to four persons rather than groups much larger in size. --- The current research This study addresses the question of whether we indeed observe the dyad emerging as the typical group size across a wide variety of social situations more than would be expected by chance. We provide preliminary empirical evidence from four survey studies that examined the size of the group in which people found themselves across social activities including dinner, movies/concert, off work chats, chats at work, projects, holidays, sports activity, and bars, as well as an experience sampling study of randomly selected situations people experience throughout daily life. Note that few scholars have tried to put forward situation taxonomies, and that these have not been successful in gaining strong support [22]. Therefore, we did not follow one particular situation taxonomy and instead our situations are eclectically derived from previous work. We define a group as "two or more persons who are interacting with one another in such a manner that each person influences and is influenced by each other person" [23] . As such, when two friends go out to watch a football game, they are a group of two, within a larger aggregate of fans. In Study 1, we focused on Dutch participants' past social activities. Study 2 replicated the results of the first study in a sample of US Americans, while Study 3 focused on US participants' preferred group size for imagined future activities. In Study 4, same-sex social activities were assessed to examine sex differences in a US sample. The final experience-sampling study involved a real-time data capture approach using Dutch participants' smartphones . --- Results --- The dyad is frequently the most common group size Full information on data screening can be found in the Supporting Information . We asked people to report the size of their social groups for a wide variety of activities. Group size frequencies are displayed in S1 pattern holds for Dutch participants in Study 1, American participants in Study 2, American participants reporting on preferred group size in Study 3, and for American participants reporting on activities involving same-sex others in Study 4. However, there are three cases where the dyad is the mode by only a small margin. These concern off-time chats in Study 1 , movies in Study 3 and projects in Study 3 . The typical group size for dinner parties was less clear. Future dinner parties of size 4 were strongly preferred in Study 3. Yet when people actually go out for dinner, they are typically with 2, 3, or 4 persons . Similarly, participants in Study 3 preferred to go to the movies with two or four persons , yet for actual movie visits, participants in Study 2 overwhelmingly chose to go as a dyad instead of as a group of four . To a lesser extent this also holds true for chats off work and projects, as indicated in S2 and S3 Figs. Moreover, as a general trend and perhaps not surprisingly, as group size increases the frequency of participants reporting to have been in such a group decreases. Except for sports activity, group sizes containing more than 10 individuals are rarely reported. --- The over-representation of dyads compared to other group sizes The most commonly used count models are Poisson and negative binomial. If smaller groups are easier to coordinate, and coordination costs increase exponentially with group size, the observed data would best fit a Poisson distribution . In contrast, we found that it does not, having better overall model fit with zero-inflated Poisson distribution. This improvement in model fit using a zero-inflated rather than a standard Poisson distribution indicates that there were more dyads in the observed data than would be expected simply on the basis of ease of coordination. We tested the over-representation of dyads using zero-inflated Poisson and zero-inflated Negative Binomial regression. As noted above, we first reparametrized the outcome as the count of additional group members beyond the dyad. That is, by subtracting 2 from every observed group size , a count measure of group size greater than two resulted. If we observe an excess of dyads, the zero-inflated part of the model thus reflects dyad inflation. We first tested the fit of the Poisson model . We then fit the two zeroinflated models and used a Vuong test to determine if they yielded better fit than the Poisson model. In almost every case they did , indicating dyad-inflation. The dyad-inflated models similarly support the notion that women are somewhat more dyadic-out of a total of 25 tests for all activities for four studies , 13 Poisson regression tests show significant sex differences, and 12 of these effects survive the move to dyad-inflated models with at least p < .10 . In no case do men show significantly more dyads or smaller group sizes relative to women. However, this also means that in about half of the activities, there were no significant sex differences and the differences we did find were generally small with some exceptions, such as same-sex sports, in which women were noticeably more dyadic than men . --- Age and group size Pearson correlations were performed to assess whether group size was associated with age. The correlations are reported in S3 Table. The significant associations were typically between small to medium in size. There was a positive correlation between age and dinner size for Dutch participants in Study 1, r = .11, p = .002. As age increases, so does the size of the dinner party. Yet, this result did not replicate with American participants in Studies 2-4. There was a negative correlation between age and group size at the movies for participants in Study 3 and men in Study 4 . As age increases, group size at the movies decreases. Note, however, that this relationship was not found for participants in Studies 1-2, and for women in Study 4. For Dutch participants specifically, there was a positive correlation between age and conversation sizes . As age increases, conversation sizes become larger. There was also a positive correlation between age and project group sizes for participants in Study 1 , Study 2 , and Study 4 women but not for participants in Study 3 and Study 4 men. Moreover, there was a negative correlation between age and holiday group size for participants in Study 2 but not for participants in Studies 3-4. Finally, Study 4 data indicated a negative relationship between age and sports group size in women , but not men ; and a negative relationship between age and group size while visiting bars for men , but not women . --- Relationship status and group size In order to assess the association between group size and relationship status , Pearson correlations were performed and reported in S4 Table . For dinner parties there was a consistent positive relationship between group size and relationship status. Participants in a relationship reported larger dinner sizes than participants that were single. This holds true for Study 1 , Study 2 , Study 3 , Study 4 women and Study 4 men . These effects can be considered small . There were no significant correlations between group size and relationship status for movies, projects, and bars. For conversation sizes the pattern was less clear with only significant results for off work chats in Study 1 ; chats at work in Study 2 and Study 4 men . The positive sign indicates that participants in a relationship had larger conversation sizes. Finally, there was a significant negative correlation between group size and relationship status for Study 2 participants' holidays , indicating that participants in a relationship had smaller holiday groups; and a significant positive correlation for Study 4 men in the activity of sports , that is, men in a relationship engaged in sports activity with larger group sizes. The significant correlations for conversation sizes and holidays should be interpreted with caution as they are not consistently found across the studies reported here. --- Real-time experience-sampling data indicate that dyads predominate in daily life Over the course of one week, participants were asked seven times a day to report the last situation they had experienced with another person since the last survey . The experience sampling data yielded 10,933 total responses, 7,248 of which involved social situations. Forty-eight of these did not specify the number of other individuals present; therefore, for this analysis N = 7200 responses. The dyad was the most common group size , followed by triads , groups of four , five , six , seven , eight , nine , ten , eleven , and 'more than 11' . We again tested whether women encountered a greater proportion of dyadic situations than men. To do so, we fitted a logistic mixed effects model with random intercepts for participants and with gender as a level-2 predictor of the number of interaction partners in social interactions . Here, the outcome variable indicated whether the situation was dyadic or involving any other non-zero number of interaction partners . We computed odds ratios for the fixed effects estimates. Analyses were conducted in R software version 3.4.4 using the tidyverse and lme4 packages. The hypothesis was not substantiated by the data. Women did not experience a greater proportion of dyadic interactions than men, OR = 1.056, d = .03. The detection of a sex difference in dyadic situations may be obscured by the overrepresentation of romantic couples for this unit size. Situations with romantic partners were indeed more dyadic than other situations, yet conducting the same analysis while excluding the situations with a romantic partner still found no significant sex difference . --- Discussion The results of the present studies provide strong evidence for the prevalence of the dyad in daily life. Our data show that dyads are most common across a range of activities obtained from three time moments , sampling both mixed-sex and same-sex groups, with three different methodological approaches in the United States and the Netherlands. With some exceptions, we also found similar patterns for men and women. These results are in line with classic research conducted in the 1950s and 1970s [14-17] which also found dyads to be the most common group size at local sites with local participants . Yet how do dyads relate to other, larger, group sizes? What theory can be used for understanding the prevalence of the dyad in daily life? To integrate the dyad with other group sizes, we briefly review Caporael's Core Configuration Model [5,6] and although we did not test causal mechanisms in this study, we subsequently provide four possible reasons for the predominance of the dyad in everyday life. --- The core configurations model Caporael [5,6] proposes that face-to-face groups are hierarchically structured in four core configurations. These represent kinds of interdependent interactions between people. The interdependency is determined by the type of task to be performed, the situated environment, and the physical attributes of the participants . Group size is an important feature of the model but the core configuration sizes have little meaning without invoking the associated activities to be performed. The notion of being "core" configurations reflects the idea that these group configurations can be repeatedly observed in hunter-gatherer groups , across an individual's lifespan, and in daily life. These configurations occur for a specific purpose and the successful accomplishment thereof accounts for their continued recurrence. Moreover, it is proposed that cognitive processes have evolved and developed in the context of the core configurations and thereby also cause their recurrence over time. The first core configuration is the dyad, an interaction between two entities . Tasks addressed by the dyad are, for example, internal fertilization in the context of sex, providing infant nutrition while the mother is breastfeeding the infant, or a child's interaction with a peer or adult. One proposed function of the dyad is microcoordination . Second, the work group, family group, or team has a modal group size of 5 individuals and a range of 3-7 people. It refers to interactions in small face-to-face groups that have a common task orientation. Examples of small group tasks are foraging, hunting, and gathering. A modern world example would be the completion of an assignment by ad-hoc groups of students working together for brief periods in class . The workgroup affords the function of distributed cognition across group members. This refers both to the sharing of cognitive resources in the pursuit of a shared representation of the task or problem and to a division of cognitive effort where there are specialties in the group over time allowing role-based trust. Third, the deme, band, microband, or conceptual deme has a mode of 30 individuals with a group size range of 25-50 people. Note that Marilynn Brewer proposed a group size range of 50-200 people for the deme/community in her 2015 keynote at the ICSD conference in Hong Kong, China [24]. The deme is similar in size to the extended family and modern-day classrooms. Common tasks of the deme are the movement from one location to another in hunter-gatherers, providing college students skill-based education , and the integration and redistribution of resources retrieved from smaller workgroups . This configuration has the function of constructing a shared reality or worldview, a common bond identity and common knowledge. It also allows cooperative alliances to emerge which may lead to the breakaway of group members to form their own group in case of conflict or when the environment's carrying capacity is reached. Finally, the macrodeme or macroband has a modal group size of 300 individuals with a range of 100-500 individuals. Brewer [24] has instead provided a range of "300-1000 and beyond" people. The task of the macrodeme is the gathering of people in the pursuit of exchanging resources, people, or information about more distant places or groups. The between-group mobility of people may involve mates in the case of hunter-gatherers or staff in the case of business or science. The function of the macrodeme has been proposed to be the stabilization and standardization of language. Stabilization indicating words referring to the same thing and standardization indicating that the members often use and understand the same words. Whereas the dyad can rely on synchronization and reciprocity to obtain goals [25,26] and whereas workgroups can rely on mutual performance monitoring [27] and self-regulation [28] to achieve their objectives, members of the deme have to resort to other mechanisms to sustain cooperation and coordination. Several mechanisms have been proposed in the literature. These include informal sanctions [29]; indirect reciprocity [30]; and descriptive norms [31]. However, whereas in demes people are all individually known, this is not the case in macrodemes. As such the likelihood of knowing someone's reputation may be insufficiently high for indirect reciprocity to serve as an effective mechanism to foster and maintain cooperation and coordination in very large groups [32]. In addition, it has been argued that there are mental constraints preventing people to keep track of the reputation of a large number of people at the same time [13,33]. Hence, other mechanisms such as proscriptive norms, institutions, and formal sanctions [34,35] may be necessary for people to work effectively together toward the accomplishment of their goals in macrodeme configurations. --- Four reasons for the prevalence of the dyad More than larger sizes, dyadic interactions enable benefits from direct reciprocity [26]. In repeated dyadic interactions, the threat of non-reciprocation supports cooperative strategies such as tit-for-tat [26,36], but this mechanism breaks down in larger groups where freeriding is possible [37]. In a dyad, such strategies are efficient, as one's choices are noticeably affecting one's own and other's outcomes. The partner may readily perceive that the other prefers stable cooperation over mutual selfishness, where exploitation would meet swift retaliation. However, even the small step from dyads to triads causes complexities in the workings of reciprocity, as one's retaliatory action can no longer be delivered to the desired target only, the third person is equally affected [38]. When group size increases further, one is increasingly less able to induce cooperation in others through strategic signalling. One individual's actions are both less likely to be perceived, and less likely to affect the others' outcomes [37]. Thus, being part of a dyad allows for more control over the social situation toward the accomplishment of mutually beneficial outcomes. Second, the detection of emotions and mental states via nonverbal cues is most likely to occur in dyads. Yet as group size increases, it becomes increasingly challenging to attend n-1 communication channels in a group comprising n members [39]. Moreover, as group size increases, the number of interpersonal linkages along which coordination may be required increases sharply [39]. Indeed, coordination with multiple individuals is computationally complex and therefore individuals should prefer interaction partners with whom coordination is easier [40]. Coordination is more efficient with familiar others as learning about others' preferences, intentions, and traits allows for improved behavioural anticipation [41]. As mentioned above, dyads are expected to be in a better position to attribute mental states , to be clearer communicators, and hence to be behaviourally more predictable than larger groups. Thus, if people prefer those with predictable behaviours as interaction partners, such as familiar others [41], people may also have a preference for dyadic activities given their relatively predictable form and hence lower cognitive effort, compared with larger group sizes. Third, group living provides various benefits. These include cooperation in the pursuit of difficult tasks, protection against danger, directing and receiving altruistic acts toward and from kin, the availability of allies, and high concentrations of mates [42]. However, group life also comes with liabilities, mostly in the form of conflict and competition. Examples include competition over material resources, high-status positions, or romantic partners. As such, individual competitors may want to exclude others who pose a threat to their interests . Being excluded in an ancestral environment yielded dire prospects for survival and reproduction. Therefore, it is not surprising that social exclusion gives rise to anxiety [43] and social pain [44] and that people may be afraid to deviate from the group. Moreover, it has been argued that humans are highly sensitive to actual and threatened rejection and may possess an ostracismdetection system biased toward overdetection [45,46]. Whereas in larger groups social exclusion is a possibility, sometimes a threat, in a dyad one cannot exclude the other person without bearing the cost of becoming alone oneself. Seeing others together when one is alone may trigger affective distress, being a reminder of social exclusion [43] and as such is uncomfortable. Thus, even if one is generally embedded in a larger group and as such reaping its benefits, the dyad specifically may provide a relatively comfortable unit for social interaction in which self-monitoring and self-censorship can be somewhat relaxed, allowing for more authentic behaviour . The final argument that may help explain the prevalence of the dyad in daily life concerns reproduction and infant-rearing activities, which take place mostly in pairs [5,6]. As mentioned above, this includes interactional synchrony during the initiation of courtship, which is a process confined to the mating couple [47]. For instance, it is difficult to conceive how three people would be able to nearly continuously look into each other's eyes. Moreover, most people choose to engage in sexual activity with one other person during a sexual encounter. Next, the provision of infant nutrition through breastfeeding typically involves the mother and the infant in a breastfeeding dyad. Finally, alloparents can assist the parents in infant-rearing by watching over or feeding the child, allowing the parents to allocate their time in the pursuit of other activities. We argue that even with extensive alloparenting, the dyad may still be the most functional unit for interaction between caregiver and young child . Yet how do these theoretical reasons connect to the activities that were probed in the present studies? Our activities may have had low risk of exploitation, low costs in terms of commitment, and involved the exchange of resources likely of little consequence. However, the four proposed mechanisms we suggest are all attuned to the two-person interaction in daily life. This applies to reciprocity where the interactants engage in exchange and turn-taking. For instance, paying for the drinks at the movies, one may expect the other to pay next time. Reciprocity can happen "on the spot" , but more frequently unfolds in a step-by-step manner. As social transactions recur and gradually expand in significance, reciprocity results in fortified interdependent social bonds [10] . Besides trust, these processes are often accompanied by a series of emotions that may serve as internal pressures to maintain interpersonal cooperation: feelings of indebtedness, personal obligation, appreciation, and gratitude after receiving a favour [51]; feelings of guilt when failing to reciprocate [52]; anger when receiving substandard exchange [53], and forgiveness to pardon having been short-changed [54]. If the above line of reason is correct, it suggests the interesting hypothesis that various feelings and emotions such as indebtedness and gratitude are experienced most commonly and most strongly toward one other person instead of toward groups. Second, in terms of coordination, in a dyadic interaction one enjoys relatively noise-free information. Compared to larger group sizes, dyadic interactions offer less room for hiding and ambiguity, and this facilitates coordination. For instance, when a group of friends are planning a holiday trip, it may require significantly more effort to agree on a departure date and destination, not to mention all the small decisions to be made once there, compared with a dyad planning a similar trip together. Third, in a range of situations , being in a dyad may also entail being tied to another person, incentivizing both persons to be accommodating and cooperative to prevent social exclusion and having to fend for oneself in a sometimes daunting world. In a bar or club, for instance, it may feel uncomfortable to have no one to fall back on if needed. Finally, dyadic interaction also allows for relatively unconstrained flirting behaviour compared to situations where third-parties are present [56], reflecting the role of reproduction in the prevalence of the dyad in daily life. During a dyadic conversation, for example, a subtle courtship attempt can more freely be sent without potential interception costs from third parties. These mechanisms help to explain the primacy of the dyad because each of them supports cooperation in smaller groups-by which all individuals involved benefit. This is not to imply that cooperation does not occur in larger groups. We suggest that it is more challenging, however, and it involves qualitatively different mechanisms, such as indirect reciprocity [30] or third-party punishment [57]. In this empirical paper we echo the importance and prevalence of the dyad in society that has recently been demonstrated in mathematical and simulation analyses by Allen et al. [50]. --- Sex differences For our mixed-sex diary data we did not find significant sex differences. There were some significant differences for mixed-sex group activities sampled in Studies 1-3, but these were generally small. However, for same-sex groups in Study 4, the sex differences were somewhat more pronounced, where women are more dyadic than men. This applies for instance to sports or going to the movies and to a lesser extent going on holiday or out for dinner . This latter finding is consistent with research on same sex friendship [58]. Using a sample of 112,000 Facebook profile pictures the authors found women to favour dyadic relationships, while men preferred larger male groups. Moreover, it has been demonstrated that men and women process social information differently in line with these differences in social structures, where women focus more on individuals while men focus more on groups [59,60]. Although the exact mechanism is unknown, various complementary explanations have been put forward to account for these sex differences . --- Limitations Studies 1-4 have three limitations in common that should be taken into account. The situations that were sampled may not have been exhaustive or fully representative of daily life. Although we asked for the last time someone engaged in the activity, for some people it may have been weeks ago. Participants may also selectively remember an instance they particularly enjoyed. These concerns were addressed by Study 5, which used a real time approach in which the reference period is the present, or the last hours, and in which situations are sampled randomly. A limitation of Study 5, however, is that we did not code the situations that people reported. This is a fruitful avenue for future research given that situation taxonomies are currently lacking consensus and may be advanced by diary studies. The current research focused on social activities and hence we have not probed individuals conducting activities alone [14,15]. Therefore, we do not know whether dyads would also be most frequent compared to individuals for various different tasks. In our studies, most participants engaged in activities with a few other people. However, there are certainly activities that require larger groups . Future research could investigate whether even there, dyads may form the most effective subcomponents . Indeed, given the aforementioned arguments , these activities may also predominantly yield the dyad as the most common subunit in a larger collective. Related to this, because we focused on direct interaction, and used a definition of a group as "two or more persons who are interacting with one another in such a manner that each person influences and is influenced by each other person" [23], our results do not address the fact that groups can have psychological significance beyond direct interaction [65] and measuring groups differently may yield different results. --- Future research One issue remaining for future research is to provide empirical evidence that the dyads that are so prevalent in various domains are also stable over time. Moreover, future research would do well to investigate the prevalence of the dyad in collectivistic daily life, to determine whether dyad inflation is generated by cultural factors as a necessary requirement for the effect to occur or whether an evolutionarily ingrained predisposition is sufficient in and of itself. If the former holds, then the effects should be observed only in individualistic countries, whereas if the latter holds, as argued by Caporael's core configuration model [5,6], then the effects should be observable across individualistic and collectivistic cultures, and more generally, around the world. --- Materials and methods --- Study 1 All studies reported were approved by our faculty's ethical review board . Participants indicated approval with our informed consent form which was displayed on a computer screen. In order to obtain a representative sample of the Dutch population we collected data through Flycatcher , an online research company in the Netherlands. The Flycatcher panel consists of 16,000 participants who frequently participate in research. Participants receive points for every questionnaire that they complete. The accumulated points can subsequently be exchanged into gift vouchers. Participants. A Dutch sample of 459 women and 508 men participated in this study and received points for their participation. Measures. The respondents were presented with five daily activities and were asked to report the number of persons present in these activities while including themselves in the count. Participants read the following prompts. 'Think about the last time you went out for dinner with friends.', 'Think about the last time you visited a movie or concert with friends.', 'Think about the last time you had a conversation with others in your spare time.', 'Think about the last time you had a conversation with others at work.', 'Think about the last time you worked together on a common project or assignment as part of your work or training.', and then they were asked with how many persons, including themselves, did they engage in the respective activity. Note that the measures for this, and subsequent studies, are reported verbatim in the OSM. Procedure. Participants filled out a large survey as part of a third wave of research on trust, relationships and moral dilemmas. The five questions on group size described above were part of this questionnaire and were added for exploratory purposes. After finishing all the questions, the respondents filled out their demographical information, were debriefed and thanked for their participation. --- Study 2 and 3 In Study 2 and 3 we aimed to replicate Study 1 while changing the population from Dutch to American participants and changing the data collection platform through which the responses were collected, from the Flycatcher Panel to an online crowdsourcing service: Amazon's Mechanical Turk . In addition to the five activities sampled in Study 1, we added holidays as a sixth activity. In Study 2 respondents were asked to recall past activities in which they took part, while in Study 3 participants were asked to imagine themselves in a hypothetical future activity. The studies were run at the same time; participants were assigned to either Study 2 or to Study 3 after accepting our HIT on M-Turk. Measures. All participants were presented with six activities. Participants in Study 2 reported on past activities and were presented with the following six prompts: 'Recall the last time you went out for dinner with friends.', 'Recall the last time you went to the movies or a concert with friends.', 'Recall the last time you had a conversation with others in your spare time.', 'Recall the last time you had a conversation with others at work.', 'Recall the last time you worked together on a common project or assignment as part of your work or training.', 'Recall the last time you went on vacation.', and then were asked with how many persons, including themselves, did they partake in the respective activity. Below each of these questions the participant was presented with a text box that read 'Number of persons:' in which a number could be filled in. In contrast, participants in Study 3 reported on imagined future activities and were presented with the same six prompts as above except that the question wording was slightly different . Next, participants were asked, with how many persons, including themselves, did they prefer to be in the respective activity. Underneath each question a textbox was presented that allowed the participants to fill in the number of persons. Finally, the following demographics were obtained from participants: 'Do you identify as male or female', 'What is your relationship status', 'How long have you been together for' , 'What is your sexual orientation', 'What is your age', 'Are you Hispanic or Latino', 'What best describes your ethnicity?' and 'Please enter your worker's ID'. Procedure. Participants entered the experiment through clicking on our HIT on Amazon Mechanical Turk. The welcome screen presented the consent form and explained the study was about decision making in daily life. After providing consent, the various six questions were presented to the respondents and they filled out the number of persons that were present or preferred for each activity. Next, the participants reported their demographical information and were given a code that would allow respondents to obtain their reimbursement. Finally, participants were debriefed about the study. --- Study 4 In the previous studies we did not identify strong sex differences between men and women in the group sizes that were reported. It can be argued that no sex differences are expected in these situations as many or all participants may have reported on mixed-sex situations. Therefore, in Study 4 we aimed to focus on sex differences between men and women in the group sizes they report in various activities by asking explicitly about same-sex daily activities and by including two additional activities in which sex differences may be more likely: sports activity and visiting bars and clubs. Participants. We recruited 450 men and 604 women as respondents in this study. The sample consisted of respondents from the United States. Participants received $0.40 for their contribution. Measures. All participants were presented with eight same-sex daily activities that had happened in the past. The questions explicitly mentioned same-sex others in boldfaced text and clarified the sex of the others in parentheses. The following two example prompts were used for women .'Recall the last time you were physically engaged in a sports activity with same sex others .', 'Recall the last time you were going out to a bar or club with same sex others .', and then were asked with how many persons, including themselves, did they engage in the respective activity. The questions for men were the same except for the information in parentheses, it specified 'men only'. The demographical questions were identical to Studies 2 and 3. Procedure. Participants accessed the study through Amazon Mechanical Turk. After reading the consent form the respondents proceeded with indicating their sex . Women were assigned to the female question format condition and men were assigned to the male format condition. Next, the participants indicated the group size that characterized the eight activities in which they had taken part in the past with same sex others. Finally, respondents provided demographical information, were provided the code to obtain payment and were debriefed about the study. --- Study 5 In Studies 1-4, we asked people to report on a wide variety of common situations. However, these categories may not have been exhaustive, and may not fully represent daily life. We therefore used experience sampling to obtain reports on random social situations. Moreover, in Studies 1, 2, and 4 we used retrospective reports to investigate group size. Although we asked for the last time someone engaged in a particular behaviour , for some participants the last event may have been weeks ago. Finally, participants may remember a dinner that they especially enjoyed rather than the last dinner per se. To address these memory-related concerns we decided to use a real-time data capture approach in which the typical reference period is brief . Participants. We used data from the Interdependence in Daily Life Study [67]. A sample of 284 Dutch individuals with a mean age of 35.55 years were recruited through panel agencies and snowball sampling. The majority of the participants were born in the Netherlands . Participants were paid €0.50 for every survey completed and obtained a bonus of €20 for a response rate of 80% or higher, as well as payments for other parts of the study not covered here, up to a total of €74.50. Measures. Participants were asked whether they experienced a situation with another person since the last text message received. If so, respondents were asked with reference to the above situation: "Apart from yourself, how many people were present?" on a scale from 0 to 10 with the option to select 'more than 10'. Procedure. After an intake session during which demographics were recorded, as well as surveys and experimental tasks unrelated to this study, participants completed a seven-daylong experience sampling phase . Each day between 08:00 and 22:00, participants received seven messages at semi-random times . Participants received a survey link by text message sent through SurveySignal [68] and were prompted to recall the most recent situation they had experienced with another person since the last survey. --- The data and R code used for the analyses can be found at the OSF: https://osf.io/kqvcu/ --- Supporting information S1

A salient objective feature of the social environment in which people find themselves is group size. Knowledge of group size is highly relevant to behavioural scientists given that humans spend considerable time in social settings and the number of others influences much of human behaviour. What size of group do people actually look for and encounter in everyday life? Here we report four survey studies and one experience-sampling study (total N = 4,398) which provide evidence for the predominance of the dyad in daily life. Relative to larger group sizes, dyads are most common across a wide range of activities (e.g., conversations, projects, holidays, movies, sports, bars) obtained from three time moments (past activities, present, and future activities), sampling both mixed-sex and same-sex groups, with three different methodological approaches (retrospective reports, real-time data capture, and preference measures) in the United States and the Netherlands. We offer four mechanisms that may help explain this finding: reciprocity, coordination, social exclusion, and reproduction. The present findings advance our understanding of how individuals organize themselves in everyday life.

Introduction Through natural learning, humans develop understandings of their increasingly technologized environments long before they attend school . In this context, the earliest sphere of human development, the family home, sets the first primary stage for children's learning experiences, before their horizons begin to widen when they enter other childhood environments including their neighborhoods, pre-schools, primary schools or other family homes . Within these broader home and childhood environments, a global sociotechnical change has occurred over the last decade through the rise of commercially available Digital Voice Assistants like Amazon's 'Alexa', Apple's 'Siri', or Google's 'Google Assistant' . One-third of US adults already own stand-alone DVAdevices ), while forecasts predict the total number of DVA-enabled devices, which had already exceeded four billion in late 2020, will outnumber the human world population by 2024 . Although DVAs may neither be the only nor the most sophisticated manifestations of intelligently behaving technologies, these individually and communally accessible voice interfaces are one of the most obvious experiences of artificial intelligence in today's home and childhood environments . In contrast to smart toys and other technologies solely designed for children, DVAs' commercial ecosystems comprise a growing number of hardware and software components , which together are becoming omnipresent parts of everyday life . In addition, preliminary research suggests DVAs' omnipresence across various devices at home is especially appreciated by families , further underpinning the relevance of DVAs in the context of developmental research. But it is not only the growing socio-technical omnipresence of DVAs which could make them an important case for developmental research, but also their potential ontological momentum as experienced by today's children. DVAs are part of the human-technology dyad which has evolved since humans first used stones to break open coconuts or fallen trees to bridge rivers . In this dyad, DVAs represent the stage of "autonomous technological beings" , which are able to emulate peculiar qualities of human beings like language and speech, and, as Nass & Brave point out, "over the course of 200,000 years of evolution, humans have become voice-activated with brains that are wired to equate voices with people and to act quickly on that identification". However, it should also be remembered humans seem to have a natural tendency to 'equate' almost anything with people, and to act quickly on this identification through something formally known as anthropomorphism. Anthropomorphism refers to the psychological phenomenon that humans tend to engage socially with non-human entities as if these entities were human. These humanlike engagements include their behaviors , their feelings and their perceptions . 1 This phenomenon is seen across various societal, cultural, religious and historical contexts , and has been scrutinized across diverse fields of research including psychology, neuroscience, psychiatry, philosophy, ethology and education . But since the growing digitization of the environment began in the second half of the twentieth century, anthropomorphism has become particularly relevant in addressing how humans engage socially with technology -with personal computers, basic robots, electronic toys etc. )). With ongoing technological progress , anthropomorphism continues to be a popular research theme in the literature ), and especially in the context of DVAs ). Despite this wealth of material, little attention has been paid to the systematic differences between commercial DVAs and the animal-like or human-like technologies and toys often used in anthropomorphism research . For example, most of these robotic technologies and toys convey the notion of being self-contained entities . In contrast, children may experience DVAs as a different phenomenon, because they appear both to be omnipresent in various environments at the same time and to be intelligent in the sense that they are closely intertwined with sources above and beyond their physical boundaries . Furthermore, robotic 1 The review does not differentiate between anthropomorphism and animism. While the latter term refers more generally to cases in which humans engage with non-living entities as if they were alive, the former term refers only to those cases in which humans engage with non-human entities as if they were human . In addition, this review mainly focuses on anthropomorphism in the context of technology, which differs from anthropomorphism in other contexts . Therefore, the review acknowledges any references to other contexts must be treated with caution. technologies and toys are often designed to convey strong visual and haptic experiences , or to evoke caring or nurturing behaviors . As a consequence, such entities are usually limited to specific age ranges, and children eventually grow out of playing with them . In contrast, DVAs are not limited to specific age ranges: their minimalistic physical designs and utilitarian 'personalities' with a constantly growing number of functionalities and applications constitute an appeal to infants as well as elderly people, therefore turning DVAs into potential lifetime companions in their own right . In the context of daily family life, this could mean that children rarely observe adults engaging socially with toys, yet they may often observe adults engaging socially with DVAs through human language and speech, therefore prompting children to conclude DVAs are serious voice-enabled engagement partners in their own right. Lastly, it should also be noted that although DVAs may be less embodied in terms of animal-like or human-like appearances , DVAs' are still physically embodied entities visually conveying socially salient cues . 2The key question remains how anthropomorphism research relates to children who grow up amid increasingly intelligent and omnipresent technologies, particularly DVAs. This article suggests the answer may be less obvious than it seems. In particular, it is argued the nature of anthropomorphism may change as humans' experiential understandings of humanness change, and this may particularly apply to today's children who grow up amid increasingly intelligent and omnipresent technologies, such as DVAs. This article uses the methodological basis of a conceptual review, which means the synthesis of the interdisciplinary body of literature follows a conceptually framed process to address the key question of interest . The process is as follows: the first section reviews anthropomorphism research according to three different research perspectives . Following a brief excursus on philosophical pragmatism, the second section then assesses how each of these perspectives relates to children who grow up amid increasingly intelligent and omnipresent technologies, such as DVAs. Finally, the last section summarizes the main arguments, outlines its limitations, and offers some directions for future research. --- What Is it? What Are its Consequences? What's Causing it? A Triadic View on Anthropomorphism Research Inspired by Epley's core questions which anthropomorphism research should address , this section reviews anthropomorphism research according to three different research perspectives. The purpose of this triadic view on anthropomorphism research is to provide a comprehensive account of contemporary anthropomorphism research in the context of human-technology engagements. Facing the trade-off between analytical scope and depth, the following review refers to 'technology' and 'social engagements' as rather broad categories comprising various types of technological entities and engagements , while special attention is only paid to the peculiarities of children's engagements with DVAs. Although this is not without limitations, taking this approach is in line with the review's overall objective to look at the bigger picture of contemporary research investigating humans' engagements with technology, in general, and children's engagements with technology, in particular. --- What Is it? Anthropomorphism from a Descriptive Research Perspective From a descriptive perspective, anthropomorphism research uses human-human engagements as a benchmark to investigate empirically observable parallels to human-technology engagements. This is, to what extent humans seem to engage with non-human technology as if these technological entities were human beings. Early empirical work which suggested human engagements with technology bear deep parallels with human-human engagement patterns )) has been followed by an array of research on the different delineations and dependencies of anthropomorphism . For example: • more humanlike appearances and behaviors of technological entities can directly increase how humans rate the entities' human-likeness and likability ); • more humanlike technological entities can convey signals of 'intention' better to humans compared to less humanlike technological entities ); • humans tend to engage in less abusive behaviors towards technological entities when the entities appear more intelligent ); • humans' trust in humanlike technological entities can be more resilient compared to their trust in less humanlike technological entities ); • voice-enabled technological entities can be held more morally accountable for their actions than simple technological entities without voice-interfaces ); • technological entities with non-verbal emotional expressions, or with humanlike rather than robotic voices, can yield higher levels of acceptance and attachments in humans )). With regard to DVAs, humans who personify these voice-enabled technologies tend to be more satisfied with their overall engagement experiences, and tend to cultivate more sociable relationships with them . Additionally, DVAs which emulate humanlike However, the role of age in anthropomorphism remains intricate, or, as states, "the difference between adults and children is not qualitative but rather a matter of complexity". For example, although infants and young children often engage socially with non-human entities by pretending the entities are human , they also possess firm understandings of the ontological borders they cross and the behavioral commitments they make . This suggests children's anthropomorphism cannot be reduced to naïve confusions or infantile pretend play . Despite this considerable research about the nature of anthropomorphism in human-technology engagements, the "promiscuous use of the term" in the literature suggests there is still little consensus about a precise definition of what anthropomorphism is and is not. Although the attribution of mental states lies at the heart of anthropomorphism for many authors ), other empirical studies suggest anthropomorphism it is not merely about the attribution of mental states; rather, it is about the holistic ontological concept of humanness people apply when engaging socially with non-human entities . However, these underlying concepts of anthropomorphism often seem too intricate to explicate, yet are appealingly intuitive to human audiences . It is, therefore, unsurprising that anthropomorphism research continues to explore almost any corner of the social experiences which humans can potentially have with other humans. This suggests the alleged 'promiscuity' more likely reflects what anthropomorphism research is about from a descriptive perspective: investigating parallels between the empirically observable patterns of human-technology engagements and the openly interpretable benchmark line of human-human engagements. Yet investigating parallelisms does not require consensus on the 'true' nature of the designated benchmark line , it only requires minimal consensus on what its designated endpoints are in order to identify parallels between alternate endpoints -and there is, of course, little disagreement about who these living organisms, widely referred to as 'humans', actually are. --- What Are its Consequences? Anthropomorphism from a Normative Research Perspective As expressed by Epley , something "that has no demonstrable consequences is not worth studying". The above discussion already suggests anthropomorphism has various consequences for human-technology engagements, but, from a normative perspective, the important question is how to evaluate such empirically delineated consequences. On the one hand, these evaluations can yield concrete recommendations about how to improve technology's usability and effectiveness in human-technology engagements through the nuances of humanlike designs ). Therefore, from a normative perspective, anthropomorphism research offers various potentials and opportunities to exploit the human tendency to engage socially with technological entities for meaningful and practical purposes. This is sometimes referred to as 'applied anthropomorphism' . For example: • naturalistic and humanlike movements can enhance technology's likeability ); • technological entities endowed with conversational humanlike design components can give human users an immediate intuition of how to engage with them ); • more humanlike technological entities can facilitate the collaborative effectiveness of human-technology engagements, or they can evoke socially desirable behaviors from humans -like refraining from harming the entity ); • engaging with humanlike technologies can serve as a 'scaffold' for children with autism to improve their social cognition ). But, evidently, whether the integration of humanlike design components is desirable also depends on the concrete applications for which technological entities are used ), especially in the context of child-technology engagements . For example, although social robots have proven their promising potential as learning technologies ) -and this seems to hold true for DVAs as well ) -children can also develop adverse responses to too much human-likeness ), therefore echoing Mori's widely-cited 'uncanny valley' theorem about the eeriness of almost perfect human resemblance. On the other hand, the consequences of anthropomorphism in human-technology engagements have prompted researchers to evaluate the ethical risks of endowing technology with humanlike design components ). At their core, such ethical criticisms, sometimes referred to as the 'forensic problem of anthropomorphism' , consider the creation of humanlike technology as a form of deception, because they 'trick' humans, especially children, into overestimating the technology's 'true' capacities , or into engaging with the technology in ways which are supposed to be reserved for genuine human-human engagements . This may even culminate in the fear that humanlike technology could constitute a threat to human distinctiveness ). At the forefront of this criticism, Turkle ) has repeatedly emphasized how the temptations of engaging with technology, in general, and humanlike technology, in particular, appeal to human vulnerabilities rather than human needs, because the technological entities in question constitute seductive appeals to replace, denigrate, deny or degrade what deserves integrity: the existential value of humanhuman engagements. According to this normative perspective, no matter how authentic a technology's embodied human-likeness becomes in terms of its experienceable social qualities, and no matter how much humans appreciate and enjoy engaging socially with this particular technology, there remains an unambiguous normative hierarchy between engaging with 'simulations' of human nature and with human nature itself . More recently, Turkle has re-emphasized this normative positioning in the context of DVAs by arguing children who engage with DVAs, and who may even develop social bonds or a sense of friendship with them, are ultimately at risk of 'forgetting what it means to be human'. --- What's Causing it? Anthropomorphism from an Explanatory Research Perspective At its core, anthropomorphism is a psychological phenomenon, and, as such, the causes of its origins and variability must be explainable by psychological mechanisms . From an explanatory research perspective, this means there must be a psychological reasoning for the question why humans tend to engage socially with non-human technology as if these entities were humans. The answer lies in the heart of human social cognition. Social cognition seems to be a particularly strong quality of human intelligence compared to other primate species , and there are good reasons to assume the foundations of social cognition are innate due to their early manifestations in human development ). As humans develop, social cognition culminates, among other things, in the ability to infer and contemplate the intricate perspectives of others in their beliefs, interests and motivations for action -usually referred to as 'Theory of Mind' . But what about non-human others? Social cognition is a pervasive psychological ability, and, consequently, anthropomorphism should not be reduced to "a by-product of misplaced social cognition […]; rather [it is] an unavoidable consequence of the functional organization of the human brain" . For example, children's early ability to evaluate the passively observed social behaviors of others seems to extend naturally to human-technology engagements, as exemplified by 18-month-old infants who can already recognize whether a robot engages in socially contingent dialogues with adults . Furthermore, neuroscientific evidence suggests the same neural mechanisms of social cognition are also at work when humans anthropomorphize non-human entities ). From an evolutionary perspective, this propensity of social cognition to be triggered by human and non-human entities can be explained as an adaptive survival mechanism. In primeval environments, which had the greatest evolutionary impact on the development of the 'modern' human brain , it was a more effective strategy to behave falsely as if something was alive and similar to oneself, and a more fatal strategy to behave falsely as if something was inanimate and without agency or intelligence ). Notably, even today individual tendencies to anthropomorphize non-human entities are positively associated with other evolutionary salient behaviors , and also other non-human primate species tend to engage socially with technological entities when the entities imitate the bodily movements of the animal engagement partner . In line with this brief account of social cognition, Epley et al.'s widelycited three-factor theory assumes anthropomorphizing non-human entities follows the same cognitive process of inductive inference which also orchestrates humanhuman engagements, namely inferring concrete assumptions and predictions about the otherwise opaque inner nature of social engagement partners. According to Epley et al. , this inference is anchored in the inductive base of humanness, in humans' deeply ingrained knowledge about the self in particular , and about humans in general . Although it is challenging to disentangle the ego-and homocentric cognitive basis of anthropomorphism ), it is important to note human social cognition offers different routes, ranging from using one's own mind for simulating the mental processes of others , to using more abstract and knowledge-based forms of inference . Furthermore, this readily accessible cognitive anchor of humanness is co-determined by two fundamental motivational forces, namely sociality motivation and effectance motivation . These motivational co-determinants are important factors in explaining the empirically observable variability of anthropomorphism across humans ), and they can also account for the presence of anthropomorphism in populations with impaired social cognition . However, the inductive base of humanness as the cognitive anchor of anthropomorphism, and its co-determinants of sociality and effectance motivation, are far from uniform, especially when comparing urban and rural populations ). Therefore, as outlined by Epley et al. , the full variability of anthropomorphism can be explained only by taking into account the dispositional, developmental, situational and cultural sub-determinants of each factor . --- A Pragmatistic View on Anthropomorphism Research in the Context of Children and DVAs The current understandings of anthropomorphism, as reviewed above, seem to assume the dualistic concepts of humanness and non-humanness, which define the existence of anthropomorphism, bear a metaphysical truth . But what if there is no metaphysical truth to such concepts? What if these concepts are epistemological instruments made up by the clever animals who use them? In other words, what if "truth, at its core, is not a metaphysical category but rather a moral and epistemological one" ? This is the provoking story of philosophical pragmatism, particularly one of its variants referred to as instrumentalism or constructive empiricism , which dates back to the classical pragmatist William James )). 3Despite the different nuances which have evolved since its emergence in the late nineteenth century, classical pragmatism shares three philosophical themes: the active and subjective role of the interpretative human mind in scientific and non-scientific knowledge acquisition; the fallible and non-apodictic nature of knowledge due to its reliance on human experience; and the rejection of foundationalist truths and certainties such as the Cartesian mind-body dualism . These themes are certainly not exclusive to pragmatism per se, because as a rather heterogenous tradition, it has shares overlaps with other philosophical paradigms. For instance, pragmatism's adherence to a mind-independent reality follows the rationale of realism. However, rather than assuming that realism allows us to get reality objectively 'right', it is considered a hypothesis we need to be able to learn by experience . For Jamesian pragmatism, this process of learning by experience is teleological in nature, meaning human cognition is always subject to individual motivations and interests . Therefore, beliefs about knowledge and truth must always be viewed in light of the believers' motivations and interests, Some key terms have been adjusted compared to Epley et al.'s original terminology . Source. Developed from Epley et al. and, as James' work on religious experiences suggests, knowledge is true insofar as it has real consequences for those who believe in it. From this pragmatistic viewpoint, the dualistic concepts of humanness and nonhumanness are understood as functional distinctions for dealing with reality, and as true only insofar as they are practically useful for those who believe them according to their motivations and interests . In other words, pragmatism it is not repudiating dualistic concepts as such; rather, it is less concerned with their metaphysical validity and considers that "supposedly sharp distinctions may be better conceived as lines of demarcation drawn at some point on a continuum" . The question remains what a pragmatistic view on anthropomorphism research means in the context of children who grow up amid increasingly intelligent and omnipresent technologies, such as DVAs. Following Servais' recent pragmatistic account of anthropomorphism research in the context human-animal engagements, the main starting point is to focus less on a priori certainties and to prioritize the nuances of human experiences, such as children's subjective experiences when engaging with DVAs in their home and childhood environments. Based on the three research perspectives reviewed above, this is discussed in more detail in the following. --- Pragmatistic View on the Descriptive Research Perspective From a descriptive perspective, anthropomorphism research uses human-human engagements as a benchmark to investigate the extent to which humans engage with non-human technological entities as if these entities were human beings, while there is still little consensus about what a definition of anthropomorphism should comprise, as reflected by the wide range of interpretations in the literature . But is it feasible, from a pragmatistic viewpoint, to call for more clear-cut conceptual boundaries of anthropomorphism )? This intention seems noble, but the philosophical subtext of defining clear-cut boundaries would be that there were corresponding metaphysical categories of the human and the nonhuman. From a pragmatistic viewpoint, this would be an epistemological arrogation, because such boundaries represent only human-made lines drawn on some continuum for practical purposes and without any finality in meaning. This is partly reflected by the gradual decline of traditional anthropomorphism research which equates humanlike engagement patterns with genuinely 'false' behaviors or beliefs . For example, the recent anthropomorphism literature on human-animal engagements suggests "much of what has been considered as anthropomorphic interpretations may in fact do more justice to the mental states of other animals than was previously believed" , such as 'behaving as if dogs really understand what humans say' ). However, should the arguments be different for human-technology engagements ? In other words, is it not possible to be more certain about the things which differentiate humans from the things they have created themselves ? A pragmatistic answer would be to ask whether these are the most useful questions to ask for the investigation of human-technology engagements from a descriptive perspective. After all, it lies in the nature of a continuum that it is "constituted as much by difference as by similarity" . On the continuum of social engagements, human-technology engagements circumscribe one of the infinitesimal variations of engaging socially with one's environment. Yet to be on this continuum in the first place is simply an all-or-nothing question, because, as Seibt points out, social engagements of any kind require a basic social commitment, which can either be undertaken or not, but it must always be real. In the words of Seibt : "pretending to undertake a commitment is simply to fail to undertake it. […] Making a promise or treating someone as a person are real social interactions by virtue of engaging in a certain declarative practice". On this basis, claiming children or adults engage socially with a technology 'as if' it was a human remains misleading: they either engage socially with it in one way or another, or they do not; but if they do, they have granted it the status of a social engagement partner in its own right. Moreover, and in line with the pragmatistic adherence to realism, Seibt emphasizes social engagements of any kind emerge from the reality of experienceable social behaviors as they are displayed by social engagements partners , regardless of whether these behaviors are deemed authentic or not. As Seibt writes: [I]t is an obvious requirement of social praxis that the performance conditions of a social interaction […] must relate only to behavioral criteria and cannot take intentional states into account. […] [C]harges about someone's performing a social action inauthentically or insincerely relate to the quality of the social action, not to [its] occurrence. Both issues raised by Seibt seem particularly relevant in the context of DVAs. Alexa, the Google Assistant, Siri or other DVAs are designed to engage with humans through sophisticated social behaviors, namely the autonomous use of human language and speech . In light of such strong "behavioral realism" , the commitments of social engagement are supposedly easily undertaken , therefore granting DVAs the status of being social engagement partners in their own right -irrespective of whether their behaviors are perceived as being authentic or not. For empirical research, the pragmatistic viewpoint suggests children's social engagements with DVAs reflect how they subjectively experience what they deem significant in terms of sociality, and that the nature of their engagements with DVAs reflects how they translate this experience of sociality into real social behaviors. Accordingly, the description that children engage with DVAs 'as if' these technological entities were friends, playmates, companions or simply humans, may not adequately capture what children subjectively experience. In summary, narrowing the scope of anthropomorphism research by defining more precise conceptual boundaries is not a feasible endeavor. From a descriptive perspective, anthropomorphism research should continue to explore the infinitesimal variations of how children and adults engage socially with human or non-human entities in their environments, and to be informed or inspired by a general familiarity with human-human engagements -but not as an absolute and a priori defined benchmark. --- Pragmatistic View on the Normative Research Perspective The above discussed pragmatistic view on anthropomorphism draws special attention to the normative issue of what humans deem worth engaging with, and whether granting the status of social engagement partners to technological entities could not only ennoble the technology but also degrade the human by deflating the unique social value of human-human engagements )). However, can these concerns stand up to pragmatistic scrutiny? The basic ethical criticism of endowing technology with humanlike design components is an inheritance from Descartes' dualism ), especially regarding the strict ontological dichotomy of the mindful and the mindless, or the human and the non-human . One of the most influential criticisms following this Cartesian legacy has been Turkle's ) hierarchical distinction between technology's inferior simulation of mentation, intentionality or emotionality, and the superior genuineness of human mentation, intentionality or emotionality . Pragmatism is generally suspicious of drawing such apodictic conclusions from Cartesian a priori dualisms , and it asks whether such an uncompromising hierarchization of superior human-human and inferior human-technology engagements can be justified by the normative virtues of human experience. This seems, at best, questionable, because such a hierarchization would ignore that, at times, humans could systemically prefer to engage with humanlike technology due to, rather than despite of, the absence of what they think constitutes 'true' humanness. For example, there is comprehensive clinical research on how humans are more willing to self-disclose sensitive personal information to technology , and most findings suggest humans systematically prefer to self-disclose more sensitive personal information with nonhuman technological entities . Given technology's absence of moral judgement, this may not be a convincing point to make on its own, because the same reasoning would also apply to non-judgmental diaries people write. However, in combination with other empirical findings, one could argue such self-disclosure patterns reflect a more general tendency of humans to differentiate and appreciate certain variations of social engagements according to the perceived strengths and weaknesses of human or non-human engagement partners. For example, Logg et al. showed how humans tend to place more trust in predictions and judgements from a technological entity , while, according to Ha et al.'s experimental research on privacy concerns and self-disclosure, humans tend to mistrust technological entities which seem too humanlike , or which present themselves as 'partners' rather than 'servants'. Similar nuances can also be found in children's engagements with technological entities. For example, some children in Turkle's ethnographic studies provided well-argued reasons why the programmable nature of technology would make them more reliable, consistent and trustworthy than humans. When it comes to DVAs, in particular, the exploratory qualitative findings by Festerling & Siraj on children's open engagements with DVAs suggest children seem to appreciate the instant social gratification and excitement they experience with DVAs, and they also associate DVAs with relative ontological strengths. For instance, children systematically attributed higher accuracy levels and faster response times to DVAs for knowledge-related domains of intelligence , and explained their attribution patterns by DVAs' connectedness to the internet and their programmable nature . This is in line with other empirical findings in the literature on children's differentiated perceptions of technology as data-based knowledge sources ). However, Oranç & Küntay found even when children think technological entities are intelligent enough to answer questions related to mechanical or electronic issues, for biological and psychological issues , children still prefer humans as knowledge sources. Similarly, Festerling & Siraj also found children associate other domains of intelligence with humans , which is further in line with Xu et al.'s experimental findings on how children seem to elevate the intelligibility of their speech according to DVAs' perceived conversational weaknesses. Lastly, Yip et al.'s found although children expect a DVA to make them laugh in response to certain commands , a DVA that would laugh itself was thought of as being utterly disturbing. In summary, the pragmatistic reading of these empirically observable nuances is that children's engagements patterns with technological entities already follow nuanced understandings and expectations regarding the entities' experienceable strengths and weaknesses. This further suggests human-technology engagements are irreducible experiences in their own right on the continuum of social engagements, and, vis-à-vis human-human engagements, the relationship is as much about relative differences as about similarities. Importantly, although this pragmatistic line of reasoning challenges traditional ethical criticisms which emerge from dualistic viewpoints on reality, it does not negate ethical issues. As Damiano & Dumouchel show in their account of 'synthetic ethics', creating technology inspired by an understanding of, and knowledge about, humanness requires ethical sensitivity towards concrete issues which could arise in the context of application, and a solution-oriented attitude to address these issues without condemning humanlike technology as a whole. For example, the above discussed use of technology in clinical contexts prompts a question about which humanlike design components could compromise a technology's usability and effectiveness ) which may also be true for other professional contexts . Furthermore, child-DVA engagements raise additional questions about how DVAs are supposed to handle morally sensitive situations: these include offensive voice commands , the close emotional attachments vulnerable children may develop with them , and the potential societal harm from using primarily female voice-interfaces . --- Pragmatistic View on the Explanatory Research Perspective In a recent experimental study , van Straten et al. hypothesized children's humanlike perceptions of a robot were due to insufficient awareness about the robot's lack of 'true' humanness, as exemplified by its programmability. To address the ambiguity of their empirical findings, van Straten et al. conclude: Perhaps it can simply not be assumed that children understand that a humanoid robot is more similar to a technology than to a human being, even if this is pointed out repeatedly and even if children understand that a robot is not humanlike in terms of psychological capacities . But what if researchers' claims that children anthropomorphize technological entities were simply artefacts of false understandings of what is unique to humans and technology -'false' in the sense it differs from the perspective of today's children who develop their own understandings of these concepts. In other words, the nature of anthropomorphism could change as humans' experiential understandings of humanness change, and this may particularly apply to today's children who are growing up amid technologies which, for example, can emulate the autonomous use of human language and speech. In this case, the claim children anthropomorphize DVAs would be an artefact of an 'anachronistic' understanding of humanness and technology -at least from the perspective of children who have been intimately exposed to these voice-enabled technological entities from the beginning of their lives. Or, as put more generally by Damiano and Dumouchel , p.4), with technological progress blurring the experienceable boundaries between humans and technology, "the question of what constitutes human identity, or particularity, is raised anew". At its core, this is a pragmatistic line of reasoning regarding the explanatory origins of anthropomorphism, because it prioritizes experiential understandings of humanness over a priori derived ones. But what are the supporting arguments for this reasoning? According to Epley et al.'s previously mentioned three-factor theory, the process of cognitive inference which causes anthropomorphism psychologically is anchored in the inductive base of humanness . However, could children's experience of engaging socially with DVAs from the very beginning of life impact the development of this inductive base of humanness? It is beyond the scope of this review to provide a definite proof for this hypothesis, but there are two noteworthy empirical studies that can support this proposition. First, in a recent neuroscientific study, Waytz et al. investigated anthropomorphism in amygdala-damaged patients to isolate two neuronal processes: the bottom-up process of anthropomorphism based on overt social behaviors , and, in the absence of overt social behaviors, the top-down process based on abstract semantic knowledge about socially meaningful stimuli in the environment . As noted by Waytz et al. , these findings support the general notion that anthropomorphism, as part of social cognition, offers different routes, and that at least one of these routes seems to be based on things humans gradually learn through social experiences as they develop. Second, in a recent experimental study, Brink et al. investigated how the 'uncanny valley' unfolded in a developmentally diverse sample . In particular, the study examined whether the uncanny valley constitutes an innate aversion due to evolutionary determined responses to physiological illnesses and defects, which would further imply that even the youngest infants in the sample should display aversions towards overly humanlike technology, or whether it constitutes a violation of postnatally acquired expectations and norms about humanness and technology, which are only developed throughout childhood as part of social learning experiences . The empirical results revealed an unambiguous pattern: aversions towards overly humanlike technology seem to develop with age, but not before middle childhood , therefore providing strong empirical support for the latter origin of the uncanny valley . Taken together, both studies suggest the human stance towards technology is at least partly shaped by children's developing understandings of the social realities they face. In consequence, this would imply that the inductive base of anthropomorphism, as circumscribed in Epley et al.'s three-factor theory, could also change through changing social realities. The reality for many children today is both that they can engage with technology through human language and speech, and that these technologies are firmly embedded in social life at home . This pragmatistic view on anthropomorphism echoes previous discussions in the literature on Kahn et al.'s ) 'new ontological category hypothesis', stating increasingly intelligent technologies could become an independent and developmentally stable ontological category in its own right which cuts across traditional ontological dichotomies . Implications, challenges and potential empirical validations of the new ontological category hypothesis have already been discussed elsewhere ), and also in the particular context of DVAs ). However, this review adds that the new ontological category hypothesis is philosophically grounded in a pragmatistic paradigm, because the underlying assumption is that children's developing ontologies, as reflected by their conceptual understandings of the world around them, may not converge to an a priori definable and metaphysically 'true' end-state. In the words of James , p.114): "[o]ur ideas and concepts and scientific theories pass for true only so far as they harmoniously lead back to the world of sense". --- Conclusion, Limitations, and Future Research This conceptual review article, addressed to researchers interested in anthropomorphism and adjacent areas related to human-technology or child-technology engagements, has aimed to provide a conceptually framed account of contemporary anthropomorphism research based on three different research perspectives . Moreover, it has applied a pragmatistic viewpoint to discuss how these perspectives may contribute to a scientific understanding of children's engagements with increasingly intelligent and omnipresent technologies such as DVAs. The pragmatistic viewpoint sheds a new light on widely held views in the literature , culminating in the argument that, from an explanatory perspective, the nature of anthropomorphism may change as humans' experiential understandings of humanness change, and that this may particularly apply to today's children as their social cognition develops in interaction with technological entities which are increasingly characterized by unprecedented combinations of human and non-human qualities. Besides the methodological limitations inherent to conceptual reviews , the article's main limitation lies in the trade-off choice between analytical scope and depth: to increase the scope of the discussion, the article referred to 'technology' as a rather broad category comprising various types of technological entities and social engagements , while special attention was only paid to the peculiarities of children's engagements with DVAs. Narrowing the scope to certain types of technology or social engagements would have partly increased the depth of the discussion. However, the aim of this article was to look at the bigger picture of anthropomorphism research investigating humans' engagements with technology, in general, and children's engagements with technology, in particular. As such, the question whether the technology under investigation is specifically designed for social purposes is not of primary importance, because, as the above discussion suggests, there are infinitesimal variations of how technology can be anthropomorphized, and these occur across a broad range of design and research contexts. The article points in three main directions for future research. First, recent empirical studies which investigated traditional anthropomorphism issues in humantechnology research from a developmental perspective ) already showed how the investigation of children can contribute to the research field more generally. Future research could use similar cross-sectional or experimental age group comparisons to investigate how children's daily exposure to certain types of technology may influence such developments. For instance, Bernstein & Crowley's original, and, up to this point, un-replicated finding that children born in the early 2000s with high exposure to robotic technologies have a different conceptual understanding of intelligence and aliveness raises overdue empirical questions in light of hundreds of millions of DVAs being installed in home and childhood environments across the globe in the early 2020s. Following the pragmatistic viewpoint of the above discussion, such future research should be less concerned with the question whether children develop an 'objectively true' understanding of technology ) and more open towards new ways about thinking about their increasingly technologized environments . Second, research on children's understandings of state-of-the-art technology is trying to capture a fast-moving target, especially in the context of commercially available technologies such as DVAs. Unlike many other types of technologies often investigated in anthropomorphism research, DVAs not only stand out in terms of how fast they have populated home and childhood environments at large scale but also how fast their designs and technological capacities to emulate human language and speech change. Therefore, DVAs remain a highly relevant yet challenging case for research in their own right, and particular attention should be paid its technological development. For instance, so far DVAs' human-like design has been limited to the autonomous use of human language and speech, but smart speakers could soon also be able to move autonomously . The utilitarian importance of such technological developments may be unclear, but given coherent motion patterns are a salient stimulus for innate social cognition in their own right ), they could be important in children's developing understandings of DVAs. In this context, children with disabilities and their caregivers will be a subpopulation of particular importance, given DVAs' potential to inform, assist, assess and support these individuals with special developmental or medical needs . Preliminary research already suggests commercial off-the-shelf DVAs are popular among individuals with disabilities ), but research is still limited, especially when focusing on concrete disabilities ). However, apart from general socio-economic and skill-related accessibility issues , children with different kinds of disabilities will experience different kinds of benefits and challenges. For instance, visually impaired children are likely to benefit from the functional advantage of DVAs' hands-free and 'eye-free' voice interfaces, but, at the same time, today's DVAs usually lack the customizability to address certain special needs, such as longer response times, faster speech rates, or custom voice commands . Nevertheless, future research should not only consider DVAs as a functional means to a social end for children with disabilities ), but also as a social end in itself ). Third, the pragmatistic view on anthropomorphism can be extended by what neopragmatists ) say about the critical role of language in inquiry. In fact, human cognition is significantly shaped by the languages spoken , and in the context of anthropomorphism, determining the 'true' meaning of what humans say about or to non-human entities remains challenging from a research perspective . For example, the use of gendered personal pronouns in human-DVA engagements has been interpreted as a direct signifier for the personalization of DVAs ). Future research could be more critical of the role of language in anthropomorphism by examining the psychological mechanisms which cause the development of particular language patterns in human-technology engagements . Lastly, it is worth pointing out that our discussion is 'bent towards a particular perception of science'. In particular, the notion of pragmatism we follow lends itself to an empirically grounded perspective to look at things, that is, a perspective which favors "a [constructivist] model of children as builders of their own intellectual structures" , rather than a model of children as recipients of a priori defined intellectual structures we already built for them. Developmental constructivism -at least in its original epistemological form -has traditionally been criticized for not having a socio-cultural lens , and readers who feel more inclined towards this criticism will naturally entertain doubts about the ideas and arguments discussed above. Yet, we hope our paper widens the thinking on children's development in the context of modern everyday technology, and how social cognition of today's children develops in interaction with technological entities which are increasingly characterized by unprecedented combinations of human and non-human qualities. --- --- --- --- Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. --- --- Authors and Affiliations --- Janik Festerling 1 • Iram Siraj 1 Iram Siraj [email protected] 1 Department of Education, University of Oxford, 15 Norham Gardens, Oxford OX2 6PY, UK

Anthropomorphism' is a popular term in the literature on human-technology engagements, in general, and child-technology engagements, in particular. But what does it really mean to 'anthropomorphize' something in today's world? This conceptual review article, addressed to researchers interested in anthropomorphism and adjacent areas, reviews contemporary anthropomorphism research, and it offers a critical perspective on how anthropomorphism research relates to today's children who grow up amid increasingly intelligent and omnipresent technologies, particularly digital voice assistants (e.g., Alexa, Google Assistant, Siri). First, the article reviews a comprehensive body of quantitative as well as qualitative anthropomorphism research and considers it within three different research perspectives: descriptive, normative and explanatory. Following a brief excursus on philosophical pragmatism, the article then discusses each research perspective from a pragmatistic viewpoint, with a special emphasis on child-technology and child-voice-assistant engagements, and it also challenges some popular notions in the literature. These notions include descriptive 'as if' parallels (e.g., child behaves 'as if' Alexa was a friend), or normative assumptions that human-human engagements are generally superior to human-technology engagements. Instead, the article reviews different examples from the literature suggesting the nature of anthropomorphism may change as humans' experiential understandings of humanness change, and this may particularly apply to today's children as their social cognition develops in interaction with technological entities which are increasingly characterized by unprecedented combinations of human and non-human qualities.

IntrODuctIOn Youth smoking has remained a public health challenge worldwide including Hong Kong, a city of China with a low daily smoking prevalence of 10.5%. 1 Recent figures have shown an upward trend in ever smoking rate among local primary students. 1 The age of smoking initiation is also declining in other regions in China. [2][3][4] Early smoking predicts long-term tobacco use, 5 more difficult quitting 6 and higher risk of all-cause mortality in adulthood. 7 Identifying risk factors of smoking initiation in Chinese youth has major local and global health implications. The family has important influence on youth smoking. Lower family socioeconomic status and having smoking family members predict while parental monitoring and authoritative parenting protect against youth smoking initiation. [8][9][10] Other family-related risk factors include non-intact family structure, low parental education level and parental approval of smoking. 9 Studies in Hong Kong also show that parental smoking, exposure to secondhand smoke at home and overestimation of peer smoking prevalence were prospectively linked to youth smoking initiation. 11 12 Associations of perceived interparental relationship, family harmony and family happiness with smoking intention in never-smoking Chinese children and adolescents: a crosssectional study Tzu Tsun Luk, 1 Man Ping Wang, 1 Lok Tung Leung, 2 Yongda Wu, 1 Jianjiu Chen, 2 Tai Hing Lam, 2 Sai Yin Ho 2 Open Access Despite the importance of family relationship in Chinese culture, little is known about its role in youth smoking initiation. Under collectivism, family harmony is deemed the foundation of family functioning in the Chinese culture. 13 It is the ideal state of family relationship emphasising the sense of togetherness and absence of intrafamilial conflict through mutual respect and communication. 13 14 Family happiness is the outcome of a harmonious family in which members care and support each other and feel emotionally secured. 13 Conflict between family members is a major threat to family harmony, which in turn compromises family happiness. 13 14 Children who have witnessed conflicts between their parents and perceived family as conflictual and unhappy have greater risk of emotional distress, depression and maladjustment, [15][16][17] all of which have been shown to predict youth smoking initiation in Westerners 18 19 and Chinese. 20 Many studies have found youth smoking initiation associated with negative parentchild relationship, [21][22][23][24] but evidence on its association with interparental relationship is lacking. Never has the role of family happiness in youth smoking initiation been reported. A cross-sectional study investigated how family disharmony was associated with ever smoking in Chinese, but the study was small and included adolescents mostly aged 13. 25 Family harmony has remained a core value and an index of societal stability in Hong Kong. 26 Studying the determinants of youth smoking in Hong Kong helps develop culturally appropriate smoking prevention programmes that may also benefit mainland China and elsewhere. 27 The theory of planned behaviour postulates that behavioural intention precedes the actual behaviour, 28 as have been observed in youths intended to initiate smoking. 9 29 We therefore tested the associations of perceived interparental relationship, family harmony and family happiness with smoking intention in two large, population-representative, cross-sectional samples of never-smoking Chinese children and adolescents in Hong Kong. --- MethODs study design We analysed cross-sectional data from the School-based Survey on Smoking among Students 2012/2013 collected during October 2012 to April 2013 in Hong Kong. Detail of the study design has been reported elsewhere. 30 Briefly, schools were randomly selected from all 18 districts in Hong Kong in proportion to the total number of schools in the respective districts. All grades 4-6 primary students and grades 7-12 secondary students in the selected schools were selected. Invitation letters were sent to their parents via the students for passive consents, where declining parents asked the student to return an empty questionnaire. Students voluntarily completed an anonymous questionnaire in Chinese, which was immediately collected and sealed in an opaque envelope by research personnel in front of the students. Core questions in the survey were adapted from Global Youth Tobacco Survey. Ethical approval was granted by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster. A total of 16 316 students from 71 primary schools and 45 857 students from 75 secondary schools completed the survey. Response rates were about 96% at the student level and 20% at the school level. School refusals were largely due to administrative issues . We randomly sampled schools with similar funding source, teaching medium and sex composition in the same district for replacement of the rejected school. --- Main measures Outcome measures Similar to previous studies, we measured smoking intention with two items: 'smoke in the next 12 months' and 'smoke if one of your good friends offers you a cigarette' each with four-point response options of 'definitely not', 'probably not', 'probably will' and 'definitely will'. 29 31 Students who reported 'definitely not' for both questions were classified as having no intention to smoke and otherwise as having an intention to smoke. --- Exposure measures As the first population study on family relationship and smoking in Hong Kong, we developed three questions based on our findings from two local qualitative studies on family well-being. 13 14 One question assessed perceived interparental relationship: "What do you think is the relationship between your father and mother?" with five-point responses of 'very good', 'good', 'fair', 'bad' and 'very bad' and 'not applicable'. Family harmony was measured with students rating the item 'My family gets along well' on a five-point scale of 'strongly agree', 'agree', 'fair', 'disagree' and 'strongly disagree', 32 which were renamed as 'very good' to 'very bad' for presentation. Perceived family happiness was assessed by asking: "All things considered, you think your family is:" with response options of 'very happy', 'happy', 'not very happy' and 'not happy at all'. These three items had satisfactory internal consistency in our samples and 8-day test-retest reliability in a separate sample of 329 primary students with intraclass correlation coefficient of 0.85 for interparental relationship, 0.67 for family harmony and 0.73 for family happiness. We created a composite variable namely number of poor family relationship factors by summing the perceived interparental relationship, family harmony and family happiness . The three items were also summed to give a poor family relationship score which ranged from 0 to 11. Although children's perceptions of family relationship may differ from those of other family members, 33 it should be their subjective assessment that would be more meaningful and relevant to the development of their smoking intentions. --- Open Access --- Potential confounders We also collected data on age, sex and perceived family affluence. The advantage and reliability of perceived family affluence as a proxy measure of family socioeconomic status for youths in Hong Kong has been reported previously. 34 We also assessed family structure by asking if the students were living with their mothers and fathers, with responses classified into intact , single-parent and no-parent . 35 Open Access Figure 1 Adjusted ORs and 95% CI of smoking intention by categories of perceived interparental relationship, family harmony, family happiness, and family relationship in primary students. of smoking intention in relation to perceived interparental relationship, family harmony and family happiness and the two composite poor family relationships using complete-case analysis. As having smoking household members may contribute to perceived family unhappiness in children 36 and family disharmony, 37 all regression analyses were adjusted for parental smoking and sibling smoking in addition to age, sex and perceived family affluence and family structure. The F-adjusted mean residual goodness-of-fit tests verified that all adjusted logistic regression models have satisfactory fit . 38 The three family factors were not included in the same model because moderate to relatively strong correlations among the three family factors were observed in both primary and secondary samples . All predictor variables were also analysed as continuous variables to test their linear relationships with smoking intention . A two-sided p<0.05 denotes statistical significance. --- results --- Primary students The prevalence of smoking intention in all primary neversmokers was 4.7% . Table 1 shows the sample characteristics. Students with smoking intention were significantly older than those without smoking intention and were associated with perceived poorer family affluence and higher number of smoking parents but not with sex , intactness of family structure or sibling smoking . The prevalence of smoking intention generally increased with more negative responses in perceived interparental relationship, family harmony and family happiness , and poor family relationship . After adjusting for age, sex, perceived family affluence, family structure, parental smoking and sibling smoking, strong graded associations between each of the three family factors and smoking intention were observed . Higher odds of smoking intention were observed in primary students with 'good' to 'very bad' perceived interparental relationship and family harmony relative to those reporting 'very good' interparental relationship and family harmony. The corresponding AORs ranged from 2.84 to 5.40 for students perceiving family happiness as 'happy' to 'not happy at all' . Significant graded association between poor family relationship and smoking intention was observed with AORs increasing from 1.80 for one factor to 3.95 for three factors . Each unit increase in poor family relationship score was associated with 26% increased odds of smoking intention. --- secondary students For never-smoking secondary students, 11.8% reported an intention to smoke, which was associated with male sex , poor perceived Open Access family affluence , non-intact family structure , having more smoking parents and sibling smoking . Mean age was similar in students with or without smoking intention . Overall, the prevalence of smoking intention increased with more negative responses in interparental relationship, family harmony and family happiness, and poor family relationship . We found significant dose-response relationship between the three family factors and smoking intention , after adjusting for sociodemographic factors, parental smoking and sibling smoking. The AORs ranged from 1.24 to 2.15 for more negative perception of interparental relationship, 1.33 to 2.98 for family harmony and 1.37 to 2.61 for family happiness . The AORs of smoking intention increased with poor family relationship factors with significant graded relationship , from 1.91 for one factor to 2.70 for three factors . Each unit increase in poor family relationship score was associated with 19% increased odds of smoking intention. To test the stability of the estimates, we further add number of smoking peers , attitude towards smoking and knowledge in smoking harm in the multivariable models. Additional adjustment for these variables did not change the estimates substantially in both primary and secondary students . --- Open Access --- DIscussIOn This study found strong, significant associations of interparental relationship, family harmony and family happiness with smoking intention in both never-smoking Chinese children and adolescents with dose-response relationships. The result for family harmony is consistent with that of the small cross-sectional study on family disharmony and ever smoking in Chinese adolescents in Wuhan, 25 and studies in other cultures that generally show family conflict is predictive of youth smoking initiation. 21-24 39 Our study contributes to previous research by showing that the association may be dose dependent and extending the understanding of interparental relationship, an understudied family factor, as a significant determinant of youth smoking initiation. The ORs were very similar with or without adjustment of potential confounders, suggesting that the observed associations were unlikely attributable to confounding effects and that family relationship may independently predict smoking intention in Chinese never-smoking youths. The robust findings on different indicators of family relationships suggest their significant influences on smoking intention in Chinese children and adolescents. Our findings were based on Chinese never-smoking children and adolescents. As risks of future smoking in relation to psychosocial predictors are higher in ever smokers than never-smokers, 40 the association between family relationship and smoking intention may be stronger in students who have experienced cigarettes. The social significance of family may render family relationship a more influential determinant of smoking intention in youths from China and other similar cultures in Asia than their Western counterparts. Nevertheless, given the high incidence of divorce in Western nations with USA having the highest rate and Europe's on an increasing trend, 41 testing the role of interparental relationship on smoking initiation in Western youth population is warranted. Poor family relationship may lead to smoking in children and adolescents through different pathways. As described earlier, psychosocial stressors may mediate children's poor perceptions of family environment and smoking initiation. However, impaired parenting and parent-child relationship may also be important mediators. Conflict between parents can disrupt parenting routines, 42 which may render parental monitoring, parental disapproval of smoking and other parenting practices ineffective against youth smoking initiation. The negative emotion of parents in hostile relation may also transfer or 'spillover' to children and compromise parent-child relationship. 43 Weak bonding with the family in turn predisposes the children to attachment with deviant peers who may promote smoking. 44 Empirical evidence from neurocognitive research supports the role of poor family relationship in adolescent smoking. Social experience can modulate the development of two neural systems in adolescents: 45 46 the cognitive control system, which regulates behaviour but remains immature throughout adolescence, and the reward processing system, which has heightened sensitivity during adolescence and underlies youths' increased temptation to try new, potentially pleasurable yet risky activity such as smoking. 47 Open Access improve adolescents' cognitive and decision-making abilities to regulate their behaviours, which were linked to reduced tendency to take risk. 48 In contrast, adolescents from families with more conflicts and poorer cohesion had longitudinal impairment in cognitive control and increased risk-taking propensity. 49 Parents should be warned that conflict among themselves or family disharmony and unhappiness may contribute to smoking ideation in children. They should be encouraged to seek appropriate help to manage unresolved family conflicts. Our findings also suggest that children's perceptions of their family relationship may be an important intervening point for preventing youth from initiating smoking. This study has several limitations. First, the cross-sectional design of the survey precludes causal inference of interparental and family relationship on children and adolescent smoking intention. Prospective studies are warranted to confirm the results. However, it is unlikely that youth smoking intention, which is not easily observable in the family, could affect interparental relationship. Second, although confidentiality is reassured to encourage candid reporting, reporting bias remained a distinct possibility. The sensitive nature of the questions could lead to random under-reporting, which might bias the associations towards the null. Third, the low school-response rate might contribute to non-response bias. However, schools' rejections were largely due to administrative issues, which was not related to student smoking. Furthermore, based on the Hong Kong 2012/2013 student enrolment statistics, our student samples were comparable to their underlying population in sex , age and grade . 50 The moderate effect size observed for age in primary students was attributed to the difference in reference time frame between the official statistic and data collection as students reported their age at the time of completing the survey. cOnclusIOns Perceived lower levels of interparental relationship, family harmony and family happiness were associated with increased odds of smoking intention with doseresponse relationships in Chinese children and adolescents. Prospective studies are warranted to confirm the findings. --- competing interests None declared.

Objective To examine the associations of perceived interparental relationship, family harmony and family happiness with smoking intention in never-smoking Chinese children and adolescents in Hong Kong. Design, settings and participants Cross-sectional surveys of 15 753 primary (grades 4-6) and 38 398 secondary (grades 7-12) never-smoking students from 71 to 75 randomly selected primary and secondary schools in Hong Kong, 2012-2013. Measurements Outcome variable was smoking intention which denoted any affirmative response to smoke within the coming year or when a cigarette was offered by a good friend. Exposure variables were perceived interparental relationship and family harmony each measured on a fivepoint scale from 'very good' to 'very bad' and perceived family happiness on a four-point scale from 'very happy' to 'not happy at all'. Potential confounders included age, sex, family structure, perceived family affluence, parental smoking and sibling smoking. results In primary students, the adjusted ORs (AORs) (95% CI) of smoking intention generally increased with more negative perception of the family relationship: up to 3.67 (1.91 to 7.05) for interparental relationship, 7.71 (4.38 to 13.6) for family harmony and 5.40 (3.41 to 8.55) for family happiness. For secondary students, the corresponding AORs (95% CI) were 2.15 (1.64 to 2.82) for interparental relationship, 2.98 (2.31 to 3.84) for family harmony and 2.61 (1.80 to 3.79) for family happiness. All p for trend <0.001. conclusions More negatively perceived interparental relationship, family harmony and family happiness were associated with higher odds of smoking intention with doseresponse relationships in never-smoking Chinese children and adolescents in Hong Kong. Children's perception of their family relationship may be an important intervening point for preventing youth from initiating smoking.

This article engages in a historical-comparative analysis of the political origins and consequences of unemployment insurance systems. Its basic motivation is the following. In most long-standing democracies, organized workers are more likely to vote and more likely to support Left parties than unorganized workers . Moreover, many comparative welfare-state scholars agree that working-class organization has been an important factor in the development of redistributive social policy, partly through its effects on the partisan composition of government . The obvious question arises: Why is it that some working classes are better organized than others? Comparative studies of unionization show that countries with unemployment insurance funds administered by unions-so-called Ghent systems of unemployment insurance-tend to be more unionized than countries with stateadministered systems. It is a commonplace that Ghent systems create selective incentives for workers to join unions. Existing literature thus suggests that the question of working-class organization might usefully be reformulated as a question about the design of unemployment insurance: Why is it that some countries ended up with Ghent systems of unemployment insurance while others ended up with state-administered systems? Could it be that Left parties introduced Ghent systems as a means to improve their own political prospects? According to Rothstein ), long-term strategic thinking played an important role in the decision of the Swedish Social Democrats to introduce a Ghent system of unemployment insurance in 1934. In Rothstein's words, the Swedish case illustrates how "state capacity and policy might be deliberately constructed by political agents . . . as a means of promoting the mobilization of workers" . We argue that the conventional wisdom about the relationship between Ghent and unionization needs to be qualified. Implementing a difference-indifference design with data for 20 to 21 countries over the period 1870 to 1970, we find some evidence that the adoption of Ghent systems of unemployment insurance followed increases in union density and no evidence whatsoever that the introduction of Ghent systems accelerated unionization. Until the 1950s, Ghent systems relied primarily on contributions by the insured and provided meager benefits. As a result, they failed to incentivize workers to join unions. With data for 1930 to 2010, we show that the Ghent effect identified by the existing literature pertains only to Ghent systems that are financed by employer contributions or general tax revenues. Turning to the partisan politics of unemployment insurance design, Sweden is the only case of Ghent being introduced by a Left government . Left governments more commonly opted for state-administered unemployment insurance and in several instances initiated the replacement of Ghent systems by state-administered systems in the 1930s and 1940s. While Rothstein is interested in establishing the potential for long-term strategic behavior by Left parties rather than predicting what Left parties do under normal circumstances, we seek to articulate a positive theory of partisan preferences for unemployment insurance design. We argue that the primary concern of Left parties in the first half of the 20th century was to provide unemployment insurance coverage to their working-class electoral base. In most countries, unionization was too low for Ghent to be a viable alternative to state-administered insurance for the Left. We argue further that the Ghent story captures a social-liberal moment in the development of the welfare state that has largely been ignored by the power resources tradition of comparative welfare-state scholarship. Liberal parties-and, to a lesser extent, Christian Democratic parties as wellresponded to early union organizing and the associated socialist threat by proposing Ghent-style unemployment insurance. In so doing, they sought to appeal to newly enfranchised workers and to cultivate ties with craft unions. Ghent was an attractive option for progressive Liberals because the scheme was voluntary and, as envisaged at the time, did not entail any significant redistribution of income in favor of workers. The Swedish case is exceptional not only on account of the weakness of liberalism, but also, and above all, on account of the high-level unionization that had been achieved before the Social Democrats came to power in 1932. Our account stands Rothstein on his head in that unionization explains why it was rational for the Swedish Social Democrats to embrace the Ghent design of unemployment insurance in the 1930s. Left parties in Belgium, Denmark, and Finland did not introduce Ghent systems, but they were important actors in postwar reforms that rendered these Ghent systems more union friendly, as were the Swedish Social Democrats. In due course, we will present evidence indicating that Left government is associated with less reliance on contributions by the insured to finance Ghent systems. Setting the Swedish case aside, the Ghent effect identified by comparative studies of unionization might thus be seen as the product of a two-step historical sequence in which Liberal parties introduced Ghent systems between 1900 and 1930 and Left parties reformed Ghent systems in the 1950s and 1960s. It is tempting to think of the Ghent effect as an unintended consequence of institutional choices made by Liberals, but what we propose here is not a story of "endogenous institutional change." Postwar changes in the financing of Ghent systems-the second step in the historical sequence-resulted from shifts in the distribution of political power that were not, at least not in any obvious way, a consequence of the presence of Ghent systems. The rest of the article is organized as follows. We begin by reviewing existing literature and setting out our core arguments, addressing the effect of unemployment insurance on unionization in the section "Ghent and Unionization" and the partisan politics of designing unemployment insurance in the section "The Partisan Politics of Ghent." We introduce our coding of unemployment insurance systems as "Ghent" and "non-Ghent" and briefly describe the financing of unemployment insurance in the section "Unemployment Insurance Systems: Descriptive Data." In presenting our statistical results, we again separate the question of the Ghent effect on unionization from the question of partisan effects on the institutional design and financing . In the section "Conclusion," we conclude by articulating how our arguments relate to historical institutionalism. --- Ghent and Unionization Ghent systems of unemployment insurance originate in voluntary insurance schemes established by craft unions in the late 19th century. For obvious reasons, private insurance against unemployment was never really a viable option and union schemes commonly became insolvent in the face of mass unemployment. In 1901, the city of Ghent intervened to subsidize union unemployment funds. A number of other municipal and provincial governments followed suit in the early 1900s. France became the first country to subsidize union-administered unemployment funds at the national level in 1905. 1 While some Ghent systems created by national governments over the following three decades retained the principle of voluntary insurance, others required workers to insure themselves with a union-administered fund. As emphasized by Rothstein and other scholars, what all of these systems have in common boils down to this: Unions set the conditions of eligibility for unemployment benefits and participate directly in the process whereby it is decided whether or not unemployed workers satisfy these conditions. It is not hard to see that such a system incentivizes individuals to join unions. Even if it is possible to belong to a union-administered unemployment fund without joining a union, many workers are not aware of this option and those who are aware of it will likely look upon union membership as a means to secure access to full unemployment benefits. 2 More positively, as noted by Western , a key feature of Ghent systems is that they encourage workers to remain union members when they become unemployed and allow unions to take credit for the unemployment benefits being provided. By our count, there are at least 20 comparative studies that invoke the Ghent design of unemployment insurance to explain rates of unionization. 3 Relying entirely on postwar data , these studies effectively compare unionization rates in countries that do and do not have Ghent systems and fail to address the question of whether or not the introduction of Ghent systems affects the rate of unionization. Based on the existing literature, we cannot rule out that countries with strong unions adopted Ghent systems, whereas countries with weak unions ended up with state-administered unemployment insurance. Cognizant of the limitations of his statistical analysis, Western invokes the comparison of Sweden and Norway as a "quasi experiment" that demonstrates the positive effect of having a Ghent system for the rate of unionization . While Sweden introduced a Ghent system in 1934, Norway replaced its Ghent system by state-administered compulsory unemployment insurance over the period 1938 to 1946. According to Western, union density in the two countries rose in parallel from 1921 to 1938 while union density grew more rapidly in Sweden than in Norway over the 1940s. Drawing on different sources and covering a longer stretch of time, the data on unionization presented in Figure 1 convey a different picture. To begin with, Swedish union density grew steadily from 1910 to 1960 and there is no indication that the introduction of Ghent accelerated the process of unionization. In Norway, union density increased following the introduction of a Ghent system in 1907, dropped in the early 1920s, and began to climb again from the late 1920s onward. Again, there is no indication that the dismantlement of Ghent adversely affected unionization. Most importantly for our purposes, Figure 1 shows that a significant gap in unionization between Sweden and Norway opened up between 1920 and 1934, when Norway had a Ghent system and Sweden did not have any unemployment insurance at all. 4 It deserves to be noted that many Swedish unions boycotted the Ghent systems until 1941, when a further reform increased the level of subsidization ). In the existing literature, the negative or ambivalent stance toward Ghent adopted by many unions in the interwar period emerges as a puzzle . This puzzle fades away once we recognize that Ghent legislation not only offered carrots to unions, in the form of subsidies, but also sticks, in the form of state regulation. In return for state subsidies, unions typically had to give up the use of unemployment insurance to support strikers. They also had to insure workers who were not union members and to submit to budgetary and legal control by state administrators. In light of such stipulations, state subsidies had to be relatively high to entice unions to join Ghent systems. We do not wish to challenge the idea that Ghent systems entail selective incentives for workers to join unions. Our point is rather that the extent of such incentives depends on the costs and benefits of the insurance being provided. We hypothesize that the selective incentives entailed by Ghent systems are inversely related to the extent to which workers themselves bear the costs of insurance. For short, we will refer to the inverse of self-financing as "subsidization." As we shall see, Ghent systems relied more heavily on selffinancing than state-administered systems in the interwar period. While some Ghent systems were replaced by state-administered systems in the 1930s and 1940s, others became more heavily subsidized, that is, more reliant of payroll taxes imposed on employers and/or general tax revenues, from the 1940s onward. The existing literature on the Ghent effect might thus be said to Source. Rasmussen . The vertical lines indicate the introduction and removal of Ghent systems. suffer from selection bias, conceiving the effect of a particular kind of Ghent system-"heavily subsidized Ghent"-as the effect of any and all Ghent systems. 5 --- The Partisan Politics of Ghent Turning to the partisan-political origins of national legislation introducing Ghent-style unemployment insurance, the Swedish story can be summarized as follows. 6 Prior to the 1930s, the Conservatives and Agrarians were the dominant political parties in Sweden. For different reasons, both parties were strongly opposed to unemployment insurance and repeatedly voted down legislative initiatives by Liberal and Social Democratic Members of Parliament in this domain. Having prepared legislation to introduce a Ghent system, the Liberal government of 1911 to 1914 fell, over defense policy and royal prerogatives, before the legislation was tabled. A Conservative government instead introduced a system of public relief works in 1918. This system became deeply resented by the labor movement in the course of 1920s, not only because it offered very low pay but also, and above all, because public relief workers were assigned to work at workplaces with ongoing industrial disputes. Not surprisingly, the introduction of unemployment insurance became a key priority of the union movement and the Social Democrats in the 1920s. While unions affiliated with Landsorganisationen increasingly favored mandatory insurance, the Social Democrats adopted a very pragmatic view of the institutional design of unemployment insurance. The Agrarian Party famously entered into a coalition government with the Social Democrats in 1933, but the coalition agreement left the question of unemployment assistance aside. Drafted by Gustav Möller, the Social Democratic minister of social affairs, the 1934 legislation that introduced Ghent-style unemployment insurance, and simultaneously abolished the commission in charge of relief work, was passed with the support of Liberals rather than Agrarians. Möller made extensive concessions to the Liberals, leaving union-administered insurance schemes seriously underfunded, and many unions responded by not applying for state subsidies. Further legislation in 1941, under a national unity government, increased subsidies and otherwise rendered the Ghent system more attractive to the unions. Relative to the standard accounts of the Swedish story, Rothstein's main innovation is to present documentary evidence showing that Möller explicitly argued to union leaders that they should support the 1934 legislation and join the new system because it would serve, over the long run, to boost unionization. The important thing, according to Möller, was to establish the Ghent principle: less desirable features of the reform would be improved upon as political conditions became more favorable to the labor movement. As noted at the outset, Rothstein's is a decidedly voluntaristrationalist account in the sense that Rothstein is interested in establishing the potential for long-term strategic behavior by Left parties rather than predicting what Left parties do under normal circumstances. That said, it is difficult to see, within the parameters of Rothstein's discussion, why any political party other than a union-based Left party would promote or support Ghentstyle unemployment insurance. The Norwegian case again provides for an interesting comparison with the Swedish case, generating alternative hypotheses concerning the partisan politics of Ghent adoption. 7 In Norway, it was a coalition government dominated by the Liberals that introduced state subsidies for union-administered unemployment funds in 1906. This legislation was a response to repeated demands by the main trade union confederation, expressed through sympathetic Liberal MPs, but many unions were disappointed with the specific provisions of the 1906 legislation. The unions objected to the fact that unorganized workers could participate in the new scheme and found the level of subsidies inadequate to cover bad risks among the unionized. Therefore, the main union confederation boycotted the new scheme for two years, a boycott that was only half-heartedly abandoned when government augmented subsidies somewhat in 1908. Like the Swedish Social Democrats, the Norwegian Social Democrats became the dominant government party in the mid-1930s. In contrast to their Swedish counterparts, however, the Norwegian Social Democrats opted to use their newfound power to introduce compulsory state-administered unemployment insurance. Interestingly, the 1932 election program of the Norwegian Labor Party explicitly argued that the coverage of union-administered unemployment funds was too low for the Ghent system to provide adequate solution to the problem of mass unemployment . 8 The Norwegian Ghent story exemplifies a broader phenomenon of liberal social policy initiatives in response to unionization and suffrage extension at the beginning of the 20th century . As emphasized by Luebbert and Boix , Social-Liberal parties such as the Liberals in the United Kingdom, the Radicals in France, and Venstre in Norway, competed with rising left parties for working-class votes, including the votes of organized workers, at this point in time. The Liberals were at a distinct disadvantage in this competition for they needed to appeal to workers in ways that did not undermine their support among small farmers and urban middle classes, and this meant that redistribution had to keep to a minimum. Not surprisingly, Social-Liberals targeted skilled workers, with relatively high wages and less exposure to unemployment risk, and were particularly responsive to the demands of craft unions. With craft unions being more able to sustain voluntary unemployment insurance schemes than industrial unions, subsidization of unemployment funds suited the electoral needs of Social-Liberal parties very well. The Ghent option also tallied well with their ideological disposition. With insurance being voluntary, at least legally speaking, the Ghent option exemplified the self-help and civil-society solutions favored by Liberals at the turn of the century. In addition, Ghent systems required very little state apparatus as most of the administrative work was carried out by the unions. The nature of the union movement arguably played a key role in determining the attitudes of Liberals toward unions and, by extension, toward unionadministered unemployment funds. When and where unions adopted the principle of industrial unionism, seeking to organize all workers by industry and engaging in militant tactics to put pressure on the state as well as employers, Liberals assumed or quickly realized that successfully appealing to organized workers would require more redistribution than what their middle-class supporters would countenance. In the Norwegian case, Venstre responded to the LO confederation's embrace of industrial unionism and syndicalism by abandoning its support for the Ghent system and proposing, in the late 1920s, to replace it with a state-administered scheme . One way to approach the puzzle generated by comparing Sweden and Norway is to ask which of these two cases is more representative. Has it been more common for Liberals or for Social Democrats to introduce Ghent systems of unemployment insurance? This framing of the problem is not altogether satisfactory, however, for we ought to aspire to an explanation that not only accounts for general patterns in the data, but also for cases that do not fit the general pattern. From this point of view, Rothstein's discussion leaves something to be desired. According to Rothstein , the Norwegian Liberals were mistaken in introducing Ghent and the Norwegian Social Democrats were equally mistaken in shutting it down. But Rothstein provides no clues as to why these strategic mistakes were made. By contrast, the explanatory framework proposed here provides insights into the Swedish case. From our perspective, the Swedish case is distinguished from other Ghent adopters by the weakness of progressive liberalism in the process of democratization. In marked contrast to the experience of neighboring countries, the Swedish Liberals held government power only once and only briefly, 1911 to 1914, prior to being displaced by the Social Democrats as the main party of the Left. Related to the weakness of liberalism, the Swedish union movement was from the beginning closely linked to the Social Democratic party and embraced industrial unionism earlier and more quickly than other union movements in Northwestern Europe . Most decisively, the preceding discussion suggests that rapid unionization in the course of the 1910s and 1920s made the Ghent system a more viable option for the Swedish Social Democrats than for other Left parties in the 1930s, allowing them to design unemployment insurance in a way that would have long-term political consequences while providing unemployment insurance coverage to a majority of workers in the near future. 9 As indicated at the outset, our quantitative analysis speaks not only to the partisan-political origins of Ghent systems, but also the partisan politics of financing unemployment insurance. Even if Left parties did not initiate the adoption of Ghent systems, they may have played an important role in generating the Ghent effect on unionization by promoting subsidization and more generous unemployment benefits in the postwar era. We can distinguish two versions of this argument. In one version, inspired by power resources theory , Left parties have promoted subsidization in Ghent and non-Ghent systems alike, but their distinctive preference for subsidized unemployment insurance has only translated into higher unionization in Ghent systems. In the second version, implying the kind of strategic behavior that Rothstein invokes to explain why the Swedish Social Democrats chose the Ghent option in the first place, the unionization effect under Ghent has led Left parties to push harder for subsidization in countries with Ghent systems, at the expense of other spending priorities . --- Unemployment Insurance Systems: Descriptive Data As indicated above, our argumentation pertains to two features of unemployment insurance: the administrative role of unions and the extent to which the costs of unemployment insurance are subsidized by employers and the state. In the quantitative analyses that follow, the administrative role of unions is captured by a dummy variable called "Ghent." Table 1 identifies the countryyears in our data set for which the Ghent dummy takes the value of 1 based on either a narrow or a broad definition. 10 Three hierarchical criteria were used to generate the sorting of countryyears presented in Table 1: whether or not national legislation provides for unemployment insurance, whether or not unions that are independent Belgium 1920-1944Czechoslovakia 1921-1939Denmark 1907-Finland 1917-1960France 1905-1950Netherlands 1916-1945Norway 1906-1938Spain 1919-1936Sweden 1934-1974Switzerland 1924-1951Belgium 1945-Finland 1960-Netherlands 1949-1951Norway 1938-1946Sweden 1974-Switzerland 1951-1976 United Kingdom 1911-1920Belgium 1920-Czechoslovakia 1921-1939Denmark 1907-Finland 1917-France 1905-1950Netherlands 1916-1951Norway 1906-1946Sweden 1934-Switzerland 1924-1976United Kingdom 1911-1920 Mean density 28. 72 38.11 51. 74 43.55 from the state play an important role in the administration of unemployment insurance, and whether unemployment insurance systems in which independent unions play an important administrative role are voluntary or compulsory. Taken together, the latter two criteria yield a threefold typology of institutional designs for unemployment insurance: compulsory insurance administered by state, with little or no involvement by unions; subsidized voluntary insurance administered entirely by unions; and compulsory state insurance with a significant administrative role played by unions. The distinction between systems with "little or no union involvement" and systems with "significant union involvement" is, of course, somewhat arbitrary. In particular, it should be noted that we code bipartite and tripartite systems, such as the Dutch system established in 1952, as systems with minor union involvement. It is important to keep in mind that the administrative role that the Ghent variable is meant to capture is one that makes it plausible for workers to suppose that whether or not they are union members affects their access to unemployment benefits. Our coding of the Netherlands as "non-Ghent" from 1952 onward is entirely consistent with existing literature as, indeed, is our coding of current unemployment insurance systems in all other countries included in Table 1. Relative to existing literature, our main contribution in this regard is to identify when different institutional designs were introduced and discontinued. It also deserves to be noted that all but one of the hybrid cases that make up our third type originated in voluntary Ghent systems . In voluntary systems, such as the Norwegian system of 1906, workers had to join union-administered funds to be insured against unemployment and these funds had to apply for state subsidies. In most cases, the law stipulated that the funds had to allow nonunionized workers to join to receive state subsidies but, again, these systems created a strong implicit linkage between unemployment insurance and union membership. We can loosely distinguish three variants of the hybrid system. Since 1944, the Belgian system is essentially a mandatory, state-financed insurance scheme administered by unions. Norway, between 1938and 1946, and Sweden, since 1974, exemplify a second variant, characterized by the coexistence of state-subsidized voluntary insurance managed by unions and some form of compulsory state insurance. The third variant, exemplified by the United Kingdom between 1911 and 1920, is a state-administered compulsory insurance system with the option for unions either to apply to administer the state program in specific sectors or to opt out and form their own funds with state subsidies . The typology presented in Table 1 provides the basis for a narrow and a broad definition of "Ghent." With Ghent being defined by strictly voluntary insurance as well as union administration, the narrow definition coincides with the second of the three types of unemployment insurance systems in Table 1. With the administrative role of unions alone as the defining characteristic of Ghent systems, the broad definition encompasses the second and third types. The regression results presented below are based on the broad definition of Ghent, but we obtain very similar results if we instead use the narrow definition. Table 2, in turn, provides descriptive data on the financing of unemployment insurance over the period 1930 to 1970. To reiterate, we posit that the critical variable here is the extent to which unemployment benefits are subsidized by mandatory employer contributions and/or by general tax revenues. For our purposes, the important point to take away from Table 2 is that Ghent systems of unemployment insurance were self-financed to a much greater extent than state-administered systems in the 1930s and 1940s. In some Ghent countries and most non-Ghent countries, the burden of financing unemployment insurance shifted toward employers and the state from 1930 to 1950. More strikingly, the Nordic Ghent countries massively reduced the reliance on contributions by the insured from 1950 to 1970. Setting the exceptional Swiss case aside, Ghent systems were, on average, more subsidized than state-administered systems in 1970. 11 --- Empirical Results: The Ghent Effect Turning to more systematic analyses of empirical data, let us begin by exploring the effects of introducing or dismantling Ghent systems of unemployment insurance on the rate of unionization. The consensus in the literature on the Ghent effect implies that the rate of unionization should increase in years following the introduction of Ghent and that it should decline in years following the replacement of Ghent by state-administered insurance. We test these expectations by estimating difference-in-difference models including leads and lags of the Ghent dummy. Specifically, we estimate ordinary least squares models with indicator variables up to 5 years or 10 years before and after changes in the value of the dummy variable. This setup allows us not only to assess the effects of changes in the institutional design of unemployment insurance, but also to explore whether such changes might themselves have been a response to unionization. According to Table 1, there is only one instance of change in the Ghent dummy since 1970 if "Ghent" is understood broadly, that is, defined so as to encompass the hybrid cases. As we are interested in the effects of changes in this dummy, the results presented in Table 3 are based on estimating our models with data for the period 1870 to 1970. For our measure of unionization, defined conventionally as the percentage of wage-and salary-earners who are union members, we rely on the data set constructed by Rasmussen from International Labour Organization reports and secondary sources. We opt for this data set because it covers more countries and provides longer time series than other multicountry compilations of unionization data. 12 Table 3 presents the results of estimating several difference-in-difference models. These models address potential autocorrelation and heteroscedasticity by clustering standard errors by country. The first two models follow the new convention of relying on fixed country and year effects to minimize omitted variable bias in estimating difference-in-difference models . Now, there are good reasons to be skeptical of the idea that changes in the independent variable of theoretical interest-in our case, the introduction or removal of Ghent-resemble a randomly assigned treatment and, in particular, the assumption of any two units following parallel trends prior to one unit being exposed to the treatment. As a first cut, we address these issues by including country-specific time trends in Models 1 and 2 . In Models 3 and 4, we add a series of control variables identified by comparative studies of unionization . As we are interested in estimating the effect of introducing or removing Ghent on unionization, the appropriate control variables to be included are variables that might affect the presence of Ghent as well as the rate of unionization and it is not altogether obvious that this holds for all of the six variables that we add. It should also be noted that data availability constrains our choice of control variables. Relative to the existing literature, the whole point of this exercise is to encompass the time period in which Ghent systems were introduced and, in some instances, removed. Model 3 includes three control variables. First, we control for the size of labor force to take into account Wallerstein's argument that the benefits that unions and unionized workers derive from organizing a certain number of workers decline with the size of the labor force while the costs are the same. Second, we control for trade openness, measured as exports and imports in percent of GDP, to capture how changes in the integration of labor markets into the world economy might influence the incentives or abilities of workers to organize as well as the incentives for politicians to encourage or The t statistics are in parentheses, calculated with panel-corrected standard errors clustered by country. The coefficient for "reform year" represents the correlation between Ghent and union density in the year when a Ghent system was introduced. "Ex ante" refers to the number of years prior to a reform while "ex post" refers to the number of years after the reform was introduced . *p < .05. **p < .01. ***p < .001. --- Table 3. discourage unionization . Finally, we include GDP per capita as a control variable, to capture the potential effects of economic growth on industrial organization and government policies that might have promoted unionization. 13 The fourth model introduces three additional control variables. From the literature on the postwar period, we know that public sector is associated with unionization . In the absence of government employment data prior the Second World War, we use government expenditure from Banks as a proxy for the size of the public sector. Expecting industrial employment to be associated with unionization , we also control for this variable with data from Banks . Finally, Model 4 controls for the presence of Left government, taking account of the fact that Left government has been more common in countries with Ghent systems and might shape unionization through mechanisms other than the selective incentives associated with Ghent . 14 Comparing the results for the first two models, we find that the estimated effects over the 5 years before and after a change in the Ghent dummy are essentially the same whether or not the model includes 10-year leads and lags. In addition, Table 3 shows that the effects of the introduction and removal of Ghent systems identified by estimating these stripped-down models are robust to the inclusion of any and all of the aforementioned control variables. Based on Model 3 in Table 3, Figure 2 graphically illustrates our results. 15 The coefficients for the 3-year lead and the 1-year lead as well as the year in which a reform was adopted are statistically significant. We also obtain a positive coefficient for the 10-year lead. On the contrary, none of the coefficients of the lags for the first 5 years are statistically significant and the coefficient for the 10-year lag is significant with a negative sign. In short, we find some evidence that unionization may have triggered the introduction of Ghent systems and we find no evidence whatsoever that the introduction of Ghent systems had a positive long-term effect on unionization. Over the long haul, introducing Ghent may actually have had a negative effect on union density. 16 As shown in the online appendix , we obtain very similar results for the entire period 1870 to 2010 and when we estimate these models with data restricted to countries that have had a Ghent system at some point in time. We also obtain similar results when we include all the country-years in the Social Policy Around the World data set for which we have observations of union density . Using the narrow definition of Ghent , we find only significant lead effects. Again, the most obvious reason why we do not observe a Ghent effect is that Ghent systems were not very generous in their infancy. High individual contributions and low benefits rendered the incentives to join union-administered unemployment funds weak and, by extension, the introduction of Ghent systems did not significantly alter the incentives to join unions. As early Ghent systems were unable to pool bad and good risks, they probably underperformed state-administered compulsory systems in protecting workers against unemployment in the interwar period. This might explain the negative 10-year Ghent effect that we observe: Quite plausibly, workers in Ghent countries experienced bigger income losses as a result of unemployment than workers in non-Ghent countries in the interwar period and were therefore more prone to drop out of unions when they became unemployed. As governments shifted the burden of financing Ghent systems onto employers and taxpayers from the 1940s onward, they arguably mitigated this problem, enabling unions to retain members even in periods of high unemployment. The graphs are based on Model 3 in Table 2. The coefficient for "reform year" represents the correlation between Ghent and union density in the year when a Ghent system was introduced. "Ex ante" refers to the number of years prior to a reform while "ex post" refers the number of years after the reform was introduced . CI = confidence interval. Table 4 reports the results of estimating a series of models in which subsidization of unemployment insurance features as a determinant of unionization over the period 1930 to 2005. The measure of subsidization used here is the inverse of individual contributions as a share of total unemployment insurance receipts . 17 Our hypothesis is that subsidization promotes unionization when unemployment insurance takes the Ghent form and not otherwise. Conversely, we expect Ghent to be positively associated with unionization only at high levels of state subsidization. Again, the results presented here are based on the broad definition of Ghent proposed above, but using the narrow definition we find similar results. As shown in Table 4, interacting Ghent and state subsidization in a standard OLS model confirms our hypothesis in a very convincing manner. With and without fixed effects and the control variables mentioned above, we find negative direct effects of subsidization as well as having a Ghent system, but the interaction of these two variables is positive, statistically significant, and substantially meaningful. Based on Model 5 in Table 4 , Figure 3 illustrates how the subsidization effect is conditioned by the institutional design of unemployment insurance. In a state-administered system, increases in subsidization are negatively correlated with density, but in a Ghent system we find a strong positive effect of subsidies. Our estimate is that going from zero to full subsidization increases 1930-2005 1930-2005 1930-2005 1930-2005 1930-2005 1930-1980 The density by 25 percentage points when unemployment insurance is organized on Ghent principles. The strong interaction effect on union density holds up when Ghent and subsidization are lagged by 1 year . Restricting the sample to countries with Ghent systems, we find that lags for state subsidies have a significant positive effect on union density, whereas leads do not, allaying concerns about reverse causality . We also report the results of two "placebo tests" in the online appendix . One might well suppose that various social provisions favoring workers go together and that union density is positively correlated with all forms of subsidization. To test for this possibility, we replace unemployment insurance subsidies with sickness or workplace accident insurance subsidies. Rerunning models presented in Table 3 with either sickness or workplace accident insurance subsidies as the independent variable of interest, we do not find a significant effect of either subsidy in Ghent countries. Going beyond the existing literature on the Ghent effect, the results presented above show that having a Ghent system of unemployment insurance does not by itself boost unionization, but having a highly subsidized Ghent system insurance is indeed associated with higher rates of unionization. Rothstein's account of the Swedish case suggests that the probability of observing Ghent adoptions increases when the Left is power. As noted above, Rothstein's account involves foresighted political entrepreneurship and does not imply that Left governments always adopt Ghent. But in Rothstein's theoretical model, Left parties are the only political actors with a long-term rational interest in the adoption of Ghent. By contrast, we argue that the interests of Left parties regarding Ghent were ambiguous and that Social-Liberals had electoral and ideological reasons to promote Ghent before the Second World War. We test these competing theories of the partisan-political origins of Ghent by estimating a series of logistic regression models with Ghent adoption as the dependent variable. --- Empirical Results: Partisan Effects Our measures of government partisanship are based on the ideology of heads of government as coded by Brambor, Lindvall, and Stjernquist . Focusing on the economic dimension of ideology, Brambor et al. code heads of government as Left, Center, Right, or Other. The "centrist" category encompasses "various centrist ideologies, especially social liberalism-a generally market-oriented ideology combined with a social reform agenda" . With some exceptions, Brambor et al. code Christian Democratic heads of government as "conservative." 18 Including dummies for Left and Centrist or, in our preferred terminology, Social-Liberal heads of government in the same model means that we identify the effects of these party families relative to Conservatives, assumed to be unequivocally negative to the introduction of any form of unemployment insurance. Before turning to the regression models, it is instructive to consider what the raw data tell us about the institutional design preferences of Left parties. In our data set, there are altogether 34 episodes of government headed by Left parties between 1900 and 1949. In 10 cases, Left parties came to power in the absence of any unemployment insurance system: The Left introduced stateadministered unemployment insurance in four of these cases and Ghent-style insurance in one case . In nine cases, the Left came to power with state-administered insurance in place and, in all these cases, the institutional design remained unchanged. Finally, there are 15 cases in which the Left came to power with Ghent insurance in place: The new government replaced Ghent by stateadministered insurance in three of these cases . Overall, it would appear that Left parties tend to prefer stateadministered unemployment insurance. Our main regression models include GDP per capita to control for differences in development and wealth across countries and over time. In addition, and perhaps more importantly, we again include international trade in percent of GDP, conceived as a proxy for the level and structure of labormarket risks. Huberman argues persuasively that international trade is decisive in shaping the size of the pro-welfare coalition, as the income volatility introduced by international trade renders workers and employers in trade-exposed sectors more supportive of risk redistribution through social insurance . Many scholars have argued that proportional representation promotes the inclusion of labor-market partners in policy decision making and implementation . In light of this literature, it seems plausible to suppose that countries with PR are more likely to adopt Ghent. Using Schjolset's classification and data, we control for such an effect by including a dummy that takes the value 1 if a country has a PR system . Our argument holds that Ghent served, at a certain point in time, as an instrument for progressive Liberals to mobilize electoral support among skilled workers and to bolster the organizational resources of craft unions. As noted earlier, an important feature of this argument is that the rise of industrial unions rendered the Ghent option less attractive to progressive Liberals. To explore this question, we include a time-variant dummy measure of union structure developed by Rasmussen . This variable takes the value of 1 when the main union confederations in a country aim to organize or have already organized workers along industry instead of craft lines. Table 5 presents the results that we obtain with our broad definition of Ghent and data for the period 1905 to 1970, with fixed country and year effects and standard errors clustered by country. Our basic model can only be estimated with cases that experienced some change and so the analysis is restricted to eight countries. 19 In the first model, we include only partisanship measures. In Model 2, we control for GDP per capita, trade openness, and PR. In Model 3, we also control for industrial unionism and in Model 4, finally, we interact industrial unionism with the variable that captures the presence of a Social-Liberal head of government. The results in Table 5 clearly support our argument about Ghent as a social-liberal initiative. In all the models, the effect of having a Left head of government is imprecisely estimated. By contrast, the positive social-liberal effect consistently clears standard thresholds of statistical significance. The size of predicted effect depends on the model specification, but taking Model 1 at face value, and setting Left government at its mean value, the formation of a Social-Liberal government increases the probability of adopting a Ghent system by 34 percentage points. Based on the second model, with two control variables, Figure 4 plots the differences in probabilities of having a Ghent system for each of the partisanship variables. As for the control variables, we do not find any effect of GDP per capita, PR, nor any effect of trade openness. We find that industrial unionism reduces the probability of the introduction of a Ghent system of unemployment insurance and, more importantly, that union structure, or perhaps union militancy, conditions the effect of Social-Liberal government on the probability of Ghent adoption. 20 Based on Model 4, Figure 5 shows the effect of Social-Liberal government on the probabilities of Ghent adoption in the absence and presence of industrial unionism. Simply put, we find that the social-liberal effect is only significant in the absence of industrial unionism. We present the results of a number of robustness results in the online appendix . To begin with, we exclude Norway and Sweden as these countries have above-average influence statistics and were used to generate our theory. We also estimate random effects models with and without fixed-year effects. The estimated effects of partisanship remain substantially the same with any and all of these modifications. We then explore the temporality of partisan effects by lagging the Left and Social-Liberal dummies up to 5 years in separate models. While we do find some indication of an immediate effect of Left government, the lags for Left government are all insignificant. By contrast, Social-Liberal government lagged by 3 to 5 years is positively associated with Ghent adoption. Finally, we do not find any direct effects of either Left government or Social-Liberal government when we rerun the models in Table 4 with a Ghent dummy based on the narrow Ghent definition . However, the positive effect of Social-Liberal government in the absence of industrial unionism remains statistically and substantively significant. This suggests that their ideological disposition in favor of voluntarism does not suffice to explain why Liberals played such an important role in the establishment of Ghent systems. Their efforts to mobilize skilled workers in the electoral arena and to strengthen craft unions in their competition with industrial unionism would appear to be a critical part the Ghent story. To summarize our analysis up to this point, we have shown that Left government does not explain the adoption of a Ghent design of unemployment insurance and that the Ghent design by itself does not boost union growth. We have also shown that subsidization of Ghent systems is associated with higher rates of unionization over the period 1930 to 2005. The obvious question becomes: Have Left parties promoted union growth by increasing subsidization of Ghent systems since the 1930s? As suggested above, union growth could be an unforeseen by-product of Ghent systems favored by Liberals being combined with welfare-state generosity favored by Social Democrats. However, it is also possible that Social Democrats who came to power in countries with Ghent systems favored subsidization for strategic reasons, realizing that subsidization would promote unionization and working-class electoral support for the Left. By the logic of the former argument, we would expect Left government to be associated with higher levels of subsidization regardless of the institutional design of unemployment insurance. By contrast, the argument about strategic behavior implies that the effect of Left government should be noticeably stronger in the presence of Ghent. In Table 6, we present the results of estimate models of subsidization that interact Left government with the Ghent dummy. We first estimate these models with data for the entire period 1930 to 2005, and then restrict the analysis to the period 1950 to 2005. The justification for this two-step procedure is that we know that the Ghent design was a subject of contestation on the Left in the 1930s and 1940s. By restricting the analysis to the period after 1950, we drop cases in which Left parties were seeking to replace Ghent by state-administered insurance. In other words, the interaction effect in the models with data for 1950 to 2005 should capture the effect of government by Left parties committed to the preservation of Ghent systems. The results in Table 6 support the strategic behavior hypothesis. For the period 1950 to 2005 as well as the entire period 1930 to 2005, the coefficient for Left government has a negative sign, but this effect is imprecisely estimated. For the entire period 1930 to 2005, we obtain a positive effect of interacting Left government with the Ghent dummy, suggesting that Left governments were more prone to increase subsidization in Ghent countries. However, the introduction of fixed country and year effects renders the coefficient for the interaction term statistically insignificant by any conventional standards. When we restrict the analysis to the period 1950 to 2000, the coefficient for the interaction term doubles in size and remains significant at the 95% level when we add not only fixed country and year effects, but also GDP per capita, trade openness, and PR as control variables. Based on Model 6 in Table 6, Figure 6 plots the effects of Left government on subsidization conditional on the institutional design of unemployment. Taking the small sample into account, we report 90% rather than 95% confidence intervals in this figure. By this standard, we observe a significant negative effect of Left government under state-administered unemployment insurance and a significant positive effect of Left government under union-administered unemployment insurance. 21 A Left government in a Ghent country is predicted to increase subsidization by 14 percentage points compared with center and right governments. The contrast between Ghent and non-Ghent countries is quite striking, suggesting that the kind of strategic thinking that Rothstein invokes to explain the embrace of Ghent by the Swedish Social Democrats in the 1930s might have informed decisions about the financing of Ghent systems in Sweden and other countries in the postwar era. --- Conclusion Much of the quantitative literature on the Ghent effect reads history backwards and, as a result, suffers from selection bias. Relying primarily on data since 1960, the literature finds that countries with Ghent systems of unemployment insurance have, on average, significantly higher union density than countries with state-administered unemployment insurance and infers that this effect derives from intrinsic features of Ghent systems. Against this background, it becomes reasonable to suppose that Left parties have been protagonists in the introduction of the Ghent design of unemployment insurance. 22 With the notable exception of Switzerland, the Ghent systems that survived into the 1960s and beyond were heavily subsidized by general taxes or payroll taxes imposed on employers by the state and provided relatively generous benefits to the unemployed. Prior to 1950, however, the extent of subsidization of Ghent systems was invariably quite limited. Subsidies allowed craft unions to provide reasonably effective unemployment insurance to skilled workers, but did little for the unskilled and semiskilled workers who were the main targets of the organizing efforts of industrial unions and the main electoral base of Left parties. The Ghent design of unemployment insurance suited progressive Liberal parties far better than it suited Labor and Social Democratic parties in the first half of the 20th century. Sweden is the only case in which Social Democrats can be said to have been protagonists in the introduction of a Ghent system. As documented by Rothstein ), at least some Swedish Social Democrats, most notably Möller, argued for the Ghent design on the grounds that it would strengthen the labor movement over the long run. It is important to recognize, however, that the Liberals were an important player, supporting the introduction of Ghent, in the Swedish case as well. As the support of at least one of the "bourgeois" parties was needed to introduce nationwide unemployment insurance, Ghent may have been the only option available to the Swedish Social Democrats in the early 1930s. More importantly, high unionization mitigated the trade-off between short-term electoral considerations and long-term strategic considerations for the Swedish Social Democrats. The idea that Ghent might serve as a means to provide decent unemployment insurance to a majority of the working class was more plausible in Sweden than in most other countries. In closing, let us briefly articulate some implications of our analysis for recent theorizing about institutional change in the historical-institutionalist tradition. In her book on the historical evolution of vocational training in Germany, Britain, the United States, and Japan and a series of more theoretical papers, Kathleen Thelen argues convincingly that the historical-institutionalist tradition has relied excessively on "punctuated equilibria" in thinking about, and seeking to explain, institutional change ). According to Thelen, large-scale institutional change triggered by crises is a rare phenomenon and we ought to pay more attention to incremental changes in the way in which institutions operate. Thelen aptly summarizes her story of vocational training in Germany as a story of "institutional conversion." In her words , "institutional created for one set up purposes and resting on the shoulders of one set actors" were "carried forward on the shoulders of another coalition altogether" . The evolution of Ghent systems of unemployment insurance in Belgium, Denmark, and Finland might similarly be characterized as cases of "institutional conversion." Again, these systems were created by Liberals in an effort to shore up the position of craft unions and curtail the radicalization of newly enfranchised workers. Through subsidization, they gradually became more union-friendly, benefiting industrial as well as craft unions and, arguably, expanding the electoral base of Left parties. In our view, however, it is important to stress that this conversion did not occur through the internal workings of the Ghent system. Rather, institutional conversion was a result of legislative acts that changed the way in which the Ghent systems in question were financed and the level of benefits that they provided. The story here seems to be neither a story of punctuated equilibria nor a story of endogenous institutional change. In particular, we have found no evidence to suggest that the existence of Ghent systems somehow brought about the political conditions that enabled subsidization from the late 1940s onward. Put differently, it seems misleading to construe the Ghent effect on unionization as an unintended consequence of unemployment systems introduced by Liberals in the first two decades of the 20th century. Equally important, our quantitative analyses encompass a set of cases in which institutional conversion did not occur. In these cases, we instead observe what Thelen refers to as "institutional replacement," that is, the replacement of Ghent systems by state-administered systems. This kind of institutional change occurred in Britain , Spain , Norway , the Netherlands , France , and, most recently, Switzerland . Setting the Swedish aside, Left parties were not protagonists in the introduction of Ghent systems, but they appear to have played a critical role in the subsequent evolution of many Ghent systems. The role of Left parties is difficult to capture through quantitative analysis for it varies across countries. On one hand, some Left parties-the Spanish Socialists, the Norwegian and Dutch Labor parties, and the Swiss Socialist Party-initiated legislation that shut down poorly funded Ghent systems, replacing them with mandatory, state-administered unemployment insurance. On the other hand, other Left parties-the Belgian Socialists as well as the Swedish, Danish, and Finnish Social Democrats-initiated legislation that increased subsidies to union unemployment funds in the 1950s and 1960s. 23 How might we explain the divergence between Left parties that opted to abandon Ghent and Left parties that opted to strengthen Ghent ? The historical record suggests that timing matters: Setting Switzerland aside, Left parties have consistently chosen the second option in cases where Ghent systems survived into the 1950s. In so doing, they avoided conflicts with unions with a vested interest in the maintenance of their unemployment funds. More importantly, the preceding discussion and analyses suggest that the Ghent design became more attractive to Left parties, relative to state-administered unemployment insurance, as unionization progressed and unemployment declined in the postwar era. That said, it must also be noted that there is not a single instance of a Left party advocating, let alone implementing, the replacement of state-administered unemployment insurance by a Ghent system. The legacy of social liberalism remains critical for understanding cross-national differences in the institutional design of unemployment insurance. 9. According to Rothstein , when the insurance scheme was established in the mid-1930s, Sweden had a fairly but not uniquely high union density. It is thus not a uniquely strong working class that creates a Ghent system. Instead, the causal link goes the other way round. A medium-strong working class establishes such a system to enhance its future strength. According to our data , Swedish unions organized roughly 40% of the workforce by the early 1930s, twice the average for industrialized countries. Only two countries-Czechoslovakia and fascist Italy-had higher unionization rates in the early 1930s. 10. Our coding of unemployment insurance systems is based on the Social Policy Around the World database, created by Rasmussen . The SPAW database contains information for a total of 55 countries over the period 1870 to 2010, but the time series are considerable shorter for some countries. The 21 countries included in Table 1 are the 21 countries in the SPAW database that are currently Organisation for Economic Co-operation and Development member states. This subset includes all SPAW country-years with a Ghent system of unemployment insurance. Note that Spain drops out of our data set for lack of reliable data on union density and that Czechoslovakia drops out of our data set when it ceases to be an independent country in 1939. Note also that our analyses include separate country dummies for imperial Germany, West Germany, and reunified Germany. 11. Though our coding of legislative texts identifies Switzerland as a hybrid case, including a compulsory insurance element, the Swiss system functioned as a strictly voluntary system in the postwar era . According to the Social Citizenship Indicator Program, only 2% of Swiss employees were insured against unemployed in 1970. Not surprisingly, the Swiss Socialist Party spearheaded the introduction of state-administered unemployment insurance in 1976. 12. For overlapping years, the correlation between Rasmussen's data and those of Scheve and Stasavage is .98. For the post-1960 period, Rasmussen relies on Visser's data for all countries except the United Kingdom. 13. Our measure of the size of the labor force is based on census data reported in Flora, Alber, and others and national statistical yearbooks, with missing observations between data points being interpolated. Our trade data are from Barbieri, Keshk, and Pollins and our GDP data are from Bolt and van Zanden . 14. Our measure of government partisanship is a dummy variable that takes the value 1 when the head of the government executive is from a Labor, Social Democratic, or Socialist party . Table A1 in the online appendix provides descriptive statistics for the main empirical models reported in this article. 15. We base this illustration on Model 3 because we believe that the control variables included in this model are more readily justified from a theoretical point of view. While Figure 2 construes our results as the effects of introducing Ghent unemployment insurance, it is important to keep in mind that our analysis pertains to the effects of abandoning Ghent in favor of state-administered unemployment insurance as well. A graph showing the effects of abandoning Ghent would be the mirror image of Figure 2. 16. Viebrock's qualitative-comparative analysis yields similar results regarding the historical association between Ghent and union density . 17. Using short-term replacement rates from the Social Citizenship Indicator Program database as a measure of unemployment generosity produces results that are very similar to the ones reported below . 18. As exceptions to this rule, the codebook mentions the German Zentrum party and smaller Christian Democratic parties in countries with large conservative parties. 19. The countries included are United Kingdom, the Netherlands, Belgium, France, Switzerland, Sweden, Norway, and Denmark. Finland drops out of this analysis as Ghent was introduced in 1917 and never removed. As we only have measures of government partisanship for Finland from 1917 onward, Finland ends up being perfectly predicted by the country dummy. 20. We do not find any significant effects of interacting Left government with either union density or union structure. 21. Replicating our analysis with interpolated annual measures of subsidization renders both effects significant at the 95% level. See Table A12 and Figure A1 in the online appendix. 22. Korpi usefully distinguishes between "protagonists," "antagonists," and "consenters" in the expansion of the welfare state. 23. The French case is difficult to summarize in this fashion on account of divisions within the labor movement and the Left. On transitions from Ghent to stateadministered unemployment insurance, see also Viebrock . --- Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. --- Notes 1. In Belgium, the national government began to subsidize union-administered funds in 1907 to 1908. See Kiehel for a detailed account of the Belgian case; Alber and Viebrock for comparative discussions of the origins of different unemployment insurance designs. 2. Clasen and Viebrock provide a detailed and very informative account of how the administration of unemployment insurance incentivizes workers to join unions in Denmark and Sweden. 3. For a sampling of this literature, see Ebbinghaus and Visser , Høgedahl , Schnabel , Scruggs , Van Rie, Marx, andHoremans , andWestern . 4. Having nearly closed in the course of the 1940s and 1950s, the gap again opened up from 1960 onward. Arguably, it was not until the 1960s that state subsidies and concerns about unemployment reached levels that were sufficiently high to generate a Ghent effect in Sweden. 5. Space does not allow us to address recent reforms of Ghent systems of unemployment insurance in this article. Suffice it to note that the radical reduction in state subsidies enacted by Sweden's new "bourgeois" government in 2006 appears to have had a big impact, accelerating the decline of union density . Recent Swedish experience is consistent with our argument that the Ghent effect is susceptible to political manipulation. 6. The following summary draws primarily on Edebalk's detailed account of political struggles surrounding unemployment insurance in Sweden between 1892 and 1934. In English, see Heclo as well as Rothstein ). 7. Here we draw primarily from the government's evaluation of the scheme and debates at the Landsorganisationen congress of 1938 . 8. As shown in Figure 1, Norwegian union density barely exceeded 15% in the early 1930s. In addition, low levels of state subsidies restricted union-administered unemployment funds to low risk sectors and very large unions, such as the Metal Workers Union. According to Bjørnson and Haavet , only 5.5% of 316,000 LO members were insured against unemployment in 1937. --- Author Biographies Magnus Bergli Rasmussen is senior researcher at the Institute for Social Research, Oslo. He is currently involved in research projects on the historical origins of distributive norms, and the effect of electoral rules on industrial conflict and collective bargaining using new historical Norwegian geo-coded data. Jonas Pontusson is professor of comparative politics at the University of Geneva. His current research projects fall into two domains: growth models and the politics of macroeconomic management in Western Europe; and the impact of rising inequality on the policy preferences of different citizens and their political influence in liberal democracies.

Many studies have found that countries with union-administered unemployment insurance have higher rates of unionization than countries with state-administered unemployment insurance. With data going further back in history, this article demonstrates that the introduction of so-called "Ghent systems" had no effect on unionization rates. We argue that the Ghent effect identified by the existing literature came about as a result of increasing state subsidization and benefit generosity in the 1950s and 1960s. Exploring the partisan politics of unemployment insurance, we show that progressive Liberals ("Social Liberals") favored Ghent designs while Social Democrats favored state-administered unemployment insurance before the Second World War. We also present some evidence suggesting that Left governments, inheriting Ghent systems that were not of their choosing, promoted state subsidization in the postwar era and thus helped generate the Ghent effect identified by the existing literature.

systematically identified or supported in healthcare settings. Consumer health information technology affords a largely untapped opportunity to formalize dementia care partner identification and engagement in healthcare delivery. One example is OurNotes, a portal-based agenda-setting questionnaire that allows patients to submit concerns for discussion ahead of medical visits. In a proof-of-concept study, OurNotes was first adapted to elicit the identity of the person completing the form and clarify their relationship to the patient, and subsequently piloted at an academic geriatric practice. Semi-structured interviews were conducted with 15 care partners to elicit their perceptions of the adapted OurNotes and suggestions for refinement. Transcripts were analyzed using deductive, thematic analysis. Care partners' perceptions of the tool were overwhelmingly positive. Major benefits included enhancing their feelings of preparedness for the visit, having a voice to express concerns, and streamlining visit communication. The most commonly endorsed challenge was the time required to complete the adapted OurNotes in addition to routine check-in items. A primary suggestion was to enable the submitted OurNotes responses to be edited/updated. Findings suggest that the adapted OurNotes is an acceptable approach to identifying care partners in the dementia context. Refinements may further facilitate visit efficacy through preparedness. Future research will examine CHIT engagement on patient and care partner outcomes. Abstract citation ID: igad104.3450 --- LGBTQ+ INCLUSIVITY TRAINING AND EDUCATION FOR SKILLED NURSING FACILITIES Jennifer May 1 , Alexis Domeracki 2 , Perisa Ashar 3 , Foxx Hart 3 , Jason Wheeler 4 , Glaucia Salgado 5 , and Melanie Wang 3 , 1. University of South Carolina,Columbia,South Carolina,United States,2. Duke School of Medicine,Durham,North Carolina,United States,3. Duke University,Durham,North Carolina,United States,4. Duke PHMO,Durham,North Carolina,United States,5. Duke Clergy Heath,Durham,North Carolina,United States The population of older adults is growing in the United States, leading to greater need for specialized care offered by skilled nursing and long-term care facilities. Among older adults that require specialized health care, lesbian, gay, bisexual, transgender, queer, and other sexual and gender minority older adults are more likely to experience health disparities and health care barriers due to lifelong discrimination and exclusion. Healthcare workers within skilled nursing facilities have expressed they do not feel prepared to care for LGBTQ+ older adults. The aim of this research was to develop the LGBTQ+ Inclusivity Training and Education toolkit to provide information to healthcare workers, staff, residents, and their families in the SNF environment. Phase 1 used a community engaged approach to develop the LITE toolkit through engagement with an LGBTQ+ Community Advisory Board. Phase 2 was the implementation and evaluation of the LITE toolkit with 25 SNFs throughout North Carolina. The majority of SNF healthcare workers and staff who responded to the survey indicated that the LITE toolkit was "easy to use," they were satisfied with the contents, and that it would improve the way they care for patients. Providing LGBTQ+ focused education for members of the SNF community addresses the need for healthcare worker and staff training, thereby contributing to equitable care and inclusive environments for the LGBTQ+ older adult community. Additional work focused on understanding the facilitators and barriers to using the LITE toolkit in the SNF setting is needed. Abstract citation ID: igad104.3451 --- LIFE COURSE TIMING OF ADVERSITY AND DIURNAL CORTISOL IN LATER LIFE Nadine Sikora 1 , Briana Mezuk 2 , James Abelson 2 , Jane Rafferty 2 , Jamie Abelson 2 , and Julie Ober Allen 1 , 1. The University of Oklahoma, Norman, Oklahoma, United States, 2. University of Michigan, Ann Arbor, Michigan, United States Exposure to adversity is believed to alter stress sensitive biological processes. One manifestation of this, flatter diurnal cortisol slopes, has been linked to cardiometabolic problems. Questions remain about whether the timing of adversity within the life course influences its effect on cortisol slopes in later life. Two longitudinal cohort studies with identical measures were merged for analysis using multilevel modeling. Measures included retrospective self-report of childhood trauma , major life stressors in adulthood , and peakto-bedtime cortisol slopes . Report of any moderate and severe childhood adversity was not associated with adult cortisol slopes . When we probed specific types, emotional abuse was associated with cortisol slopes . For adult adversity, total number of stressful events and recent stressful events were associated with cortisol slopes. When simultaneously modeling all measures of adversity over the life course, number of stressful events in adulthood and childhood emotional abuse were the strongest predictors of flatter diurnal cortisol slopes in adulthood. Given that those reporting childhood adversity were also more likely to have experienced adulthood adversity , early life adversity may be an important risk factor for flattened cortisol slopes and associated cardiometabolic outcomes in later life via multiple pathways. Implications for intervention will be discussed. Abstract citation ID: igad104.3452 communication, and documentation of life-sustaining care and quality of life preferences after personal reflection, but ACP rates remain lower among older adults with limited income. To better understand how life-course factors influence these low ACP rates, we aimed to explore participant perceptions of healthcare and ACP barriers and facilitators through different life stages. Methods: We sampled older adults with limited income from six community-based sites . Qualitative descriptive design was utilized to perform semi-structured, in-person interviews. Results: Themes emerged from the inductive inclusion of in-vivo codes and the deductive application of the Cumulative Disadvantage Theory. Mean age: 64.8 years old ; 11 participants identified as female ; 16 identified as Black/African American . Four themes emerged: 1) structural, lifestage, 2) social stressors and resources, 3) individual stress responses and 4) ACP readiness. Participants' perceptions denoted that proactive planning inequities among older adults with limited income result from the dynamic interplay of multi-dimensional stressors, protective factors, and personal stress responses. Implications: Identified themes serve as a conceptual basis for future intervention development that is responsive to traumatic life-course experiences and supportive of building ACP resilience , as well as policy and practice implications. Abstract citation ID: igad104.3453 --- LONELINESS AMONG UNDER-REPRESENTED OLDER ADULTS IN THE UK: A STUDY OF MINORITY ETHNIC AND LGBTQ+ POPULATIONS Christina Victor, and Isla Rippon, Brunel University London, Uxbridge, England, United Kingdom Internationally loneliness has been identified as a major public health problem. Although there is a substantial body of research about loneliness in older adults in the UK, there is a significant evidence gap reporting experiences of loneliness of older people from ethnic minorities and those who identify as lesbian, gay, or bisexual and transgender . These two groups, under-represented in UK gerontological research, are included in our recently funded project, Socially Inclusive Ageing across the Lifecourse. In this poster we explicitly focus upon the experiences of loneliness for older adults, aged 50+, from the LGBTQ+ and minority ethnic communities. Using wave 9 data from the UK Household Longitudinal Study we measured loneliness using the three-item UCLA scale with a score of 6+ out of 9 defining loneliness. Of our total sample of 16,805 who completed the loneliness measure, 1.5% of respondents identified as LGB and 5.4% as Asian, 2.9% as black and 1.5% as other or mixed ethnicity. Overall, 21.7% of the population aged 50+ were lonely. Participants from a black, Asian or other ethnic minority reported higher loneliness than white respondents: 25.8%, 29.6%, 31.0% and 21.0% respectively. Respondents identifying as gay or lesbian or as bisexual reported greater loneliness in comparison to those who identified as heterosexual . Our study is novel and timely in demonstrating the higher prevalence of loneliness in two under-represented groups of older adults with the potential consequences this may have for their health and wellbeing in later life. When caregivers struggle to manage their own health, they may lose confidence in their health care abilities, i.e., caregiver self-efficacy. Using loneliness as the primary factor, the current study tested the mediating effect of caregiver self-efficacy on distinct outcomes of anxiety and depression. In a hospital setting in southern United States, the team in medicine, nursing, and social work collected data from 95 PWD caregivers on measures of loneliness, anxiety, depression, and caregiver selfefficacy. SEM with bootstrapping tested the two hypothesized mediation models, controlling for age, gender, and relationship to care recipient. The typical respondent was a 62-yearold Caucasian female providing care to a parent. Mediation model 1 direct effects were significant for loneliness on selfefficacy ; for self-efficacy on depression ; and for loneliness on depression . Indirect mediation effect of loneliness on depression through self-efficacy was significant . Model 2 direct effects were significant for loneliness on self-efficacy , and for self-efficacy on anxiety ; direct effect of loneliness on anxiety was not significant . Indirect mediation effect of loneliness on anxiety was significant . Results underscore the importance of emphasizing social engagement, and including self-efficacy as a mediating factor, when treating caregivers' anxiety and depression. Utilizing the Health and Retirement Study data, this research exclusively centered on a sample of 2953 African-American participants to discern longevity predictors, emphasizing the importance of intra-group analysis. This targeted approach is essential as it provides a nuanced understanding of factors affecting longevity within a specific demographic group, avoiding broad generalizations derived from diverse populations. Two analytical models were developed in Mplus. The initial fully saturated model was subsequently trimmed to exclude non-significant paths. Key findings include: Education showed a positive correlation with cognition and a negative association with subjective health. Age had a negative impact on cognition and on subjective health. Depression was negatively associated with cognition and subjective health, and positively with age. Notably, age at death was majorly

Background: Cumulative disadvantage across the lifespan can lead to income, health, and advance care planning (ACP) disparities, resulting in increased caregiver decisional distress and reduced healthcare equity during times of cognitive incapacity. ACP is the process of patient and familial learning,

Introduction Over the course of 2020, governments have adopted a range of strategies to reduce the spread of COVID-19-an infectious disease that resulted in a pandemic-while trying to keep their economies afloat. Since mobility restrictions were gradually lifted during the last phase of the first pandemic wave in Europe, contact tracing has been considered to be an effective method for disease control, particularly for preventing disease transmission via contagious individuals who are not symptomatic [1][2][3]. Several governments have rolled out a version of a COVID-19 contact tracing app to help identify individuals who have been in close physical contact with an infected individual. However, as contact tracing apps inevitably rely on the collection of personal health data and mobility data, privacy concerns have been raised among the public [4]. In Europe, where participation in contact tracing via smartphone apps is voluntary, the effectiveness of contact tracing is dependent on the uptake of such apps. One study showed that in order to successfully suppress virus transmission during the peak of an outbreak in a hypothetical city with 1 million inhabitants, about 80% of all smartphone users or 56% of the population aged under 70 years would have to install the contact tracer [5]. Further, by modeling data from Washington State, researchers found that over the course of 300 days in 2020, infections and deaths could be reduced by 8% and 6%, respectively, if only 15% of the population were to participate in digital contact tracing [6]. Other researchers have also found that app-based tracing remains a more effective system than conventional contact tracing if coverage exceeds 20% [7]. Thus, although COVID-19 tracing apps are relatively ineffective during pandemic spikes, they can still help to slow the spread of SARS-COV-2 in subsequent periods, even though these apps have relatively low coverage. This study concentrated on the association between individuals' general health status and their willingness to use a COVID-19 tracing app across several European countries and focused on Spain, Italy, Germany, and the Netherlands. These attitudes were measured in the fall of 2020, which was when the daily number of new cases was increasing . We ask the following question: are poorer health statuses associated with a higher willingness to share personal information in COVID-19 tracing apps? This dynamic could occur if individuals prioritize personal or public health concerns over possible data privacy concerns. Such health risk calculations tend to operate differently for individuals who perceive themselves to be more vulnerable. The higher sense of danger among at-risk groups is often found to positively correlate with distress [8]. Furthermore, research has shown that individuals' level of engagement with disease prevention behavior increases as soon as they are able to translate an abstract societal risk into a likelihood of experiencing a disease's most severe consequences [9,10]. We also concentrated on the moderating role of COVID-19 proximity in the relationship between general health status and the willingness to use a contact tracing app. This is because risk perceptions are known to be partially influenced by the experiences of other individuals in one's social circle, such as family, friends, and colleagues [10,11]. Recent studies have also suggested that individuals' risk behaviors are rather susceptible to information treatments about COVID-19 during the pandemic. For instance, learning about the severe symptoms of COVID-19 positively influences a range of protective behaviors [12] and results in individuals being less accepting of the incautious behavior of others [13]. In other words, first-hand physical and psychological experiences and observations of nearby people being affected by COVID-19 are likely to impact individuals' attitudes and risk behaviors. Hence, we examined the relationship between a set of COVID-19 proximity indicators and individuals' willingness to install and use a contact tracing app. We used indicators such as being tested for COVID-19, having a close family member or colleague with COVID-19, and self-reporting depression and anxiety symptoms resulting from the pandemic. We also addressed the role that individuals' socioeconomic characteristics have as covariates of COVID-19 tracing app support. Such characteristics included gender, migration status, age, household status, and labor market status. It is important to account for these factors because of their expected relationship with the dependent variable . As contact tracing systems are being rapidly rolled out by current administrations, skepticism toward COVID-19 tracing apps may be rooted in general distrust toward the government, which is why the selected sociodemographics served as necessary control variables [14]. Sociodemographic factors are also predictive of general smartphone app usage and COVID-19 tracing app installation, as shown in recent studies [15]. Furthermore, it is important to account for possible general health effect heterogeneity across the aforementioned individual-level socioeconomic characteristics [16], which could also vary across European countries [15]. In sum, we expected to find significant self-reported general health status gradients in individuals' willingness to use a COVID-19 tracing app that are dependent on a range of socioeconomic characteristics. This relationship could be mediated by country-specific associations between socioeconomic attributes and COVID-19 tracing app support. We also hypothesized that as observable pandemic-related health risks increase for individuals, their willingness to use a COVID-19 tracing app also increases. --- Methods --- Data Observational data were drawn from the WageIndicator Living and Working in Coronavirus Times survey, which was filled out by web respondents between week 42 and week --- XSL • FO RenderX 49 of 2020 [17]. Respondents provided consent for their data be used in scientific research and did not receive financial compensation for their participation in the survey. All individual-level data were anonymized by WageIndicator prior to their use by academic researchers. The data set used and the analyses conducted did not contain identifiable information. We selected respondents aged between 18 and 70 years from Spain , Italy , Germany , and the Netherlands . This was because adults aged up to 70 years have relatively large social networks and stronger connections to the labor market . Contact tracing is also believed to be the most effective when it is performed with this population [5]. The LWCV survey collects data about family structure, COVID-19 testing, self-perceived health status, and depressive and anxiety symptoms. It also contains a series of questions about individuals' willingness to use a COVID-19 tracing app as well as data on relevant confounders, such as general smartphone and app usage. Multimedia Appendix 1 contains the sections of the questionnaire that were used for this study. Voluntary web surveys have become common data collection tools during the pandemic. A range of policy-relevant studies that documented the initial impact that COVID-19 has on health, work, personal, and family situations relied on data from voluntary web surveys [18][19][20][21]. Two important advantages of this data collection method are that sampling is continuous and that questionnaires can be adjusted to rapidly changing situations, such as the 2020 COVID-19 pandemic. A significant drawback of voluntary web surveys is that the samples are not representative of the full population . The results of such surveys therefore have to be interpreted with caution. The application of poststratification techniques can help to partly correct the bias resulting from self-selection and underrepresentation [22]. The WageIndicator Foundation is a global research organization that relies on a long-standing survey of workforces across 150 countries. The WageIndicator Foundation website receives millions of visitors annually. The WageIndicator Foundation has produced reliable estimates of mental health, data on subjective feelings such as well-being and insecurity, and web survey weighting techniques for balancing selectivity bias [23][24][25]. During the COVID-19 pandemic, it has enabled the exploration of mental health, anxiety, and life satisfaction determinants [26,27]. In Multimedia Appendix 2, we benchmark the LWCV study samples against those of the European Social Survey based on key sociodemographics; relatively comparable sample distributions across age and respondents' highest education level are displayed. However, the LWCV study samples contained slightly more individuals from the 30-to 54-year age groups than those in the general population, with the exception of Spain's population . Multimedia Appendix 3 indicates that including European Social Survey-based weights led to same substantive conclusions. In accordance with recent studies that used LWCV survey data [26,27], we report unweighted estimates for the main findings. Multimedia Appendix 4 documents the model statistics and model specification checks. --- Measures Data on COVID-19 tracing app support were derived by asking whether a respondent was willing to share both their health status and geographical location on a COVID-19 tracing app . The key independent variable was self-reported general health status, which was based on the following question: "How would you rate your overall health?" Respondents answered with "very good" , "good" , "fair" , "bad" , and "very bad" . We merged the smaller categories-the "bad" and "very bad" categories-with the "fair" health status to aid with interpretation and used the "very good" category as the reference. Indicators of COVID-19 proximity were measured with questions on whether a close colleague or a family member has ever contracted COVID-19, one's own COVID-19 test-taking status and their results , and self-reported COVID-19 pandemic-induced depression symptoms and COVID-19 pandemic-induced anxiety symptoms . Socioeconomic variables included gender, age , migration background , partnership status , whether children aged under 18 years were present in the household, the highest education level obtained , urbanicity , labor market position , and how labor market position has been affected by the COVID-19 pandemic in terms of job loss and income reduction. All models accounted for the timing of the survey . Multimedia Appendix 5 depicts the correlation matrix. --- Estimation In order to gain a thorough understanding of the critical structural pathways for contact tracing app uptake and its potential country-based variation, we first estimated the marginal effects of the socioeconomic factors and COVID-19 proximity indicators by using 2 separate series of logit models. We present the results of the bivariate models and multivariate models . These analyses also included an overall model with country-fixed effects, which allowed us to account for dynamics that are altogether country specific . Aside from learning about the relevance of these covariates for social and health policies, they also informed us about how the relationship between general health status and contact tracing app usage should be modeled. Two-sided significance tests were performed for all analyses. We also estimated COVID-19 tracing app support based on the general health status indicator in nested models; socioeconomic variables and COVID-19 proximity variables were added in separate steps . The use of nested models allowed us to examine the mechanism for explaining how general health status is related to the level of contact tracing app support. The baseline model only contained country-fixed effects and a control for survey week . In a second series of models, we added the socioeconomic matrix and a variable matrix for respondents' regular smartphone usage , that is, the ownership of other apps that collect health and geographic location data and the total number of phone apps. The third series of models were further adjusted for the COVID-19 proximity indicators . We calculated average marginal effects to aid our interpretation of the coefficients, as per social science conventions [28]. In Multimedia Appendix 6, we replicate the key results by using country random intercepts, which present the same quantitative results as those of the reported country-fixed effects models. Equation 1 is as follows: Y prob = β 0 + H i β 1 + F i γ + D i ω + C i ϕ + P i δ + ε i --- Results --- Descriptive Statistics --- Socioeconomic Factors Table 2 presents the marginal effects that socioeconomic factors had on the willingness to use a COVID-19 tracing app among the full sample and the four countries separately. The bivariate associations in the Country-fixed effects column suggest that older individuals are significantly more willing to use a COVID-19 tracing app than young adults . Partnered individuals who also live in the same household are also more likely to use a contact tracing app than nonpartnered individuals, as indicated by the 4% marginal effects gap. Individuals with medium and high levels of education had a significantly higher willingness to use a COVID-19 tracing app compared to that of individuals with low levels of education . Furthermore, compared to employees , individuals who are not active in the labor force and those who lost their job or income during the COVID-19 pandemic are significantly less likely to use a contact tracing app. These independent socioeconomic variables remained statistically significant in the multivariate model. Notably, gender, migration background, and urbanicity are not associated with the probability of using a COVID-19 tracing app. The columns of Table 2 present the bivariate and multivariate model results for each of the four countries. The main finding from these models is the striking cross-national variation in the relationship between socioeconomic variables and COVID-19 tracing app usage. The impact that respondents from Spain had on the full-sample results appears to be substantial, as the country largely exhibits the same socioeconomic relationships in terms of the significance levels and magnitudes reported by the multivariate models. However, in Italy, the most important socioeconomic factors are labor market status and urbanicity. Relative to employees , respondents who experienced a recent job loss and inactive respondents had significantly lower probabilities of being willing to use a COVID-19 tracing app. These substantial differences were derived from multivariate models that accounted for important confounders, such as age and household status. Witteveen & de Pedraza JMIR PUBLIC HEALTH AND SURVEILLANCE --- XSL • FO --- RenderX For Germany and the Netherlands, where COVID-19 tracing app support is, on average, much lower than that in the two Southern European countries, educational level and labor market position were the only statistically significant independent variables of the willingness to use a COVID-19 tracing app in the multivariate models. Holding higher education credentials in Germany and the Netherlands yielded a 6.6% higher marginal effect and an 11.4% higher marginal effect , respectively, on COVID-19 tracing app support compared to those for holding lower education credentials . In Germany, individuals who recently experienced a job loss or income loss are significantly less likely to use a contact tracing app than employees . In the Netherlands, the marginal effect for support for a COVID-19 tracing app is only 7.8% for those who remained inactive during the pandemic, regardless of other socioeconomic factors. Finally, in Germany, where age remained a nonsignificant independent variable, the presence of a partner and children in the household are positively associated with respondents' willingness to use a contact tracing app. --- COVID-19 Proximity We also examined key indicators of COVID-19 proximity. Table 3 presents the marginal effects that these indicators had on the willingness to use a COVID-19 tracing app. As shown in the Country-fixed effects column , all COVID-19 proximity factors had significant positive associations with respondents' support for a COVID-19 tracing app. Specifically, having contracted COVID-19 was suggestive of a higher willingness to use a COVID-19 tracing app, as indicated by the substantially higher marginal effects of being tested . However, these marginal effects ceased to be statistically significantly different from those of the reference group in the multivariate models. The multivariate model further indicated that reporting anxiety symptoms was significantly associated with contact tracing app support. Similarly, having a family member or colleague who has ever contracted COVID-19 was also associated with greater contact tracing app support compared to not having such a family member or colleague . Similar to socioeconomic factors, the relationship between COVID-19 proximity and the willingness to use a COVID-19 tracing app appears to vary across the four countries studied. For instance, having no anxiety symptoms had a positive effect on COVID-19 tracing app support in Germany. Furthermore, having a colleague who tested positive for COVID-19 was associated with a higher likelihood of using a contact tracing app overall, but this was not the case in the Netherlands. In fact, having a family member who has ever contracted COVID-19 was the only indicator that yielded a significant positive marginal effect on contact tracing support across all countries. This implies that, in addition to including country-fixed effects and controls for sociodemographics and COVID-19 proximity in models, modeling the association between general health and COVID-19 tracing app support required us to ensure that the controls interact with the country dummies to account for heterogeneity. --- General Health Status We illustrate the association between general health status and the willingness to use a COVID-19 tracing app in Figure 1. For the leftmost graph of Figure 1, we estimated a baseline model that adjusts for timing and country-fixed effects. The graph suggests that both the "good" and "fair or bad" health statuses are positively associated with COVID-19 tracing app usage when compared to the reference category . The negative association between self-reported general health and COVID-19 tracing app usage persisted in terms of significance and magnitude when we also controlled for socioeconomic variables. However, as shown in the rightmost graph of Figure 1, the effect sizes of the poorer general health statuses were reduced by about one-third when adding the COVID-19 proximity variables; the estimated probability of contact tracing app use based on having a "good" health status and "fair or bad" health status increased by 4.6% and 6.3% , respectively. This attenuation suggests that the relationship between general health status and support for a contact tracing app partially operates through recent personal health scares or forms of stress that are related to direct experiences with the COVID-19 pandemic . In order to correct the estimates for country heterogeneity in the relationship between the socioeconomic and COVID-19 proximity variables and the outcome variable, we analyzed the marginal effects that poorer general health has on COVID-19 tracing app support while adjusting for the interactions between all covariates and the country dummies . As the point estimates in the baseline and socioeconomic models presented in Figure 2 are nearly identical to those in Figure 1, we conclude that neither timing or socioeconomic heterogeneity influences the identified relationship between general health status and the willingness to use a COVID-19 tracing app across the four European countries. However, adjusting for the country heterogeneity in the COVID-19 proximity factors yielded evident null effects . Hence, the association between general health status and COVID- 19 --- Discussion --- Principal Findings An important component of public health policy with regard to the spread of COVID-19 is the possibility of using mobile phone-based contact tracing in response to a positive COVID-19 test [2,3]. This is particularly relevant outside of major peaks in SARS-CoV-2 infection rates and lockdowns because it could help with avoiding rapid and uncontrollable disease transmission within communities [4]. COVID-19 tracing apps have been introduced in several countries. The extent to which these contact tracing apps can have a positive effect on public health is dependent on their uptake [5][6][7]. We found considerable country-based variation in the willingness to use a COVID-19 tracing app, which ranged from 16.2% in Germany to 50.2% in Italy. In addition, we found evidence indicating that several socioeconomic and demographic factors are associated with the willingness to use a COVID-19 contact tracing app. Such evidence was also found for the following COVID-19 proximity variables: exhibiting depression and anxiety symptoms, being tested, and having family members or colleagues who have ever contracted COVID-19. Importantly, based on these dynamics, we found that poorer health statuses are associated with significantly higher support for COVID-19 contact tracing apps. We examined indicators of COVID-19 proximity because, based on exiting literature, we presumed that being confronted with the detrimental effects of the pandemic within one's social network may trigger the motivation to install and use a COVID-19 tracing app [8][9][10][11][12][13]. Although we cannot observe changes in individuals' attitudes or behaviors in response to COVID-19 proximity over time, our findings support this relationship. We observed some cross-national variation in the significance levels of the COVID-19 proximity indicators. Nonetheless, having a family member has ever tested positive for COVID-19 appears to the strongest and most consistent independent variable of the increased intended usage of a contact tracing app. We argue that these results are suggestive of a relationship between the social context of the consequences of the COVID-19 pandemic and individuals' perceived risk. Given the cross-national variation in the associations between our two sets of control variables , we fitted a comprehensive model that adjusted for this heterogeneity [14][15][16]. The results from this model were straightforward. We found that having a poorer general health status positively affects the willingness to use a COVID-19 tracing app, even after adjusting for socioeconomic factors, indicators of how close the pandemic came to an individual , and regular levels of smartphone usage. Our results indicate that this relationship is moderated by COVID-19 proximity. In other words, it is plausible that the association between general health status and contact tracing app uptake is affected by how close the pandemic has come to an individual. It is also important to note that the recent loss of one's job or main income source during the pandemic yielded a significantly lower marginal effect on the willingness to use a contact tracing app in Spain, Italy, and Germany , and similar results were observed for long-term inactivity in the Netherlands. We suspect that some of these associations reflect the impacts that economic security and insecurity have on sentiments regarding the pandemic or even a broader rejection of government policies. These dynamics require much more in-depth research in the social science field. --- Limitations Our analyses relied on cross-sectional data that were obtained during the COVID-19 pandemic. We conducted several robustness checks to avoid having strong selection bias in the reported marginal effects of general health status, socioeconomic characteristics, and COVID-19 proximity. Such bias is likely to result from the underrepresentation of individuals with the poorest general health conditions-a demographic group that is difficult to include in all kinds of social surveys. Nationally representative panels would be the preferred data source for future research on the relationship between health and any kind of COVID-19-related measures. This is because longitudinal data are better equipped to measure health-related attrition. Furthermore, panel data are also best suited for effectively measuring the degree to which attitudes and behaviors of individuals change over time as a function of their health status or in response to the acquisition of new information. Future research may also benefit from expanding the operationalization of health risk and risk perceptions, such as those related to wearing a mask in close proximity to an individual. --- Conclusions This paper builds upon existing evidence indicating that contact tracing apps are an important element of public health [2,3] and that their positive effect is dependent on their uptake [5][6][7]. We studied whether general health status and COVID-19 proximity can be linked to contact tracing app uptake. This research question was motivated by a discussion in public health literature about the necessity of effective contact tracing in combating the COVID-19 pandemic as well as research in the social science field regarding the individual-level drivers of attitudes toward contact tracing apps [8][9][10][11][12][13]. We conclude that poorer general health statuses are positively associated with the willingness to use a COVID-19 tracing app. Moreover, the extent to which one's general health status impacts their likelihood of using COVID-19 tracing apps partially operates through the pandemic-related experiences that occur in their social circle. To date, public debates have mainly revolved around issues regarding apps' capacity to meet data privacy goals and legislation criteria. We suspect that the country-based variation we found in people's willingness to use a COVID-19 tracing app reflects path-dependent societal dimensions, such as large personal data leaks in recent history or underlying distrust in the government [14]. This implies that public policies that are intended to expand the usage of digital COVID-19 contact tracing apps always have to consider country-specific societal concerns. Our study suggests that once these conditions are met, public health policies that aim to increase contact tracing app uptake would benefit from campaigns that stress these apps' benefits for users , their family members, and the economy [1]. --- Conflicts of Interest None declared. --- Multimedia Appendix 1 --- Survey questions. [DOCX File , 20 KB-Multimedia Appendix 1] --- Multimedia Appendix 2

Background: Contact tracing apps are considered useful means to monitor SARS-CoV-2 infections during the off-peak stages of the COVID-19 pandemic. Their effectiveness is, however, dependent on the uptake of such COVID-19 apps. Objective: We examined the role of individuals' general health status in their willingness to use a COVID-19 tracing app as well as the roles of socioeconomic characteristics and COVID-19 proximity.We drew data from the WageIndicator Foundation Living and Working in Coronavirus Times survey. The survey collected data on labor market status as well as the potential confounders of the relationship between general health and COVID-19 tracing app usage, such as sociodemographics and regular smartphone usage data. The survey also contained information that allowed us to examine the role of COVID-19 proximity, such as whether an individual has contracted SARS-CoV-2, whether an individual has family members and colleagues with COVID-19, and whether an individual exhibits COVID-19 pandemic-induced depressive and anxiety symptoms. We selected data that were collected in Spain, Italy, Germany, and the Netherlands from individuals aged between 18 and 70 years (N=4504). Logistic regressions were used to measure individuals' willingness to use a COVID-19 tracing app.We found that the influence that socioeconomic factors have on COVID-19 tracing app usage varied dramatically between the four countries, although individuals experiencing forms of not being employed (ie, recent job loss and inactivity) consistently had a lower willingness to use a contact tracing app (effect size: 24.6%) compared to that of employees (effect size: 33.4%; P<.001). Among the selected COVID-19 proximity indicators, having a close family member with SARS-CoV-2 infection was associated with higher contact tracing app usage (effect size: 36.3% vs 27.1%; P<.001). After accounting for these proximity factors and the country-based variations therein, we found that having a poorer general health status was significantly associated with a much higher likelihood of contact tracing app usage; compared to a self-reported "very good" health status (estimated probability of contact tracing app use: 29.6%), the "good" (estimated probability: +4.6%; 95% CI 1.2%-8.1%) and "fair or bad" (estimated probability: +6.3%; 95% CI 2.3%-10.3%) health statuses were associated with a markedly higher willingness to use a COVID-19 tracing app.Current public health policies aim to promote the use of smartphone-based contact tracing apps during the off-peak periods of the COVID-19 pandemic. Campaigns that emphasize the health benefits of COVID-19 tracing apps may contribute the most to the uptake of such apps. Public health campaigns that rely on digital platforms would also benefit from seriously considering the country-specific distribution of privacy concerns.

BACKGROUND In 2013, it was estimated there would be 187,600 new cases of cancer and 75,500 cancer deaths . Since cancer prevalence and survival are both on the rise, more Canadians will require ongoing cancer treatment, surveillance, and supportive care . As a result, individuals living in rural areas will continue to experience health disparities while coping with the burdensome travel associated with a cancer diagnosis, treatment, and follow-up care . Although most individuals make a conscious choice to reside in a rural community, rural health care is nonetheless an important public health concern. The literature suggests that individuals enjoy rural life for many reasons, including a close-knit community of support, positive cultural experiences, a sense of peace or harmony, and enhanced autonomy in daily life . However, individuals living with cancer may choose to leave rural areas so they are closer to specialized medical care. Even though individuals may prefer to remain in their place of residence, rural geography can become a major obstacle to receiving necessary care. Canadians who live in rural or remote communities have ongoing challenges in the delivery of health care and access to services. Since delivery of rural oncology care varies across provinces and territories, the scope of available services will differ across the country . Rural areas tend to have limited access to primary health care services, difficulties with recruitment and/or retention of health care providers, and scarce access to specialized treatments and Supportive care needs of rural individuals living with cancer: A literature review diagnostics services . Additionally, the ongoing reality that urban centres will remain the centre of excellence for oncology care in many provinces across Canada further necessitates the need to explore oncology experiences in rural health care. The findings from this literature review suggest that individuals living with cancer in rural areas experience high levels of unmet needs that are dissimilar to their urban counterparts. Margaret Fitch and colleagues formulated a domain-specific, supportive care framework to assist cancer care professionals in conceptualizing the type of assistance required by their patients and engage in efficient planning for service delivery. According to this framework, all cancer patients require ongoing supportive care from a multi-domain perspective. Key constructs associated with the supportive care framework include "human needs, cognitive appraisal, coping and adaptation as a basis for conceptualizing how human beings experience and deal with cancer" . Patients' supportive care needs can be further categorized based on the nature of the individual's illness, from basic service provision to highly specialized or intensive and complex intervention. Humans have various needs related to physical, emotional, psychological, practical, social, and spiritual entities and develop a collection of skills throughout the lifespan to meet their own needs. A cancer diagnosis can create new demands and sources of anxiety, as well as feelings of vulnerability and uncertainty when basic human needs are not met. Unmet human needs can lead to significant distress. Cognitive appraisal is a concept inherent to the supportive care framework and represents the series of stages that individuals go through in order to categorize an event. If the event evokes a threat that is beyond the individual's capacity to manage, then emotional distress will escalate. Coping with stressful situations, unmet needs, and emotional distress can be difficult in any circumstance. In the event of a cancer diagnosis, adaptation and the development of coping strategies is a complex process. Individuals will respond to a cancer diagnosis or its treatment in their own distinctive way. As a result, interventions tailored to each individual and based on the multi-domain framework can assist with meeting basic human needs and, thus, reduce emotional distress and improve adaptation to the illness experience. The supportive care conceptual framework can be used as an instrument for service or program development, a foundation to organize educational approaches in cancer care, and a model for designing research projects . --- PURPOSE The aim of this literature review is to explore the impact of rural or remote residence on the supportive care needs of individuals living with cancer. The supportive care framework was used in compiling the data, evaluating the research articles, guiding the analysis and discussion, and considering the relationships between the supportive care needs being reviewed. --- METHOD Computer databases were searched for research studies, dated from 1990 to present, that examined adult cancer patients' experiences in a rural setting. Grey literature, including web-based resources and reference lists, were also explored for studies that could be included in the review. Key words used for the search included cancer, supportive care, rural, qualitative, quantitative, research, adult, and travel. Other similar MESH terms were included in the search strategy. Criteria for inclusion of articles in the review included the following: 1) written in English only; 2) published in a peer-reviewed journal between 1990 and present; 3) primary sources cited a specific qualitative, quantitative, or mixed research design; 4) adult patients with a diagnosis of cancer as the sole research participants, or participants within the context of the family; and, 5) patients from rural or remote areas as the sole research participants. Comparisons to urban populations were accepted in the literature review to provide insight into the differences in the urban/rural challenge variations. The search strategy revealed more than 200 articles that were selected for possible inclusion. Abstracts were scanned to determine if they were suitable for the purpose of the review, and 60 articles were selected for further evaluation. Twentythree articles met the outlined criteria and were included in the review. Articles were excluded based on the following: secondary sources, service provision, non-cancer diagnosis, survival outcomes, and descriptive reviews. Table 1 outlines the following information about each study included in this review: authors' names, study design, location of the research, sample, cancer indication and research question. The authors individually reviewed each of the articles, summarizing the data according to author, study design, location, sample, research question, data collection/analysis, and findings. The data were categorized according to the domains in the supportive care framework and a collaborative thematic analysis was completed to identify any variations or similarities in the materials. Once consensus was achieved, the authors of this review concluded that data saturation had occurred. --- LITERATURE REVIEW FINDINGS The findings in this review have been categorized according to the supportive care framework , under the following domains: physical, informational, social, practical, spiritual, psychological, and emotional. The data extracted from the articles are specifically related to the challenges or benefits of living in rural or remote communities during the cancer experience. Although the results of this review suggest that individuals in both rural and urban settings often have similar unmet needs, the aim of this review was to focus on specific details related to individuals living with cancer in rural areas. --- continued on page 161… --- Breast To assess the prevalence and predictors of perceived unmet needs among women diagnosed with breast cancer, with a particular focus on the impact of location. . Qualitative. In-depth interviews. --- Toowoomba, Australia Convenience sampling. n= 14 women and n=3 men. --- Variety of cancer diagnosis. To evaluate experiences of rural people with cancer travelling to a major city for radiotherapy treatment. Martin-McDonald, Rogers-Clark, Hegney, McCarthy, & Pearce, . Qualitative. In-depth interviews. --- Queensland, Australia Purposive sample. n= 39 female and n=7 male Variety of cancer diagnosis. To explore the issues related to rural people with cancer whose choice of radiotherapy treatment necessitated travel and accommodation in a metropolitan centre. --- Queensland, Australia Purposive sampling. n=115 Hematological Malignancy. To explore the experience of patients and their families after returning home to a rural or remote area after receiving treatment for hematological malignancies. McGrath, Patterson, Yates, Treloar, Oldenburg, & Loos, . Qualitative. In-depth interviews. Queensland, Australia. Purposive sampling. Women with breast cancer n=24 and service providers n=3. --- Breast To provide a preliminary indication of the support needs of post diagnosis women with breast cancer in remote and isolated areas. --- Breast To study the effect of distance to the nearest radiation treatment facility on the use of post mastectomy radiation therapy in elderly women. continued on page 161… Physical domain. Symptom management of cancer-related problems and treatment-related toxicity is an integral role of the nurse . Management of symptoms, such as pain, nausea, fatigue, diarrhea, myelosuppression, dyspnea, sleep disturbances, and mucositis can pose significant challenges for nurses and directly impact the quality of life of the individuals living with cancer . Many of the studies included in this review found that individuals who had to travel lengthy distances to receive care experienced exacerbations of treatment side effects and disease symptoms . Three of the studies exclusively examined individuals who travelled to urban centres for radiation therapy. It was found that side effects of the radiation treatment, such as burning, fatigue, pain, and nausea, made travel very problematic . In addition, unmet needs were noted in the physical domain as a result of the lack of medical or nursing support services or palliative care expertise in the rural/remote settings . In rural communities, management of the physical symptoms of illness places a great demand on family members. Family caregivers living in rural areas are often required to take on a momentous role in assisting their ill family member with physical care and management of distressing symptoms. Family members reported this role was stressful and emotionally draining. Moreover, they felt unprepared to deal with medical problems, care for certain devices such as central lines, and administer certain medications such as morphine . --- Informational domain. The operational description of this domain is related to the information required during the cancer trajectory. This includes information regarding side effects of treatment, drugs, treatment regimens, care processes, and other services. Therefore, this domain is directly related to service provision and access to professional assistance . The most noteworthy finding in the reviewed studies was that participants face challenges related to service provision in their rural communities. Of the 23 studies reviewed, 10 articles revealed service provision challenges such as lack of access to cancer information, limited access to professional or hospital assistance, and isolation from professional support or other supportive services were enduring challenges . As well, the participants expressed feeling unprepared to deal with medical problems or crises and reported that delays in diagnosis, inadequate support for symptom management, and limited access to additional home care supports were also significant challenges . In the rural setting, there is limited access to information, support groups, home care, trained specialists, treatments, and tests. Limitations in service provision can often lead to delays in diagnosis or elicit treatment-related decisions that may not be in the patient's best interests . It was reported that rural family doctors may not possess appropriate or current information, recognize symptoms of illness, or have the expertise to make a cancer diagnosis, all of which may lead to delayed diagnosis and/or compromised quality of care . Alternatively, Baldwin and colleagues revealed that individuals living with colorectal cancer preferred to travel to urban centres because they were more confident with the care provided by highly specialized teams in these settings. In two studies by Wilkes and colleagues , it was reported that nursing played an integral role in providing support and information services to clients and families living in rural settings. The nurses acted as coordinators of care and sources of strength for the patients and their families. Many had some expertise in cancer or palliative care situations and were able to facilitate access to necessary equipment and provide education regarding service availability. Other sources of informational support included books, pamphlets, internet, and networking with other family members or friends in similar circumstances . --- Social domain. In many instances, a cancer diagnosis affects the entire family. The social aspect of cancer involves the immediate or extended family, close-knit neighborhood groups, and the wider community. The studies in this review revealed that nurturing supportive networks and close-knit communities often exist within rural areas . Alternatively, challenges related to lack of privacy, feelings of isolation, and difficulties finding the appropriate support or support groups at home were also identified . Studies noted that patients found it difficult to cope with the burden of care placed on their family or friends. Additionally, patients found it equally difficult to be isolated from their loved ones. Families reported that the cancer experience was associated with disruptions in family life, changes in living arrangements and roles, caregiver burden, and relationship problems associated with spending extended periods of time apart. In addition, various studies revealed that patients were often worried about their partners or children while they received treatment and were concerned about the burden placed on the family to maintain the home or organize child care . Furthermore, patients had frequently reported feeling less connected with and supported by their families during extended periods of absence, noted they had no one to accompany them to the city for treatments, and indicated that aspects of social life often had to be put on hold, leading to feelings of isolation . Alternatively, some positive aspects of support were also revealed in the studies. For example, living away from one's home in a hostel setting played a vital role in the overall cancer treatment experience. In these settings, patients were able to establish supportive networks with others in similar situations and share the cancer experience. These supports helped counteract the loneliness associated with being away from home for extended periods and served as an immense source of emotional support. When patients returned home to their rural communities, they often reported that those close bonds were sorely missed . The role of social support in health is widely documented in the literature. Positive supportive networks play a role in improving health outcomes, decreasing physical and mental illness, and reducing mortality rates . Practical domain. Practical needs associated with the cancer experience encompass a broad range of tasks, such as housekeeping, finances, child care, transportation, housing, shopping, or legal needs . The financial hardships of cancer are related to two factors: 1) the dramatic decline in family income due to disability; and 2) the rising out-of-pocket expenses associated with a cancer diagnosis. This can lead to severe financial destitution from which some families never recover . Rural and northern residents who need to travel for medical treatments are at the greatest risk of financial devastation. For many rural individuals, travel expenses can exceed the cost of medications . Financial disparities associated with living in a rural setting include the costs associated with wear and tear on the family vehicle, fuel, accommodations, maintenance of two households, meals in restaurants, parking in the city, time away from work, telephone usage, and additional child care . Housing and accommodation are a basic human need. Affordable and supportive housing is a service that is vital to the well-being of rural cancer patients who must relocate to urban centres for care. When these services were available in the urban treatment centre, patients reported high degrees of satisfaction . Maintaining daily responsibilities at home and work while undergoing cancer treatment can be challenging for patients. Individuals living in rural settings must cope with the additional stress related to long travel days and frequent commutes to the city for treatment . Studies by Martin-McDonald and colleagues and McGrath both reported that patients struggled with maintaining their work responsibilities, feeling uncertain regarding job status, losing their workrelated roles, and coping with work-related challenges while undergoing treatment. Patients reported these were significant stressors that affected their overall quality of life . A large Canadian study of rural women with breast cancer identified that coping with changing work-life was a major theme during the cancer experience. Employment in rural and northern settings is often self-determined, part-time, or temporary and, therefore, associated with a lack of job security and benefits. The women in the study were frequently left without employment options during their illness because of the need to relocate to an urban centre for care . Other practical needs that markedly affect rural patients and their families include the time required to prepare for travel, arrange accommodations, and organize test schedules and medical appointments. Extensive travel, heavy city traffic, and poor weather conditions can intensify these challenges. Patients also expressed concerns related to feeling unsafe while walking in an unfamiliar city environment and having difficulty navigating the city or large hospital setting . The burdens associated with travelling to urban centres for cancer treatment are a significant finding in this review. In two of the studies reviewed, it was disturbing to note treatment decisions were made based on the burdens of travel. The first quantitative study by, Punglia, Weeks, Neville, and Earle described the effect of distance to the nearest radiation treatment facility on the use of post-mastectomy radiation in a large cohort of elderly breast cancer patients . The authors concluded that patients who resided the furthest distance from the treatment centre were least likely to use adjuvant radiotherapy based on the burden of travel . This decline in treatment access became statistically significant for individuals who lived greater than 75 miles from the nearest cancer treatment centre. Similarly, a concurrent study found that women with early stage breast cancer were less likely to proceed with post-operative radiotherapy after breast conserving surgery when they had to travel great distances to a treatment centre or when diagnosis occurred in the winter months when long-distance travel was more challenging . Radiation therapy is a standard of care for women with limited stage, invasive, or advanced breast cancers. It remains an important treatment modality alone or in combination with other therapies. Radiotherapy is often used to achieve local control of disease or prevent locoregional recurrence . It is alarming that treatment decisions may sometimes be made based on geography rather than on the best interests of the patient or an understanding of the treatments that are most likely to succeed. As a result, the burden associated with travelling for cancer care may carry serious implications. --- Spiritual domain. Spiritual wellness is a central dimension of this domain. Personal or religious beliefs and finding the meaning of life are inherent human needs. A cancer diagnosis can often trigger spiritual distress, existential despair, pain, and suffering . The spiritual domain was addressed in two of the studies included in this literature review. A qualitative study by Rogers-Clark explored the features of resiliency and suffering among rural women who were longterm survivors of breast cancer. The results revealed both positive and negative features in the context of spirituality. Rural life was described as a peaceful sense of balance and personal connection to the land. In addition, the authors also identified that patients experienced a sense of resiliency, strength and inner peace, community connectedness, independence, strength, and stoicism. The women in the study most notably articulated their connection to and appreciation for rural life and their "dislike" of urban life. Although the women also reported feelings of loneliness, threats to personal privacy, and a sense of isolation, the positive features of rural life far outweighed the difficulties associated with relocation to an urban centre. The authors concluded that rural living in the context of the illness experience should be considered an important health topic, yet tends to be ignored in the literature. Rural individuals may be able to draw upon a number of these strengths in response to stress. Therefore, interventions targeted at individuals living with cancer in rural settings should be geared toward programs based on supportive care that can enhance adaptation to the cancer experience . Pesut and colleagues conducted a mixed method study to gain an understanding of the experiences of rural cancer patients who had to travel to a regional centre for palliative treatment. The participants in this study described the features of rural life that were most important to them, including physical beauty, peaceful living, privacy, community support, and proximity to religious affiliations. The authors concluded that rural lifestyle is an important component of overall well-being for these individuals and that disruption, such as re-location to an urban setting, can lead to emotional and spiritual distress . These studies by Pesut and colleagues and Rogers-Clark add to the body of literature that emphasizes the value of rural living in the context of overall well-being, particularly spiritual and emotional wellness. Many of the rural attributes and features revealed in these studies are consistent with other research and scholarly works . This further demonstrates the value of community-based supportive care programs and access to oncology care resources in rural areas both of which can contribute to positive patient outcomes and enhanced well-being. Psychological and emotional domains. The primary feature of the psychological domain is the ability to cope with the illness experience and its associated disparities. The emotional domain is characterized by a sense of belonging, comfort, and understanding that exists during times of vulnerability. Challenges within these domains may include grief, uncertainty, fear, anxiety, loss, and depression. Within these domains, patients may experience fears of recurrence or changes in body image or self-worth. All cancer patients experience some degree of psychological and emotional stress during the course of the illness experience. Patients' ability to cope with these stressors or receive the required emotional support facilitates adaptation . It was frequently reported throughout studies in this review that, in the rural context, the cancer experience creates significant psychological and emotional disruptions for patients and their families. Cancer care of rural patients presents a multifaceted and complex problem. Many disparities that have been highlighted, including the burden of travel and its associated challenges, the financial, practical, and social disruptions in everyday life, the lack of services and supportive care in rural settings, and feelings of isolation, will have an impact on the psychological and emotional well-being of individuals and their families. The following are some specific examples from the literature review in which living with cancer in a rural setting directly impacted individuals' emotional and psychological well-being: lack of services in the rural setting led to feelings of anxiety among family members and caregivers ; difficulties with managing practical needs created undue stress ; the need to travel into an urban centre led to feelings of fear and uncertainty ; the lack of privacy in the rural setting led to embarrassment and isolation ; disturbances to family life were associated with a sense of loneliness ; long waits and delays in treatment created uncertainty and anxiety ; the isolation associated with treatment in urban settings led to "homesickness" and the need to discuss emotional concerns with people in similar circumstances ; and the challenges of coping with medical crises in a rural community caused significant distress to patients and families . These are examples of challenges that disproportionately affect rural cancer patients and have the potential to lead to severe emotional and psychological distress. In one study, Burris and Andrykowski compared the mental health outcomes of rural and non-rural cancer survivors. The rural participants reported higher rates of anxiety, depressive symptoms, overall distress, and poorer mental health functioning and life satisfaction when compared to urban participants. These results are consistent with the many disparities noted in this review in the context of rural living and the cancer experience . Alternatively, results of this review suggest that there are many positive features of rural life that contribute to the psychological and emotional well-being of individuals living with cancer. Examples of positive features of rural life include: the positive culture and supportive community networks ; a sense of peace and the benefits of being part of a close-knit community ; survivorship and stoicism developed through past experiences of living in a rural setting ; individuality and connectedness to the community ; and support from the wider community . Moreover, emotional and psychosocial well-being was often maintained even when relocation was required. Factors that contributed to a greater sense of well-being while away from home included staying in affordable and comfortable accommodations, developing close ties with individuals in similar circumstances, fostering a sense of comfort by being close to expert oncology services, alleviating the demands of driving every day, or spending additional quality time with significant other . --- DISCUSSION This review of the literature suggests that there are distinct and unique experiences that present additional stressors and challenges for individuals living with cancer in rural areas. As such, rural cancer patients face distinctive social disruptions and service provision challenges when compared to their urban counterparts. Rural living is often associated with delays in diagnosis, challenges of travelling long distances to urban centres, limited access to oncology services, economic disparities, and treatment decisions that are made based on geographic location. The cancer experience can be long and difficult. In addition, the unpredictable nature of the disease creates uncertainty, anxiety, and ongoing emotional, spiritual, or psychological distress. As a result, it is imperative to isolate the additional burdens and unmet needs of rural Canadians living with cancer. This is invariably a complex process and is beyond the scope of this review, as only four of the studies under review were completed in Canada. The supportive care framework is a valuable tool to understand the global needs of cancer patients from a person-centred care approach. As well, aspects of the framework are essential components of care for all health care practitioners and associated agencies . Cancer patients receive care in many settings, from primary care clinics to outreach programs and urban cancer centres. It is clear from this review that the rural challenges in cancer care service delivery are multifaceted, comprehensive, and occur across many dimensions of health care. In order to improve the quality of care provided to rural individuals, we must recognize service gaps, develop tailored, multi-domain intervention strategies, and engage multiple disciplines within the global health care community. Considering the rising incidence and prevalence of cancer in Canada, as well as the number of individuals living with cancer in rural and remote areas of the country, health care practitioners, administrators, researchers and policy makers must pay particular attention to this important health care topic. Clinical standards can guide the establishment of interventions and strategies. The supportive care framework outlines that ongoing assessment of supportive care needs, access to resources, capabilities for self-referral, and ongoing informational support, while also considering unique individual variations, should become a standard of care . Interventions that should be at the forefront of health care activities include incorporating supportive care into the organizational structure, improving the ability to recognize and respond to patient and family distress, and enhancing access to specific services and informational support . The supportive care framework provides useful information at an organizational level to assist in developing tailored interventions for this vulnerable population of cancer patients. It can also provide nurses with a framework for assessing the unique challenges encountered by this population. This review highlights that people choose rural life for reasons that directly affect their quality of life and sense of well-being. As a result, the benefits of rural living need to be considered when caring for patients from rural communities. Front-line nurses, both in the urban and rural settings, remain integral to the management of oncology populations, as they are well positioned to implement positive changes in the provision of care. Tailored or individualized interventions represent a customized approach to intervention management, while considering the unique characteristics of a specific client group . Delineation of this process requires assessment of the client characteristics, development of interventions that address a targeted problem based on those characteristics, and detail an algorithm that customizes the application approach. This thoughtful process to intervention development contributes to personally relevant information or services meaningful to the specific client group. In addition, it has the capacity to motivate and empower individuals and, ultimately, impact health, well-being and quality of life . It is essential that we develop a deeper understanding of this important health care topic to tailor interventions to the needs of this client group. There are several methodological limitations that affect the generalizability of the findings in this review. Studies that utilized multiple sampling methods, examined diverse populations of individuals with varied diagnoses and ages, included English language only, investigated multiple treatment methods and used different research designs were incorporated in this review, thus making it difficult to generalize to a specific population of cancer patients. Relatively few studies were found from the Canadian population. The majority of research retrieved was from Australia, the United States, and international populations. Canadian living is associated with distinctive challenges, such as long winters, poor driving conditions, and great distances between towns and cities. Since many of the studies reviewed were international-based, it is also difficult to ascertain the variations that exist between service provision and supportive care services in the different settings. Based on these limitations, further research is warranted for Canadians living with cancer in rural areas. Future research should focus on studies that examine supportive care services in rural Canadian communities, based on location, economic disparities, and supportive, psychosocial, medical, and nursing services required to improve the health and well-being of rural Canadians living with cancer. Based on this literature review, the authors will explore a qualitative research proposal. The aim of the proposed study will be to examine the supportive care needs and health service challenges of individuals with cancer who live in rural and remote communities in Manitoba. --- ConCLUSion Cancer diagnosis and treatment represents a life-altering and extremely difficult experience for both patients and their families. Geographic location of the patient and family creates additional obstacles that must be considered. This literature review outlines some key findings that could significantly impact quality of life, health, and wellness of individuals living with cancer in rural settings. The review contributes to a body of evidence-based knowledge and may help inform future practice and research endeavours. The supportive care framework utilized in this review is a well-designed model that can guide work in this area of practice and research. It is beneficial to advance the understanding of the additional burdens experienced by rural Canadian populations and the unique characteristics of rural individuals, as well as specific services required, so that tailored cancer care interventions can be incorporated. In addition, the benefits of rural living cannot be overlooked and must be considered in this process. In order to develop meaningful health and social services, and help mitigate the impact of cancer and improve the quality of life for rural individuals living with cancer, research studies must situate the experiences and needs of rural Canadians living with cancer at the centre of inquiry.

Regardless of geographic location, the cancer journey is an extremely difficult experience for both patients and their families. The aim of this literature review is to explore the impact of rural or remote residence on the supportive care needs of individuals living with cancer. This review included ten qualitative, seven quantitative, and six mixed design studies. Data collection, analysis, and evaluation were guided using a multi-domain supportive care framework based on seven domains : physical, emotional, informational, psychological, spiritual, social, and practical (Fitch, 2009). This review will suggest that there are distinct experiences that present both challenges and benefits to individuals living with cancer in rural areas. These findings will be detailed with recommendations, and grounding for future research outlined.

performed qualitative interviews with adolescents and SRH service providers to know their perspectives on how the pandemic influenced the SRH of adolescents. This paper provides insights into the barriers faced by adolescents while accessing SRH services during the pandemic, as well as their perceptions of digital strategies, such as mobile applications, and other recommendations for supporting SRH education and services. Based on the study findings, an adolescent-friendly mobile application will be developed to provide a virtual platform connecting adolescents to SRH educational resources, services, and support. --- Background Globally, the COVID-19 pandemic led to drastic changes in the access and delivery of healthcare services, influencing the physical, psychological, social, and emotional well-being of individuals, and profoundly impacted adolescents [1,2]. The World Health Organization defines adolescence , as well as preventing negative consequences to SRH [5][6][7]. A Canadian Community Health Survey reported that more than half of 15-24-yearolds in Canada were sexually active and at risk of unintended pregnancy and/or STIs [8]. Adolescents in Canada give birth to close to 13,000 newborns each year [9]. Adolescent mothers are at increased risk of mental health problems, substance use disorders, domestic violence, STIs and recurrent pregnancies, which are often associated with lower socio-economic status, poverty, and lack of social support [9,10]. Wong et al. [11] found that youth experiencing economic or social marginalization were more likely to have poorer SRH outcomes. Adolescents often experience a range of intersecting barriers making it difficult to access appropriate SRH services [6,12]. The pandemic imposed additional barriers further reducing access to SRH services. Prioritization of resources for essential COVID-19 response over SRH services resulted in the closure of safe abortion sites, reduced supply of contraceptives, and suspension of SRH services. As a result, this is expected to increase cases of genderbased violence , unplanned pregnancies, unsafe abortions, and maternal mortalities [2,13,14]. The closure of schools and SRH clinics limited opportunities for adolescents to gain comprehensive SRH education and utilize SRH services or products [15,16]. Economic constraints, loss of income, and limited health insurance coverage during the pandemic further impeded seeking SRH care, undergoing STI testing/treatment, or purchasing contraceptives [7,17]. The impacts of the pandemic are still unfolding and there are potential long-term consequences that will shape adolescents' SRH. However, the impacts of these measures on the SRH of adolescents in Alberta, Canada is relatively unknown. Understanding the effects of these measures is a crucial step towards informing effective services, supports, and strategies to improve adolescents' access to SRH services. Therefore, the overall purpose of our study was to investigate the impact of the COVID-19 pandemic and associated public health measures on the SRH of adolescents in Alberta. Specifically, the study addressed three research questions: 1. What are the perspectives of adolescents and service providers on how the pandemic has influenced the SRH of adolescents in Alberta? 2. How did pandemic-related public health measures shape access to SRH services by adolescents? 3. What types of services, programs, or supports for SRH did adolescents use during the COVID-19 pandemic? --- Methods --- Theoretical framework We used Levesque's conceptual framework of access to health care [18] to understand and examine adolescents' access to SRH services during the pandemic. It provides a comprehensive and multidimensional view of healthcare access in the context of the healthcare system with dimensions of approachability, acceptability, availability/ accommodation, affordability, and appropriateness, along with considering the population's socioeconomic determinants and abilities to perceive, to seek, to reach, to pay, and to engage, in healthcare [18,19]. We utilized these principles to understand adolescents' access from both provider and adolescent dimensions. --- Study design We conducted this study in two phases over 2 years . The first phase of the study consisted of qualitative interviews with adolescents and care providers, whereas the second phase consisted of co-designing an adolescent specific SRH website based on the knowledge generated from the interviews. In this manuscript, we are presenting the findings from the qualitative interviews. --- Phase 1: qualitative interviews To enhance the integration of research findings into practice, programming, and policy and to promote the co-creation of knowledge within the study, we used an interpretive description approach combined with integrated knowledge translation [20]. ID facilitates knowledge development by capturing the subjective experience and perception of the group under study advancing the understanding of the phenomenon under study [21]. We applied the principles of ID in our study through purposeful sampling, concurrent data collection and analysis, and ensuring rigour [20]. An Adolescent Advisory Group , pre-established by SM in Edmonton, was engaged in this study to contribute to the recruitment process, data collection and development of KT activities . --- Participants and recruitment We used a multi-faced community-based strategy to recruit adolescents, which involved social-media marketing campaigns, university-wide weekly email campaigns, and promotion of the study in the university campus and clinics. In collaboration with our community partners who provide care to vulnerable youth, we recruited participants from various backgrounds, socioeconomic standing, and gender identities to ensure adequate representation. We also recruited SRH service providers through Adolescent Medicine Clinic and Birth Control Centre in Edmonton. Participants were included in the study based on the following inclusion criteria: Adolescents aged 15-19 years old who were seeking or had accessed SRH services in the past 5 years or healthcare professionals who provided SRH care to youth, living in Alberta, and fluent in English. We included older adolescents in this study as their SRH needs are different from younger adolescents . We used purposive sampling to deliberately choose participants based on the inclusion criteria, who would provide relevant and rich descriptions of their SRH experiences. Informed consent was obtained from all participants and additional parental consent was obtained if it was determined that adolescents lacked decision-making capacity. A total of 18 adolescents and 15 healthcare providers were recruited for the study. This sample size was determined a priori and deemed adequate to gain a deeper understanding of participant viewpoints while acknowledging that there may be differences in perceptions and outliers [20]. Due to COVID-19 restrictions, all recruitment and subsequent data collection occurred online via E-mail and Zoom between November 2021-January 2022. --- Data collection Data collection was conducted from 2021 to 2022. Ethics approval was obtained from the University Research Ethics Board. Data were collected through a semi-structured interview guide , which was co-developed with the AAG, and included open-ended questions on the impact of COVID-19 on everyday life, SRH, access to services, and sources of support. Interviews were conducted by SM with expertise in qualitative methods and SRH of adolescents along with a trained research assistant. Interviews lasted for 30-60 min and were audio-recorded through Zoom. Audio recordings were transcribed to remove all identifying information. The demographic data of participants were collected through secure Google forms. --- Data analysis Conforming to ID methods, we performed data collection and analysis concurrently through an iterative process [20]. Data analysis was conducted in four stages: Audio recordings were transcribed --- Rigour To achieve reliability and validity, we employed verification strategies to identify when to continue, stop, or modify the research process. Verification strategies included: Methodological coherence; ensured congruence between the research questions and components of ID, appropriate and diverse sampling to ensure efficient and effective saturation of categories with optimal quality, collected and analyzed data concurrently, developed a coding system that was discussed and verified with research team members including AAG, recorded field notes and evaluated the quality of the interviews consistently by analyzing the participant's responses to the questions and modifying the questions accordingly. --- Results --- Participant characteristics See Tables 2,3. --- Key findings Three broad themes emerged from the adolescents' and service providers' qualitative interviews: COVID-19 SRH experience, barriers to SRH, and adolescent SRH service access strategies. These themes were then further divided into sub-themes, which are presented below. --- 1) COVID-19 SRH experience The COVID-19 pandemic influenced many institutions including adolescents' access to sexual and reproductive health. We found that adolescents had both positive and negative experiences during the pandemic. --- Impact of the COVID-19 pandemic Many adolescents reported both positive and negative changes in their SRH services before and during the pandemic. These changes were associated with the challenges of accessing walk-in clinics and longer times to seek SRH services: …It was definitely a lot harder to find that kind of access as well because of I assume, less staff working --- Number of years in Canada Less than 1 year and more pandemic measures, so, while I assume it should have taken about a week it took a month to get the service that we did. 1 Service providers commented on the de-prioritization of SRH services during the pandemic: "I think this kind of got pushed aside -like sexual health like wasn't people's priority, of course, with COVID going on." . Another service provider also mentioned: Definitely As consultation moved from in-person to online, some adolescents expressed their dissatisfaction with online services, especially with matters regarding counselling and family planning. They reported feeling uncomfortable with a healthcare provider in a virtual setting: --- I needed to talk to my doctor about something and I wanted her to like look at something or like have a visual she only does phone appointments now most of the time, and then, if she didn't do for an appointment it's like she's only in office, I think, like two days a week, and that's changed definitely because she used to be an office I think. Alternatively, many adolescents indicated online services to be a positive outcome of the pandemic: "Discussing with my doctor over the phone was very efficient and faster than it would have been in person" . Some participants stated that there was no impact of the pandemic on their SRH: "Personally, no. I don't think it's done anything like about my reproductive health, really. Like before and after, or even during, it was nothing's really changed. " . --- Availability and accessibility of SRH products and services Several adolescents reported the easy accessibility of using alternate services for their SRH needs and highlighted the use of newer resources like mobile applications to support their SRH. Although some adolescents appreciated the online services, many struggled with lengthy waiting times and securing SRH products . They recalled the ease of obtaining free contraceptives and medical advice from clinics on a walkin basis before the pandemic: "It's also a lot harder to find free like birth control, like condoms or whatever versus before I just would pick them up in my local clinic. " . Another participant also expressed: There's definitely a decrease in in-person services and I also think that there's much more of a wait time when it comes to picking up stuff like birth control or asking basic questions that I could easily have asked my physician before the pandemic. Service providers talked about the challenges that were introduced in the system during the pandemic: "During the worst of the pandemic, there was only one nurse. It was by appointment only, which kids don't do great with appointments. And it was symptomatic testing only." . Additionally, workforce shortages because of redeployment to COVID-19 response further compromised SRH operations: A service provider commented on how the pandemic led to the suspension of SRH outreach education initiatives: So, before the pandemic, our clinic did a lot of outreach, so we'd go to like -it's at the high school for kids that struggle in regular high school. We would go there. We would go to other little places around the community, and then with COVID, that obviously stopped. During the later stages of the pandemic when educational institutions switched to online platforms, SRH education was not given a higher priority: --- It definitely like affected it, because I missed out on my grade ten and eleven -well, sorry, eleven and half of twelve school year, so it was like -like that's like prime age for kids to start like exploring, and like I didn't get that sex ed during those years, so like it was definitely more of a figure it out yourself kind of thing that happened, to like the age group that I belonged in, because we didn't get taught it, because it wasn't a priority, so they didn't teach it over Zoom. Limited store hours Participants reported that reduced hours of stores during the pandemic decreased access to SRH products, like contraceptives and menstrual hygiene products: "I guess during the pandemic it definitely got a little bit harder, because of like the mandates and stuff like that, some stores were closed . Another participant expressed: The ability to go to the store became a little bit harder because of limited capacities and things a lot of stores have only 25% capacity or 30% capacity, so it was hard just having to like wait in lines and things but that's just like a general inconvenience. --- Another participant reported difficulties securing contraceptives: "Like if I was getting contraceptives, it was definitely harder to get during the pandemic because stores were closed down. " . Reduced supply of SRH products during the pandemic, such as contraceptives was also reported: "Lower store capacities or stock of items, especially when like toilet paper and, like other hygiene products things became a little bit harder. " . Lack of social support Many adolescents expressed the loss of social support and connectedness during the pandemic. Social distancing mandates made it difficult to meet peers and share common concerns: --- Like it's a lot harder to find a group like I said and to find people that you can rely on to support you as well, I felt like it was like I was discouraged during the beginning of the pandemic and during peak or like when there was more cases to go and get tested. Another participant talked about losing other forms of support, such as counselling: "Yeah, kind of both, because the only counselling I would have had access to would have been school, and when schools went online, there was really no way to do that. " . One participant expressed the difficult journey of exploring their sexual orientation during the pandemic without any support: --- So I found that has really impacted me and my sexual and reproductive health, during the pandemic I can't really talk about that, before because I don't think I would have gone through the journey that I've gone through with my sexuality if it wasn't for this pandemic but I do know that resources have been a lot harder for LGBTQ+ individuals during this time. --- 3) Adolescent SRH service access strategies We inquired about adolescents' perceptions of the development of a mobile application to access SRH resources and what strategies proved to be useful to them during the pandemic. Participants recommended strategies that they believed would be more beneficial for the adolescents. Outlook on SRH digital strategies The majority of the participants in this study supported the idea of developing a mobile application for SRH resources and indicated that it would increase access to SRH resources. One participant stated their preference for mobile applications over other digital platforms: --- Yeah, I definitely would recommend that I think would be really useful, especially because I'm on websites sometimes the university website sometimes just in general websites on the Internet, it can be hard to navigate them like they aren't made in a way that it's easy to know where things are located. Service providers shared their experiences of using digital strategies for SRH education: --- Like there's actually a lot of good technology out there that I knew about and could use. Most of the time it was tech problems on the school side. Like there was lots of problems getting -if there was two classes, getting both teachers to log in, because the person who made the Google lead had to let them in. I couldn't let them in. So, like there was like those little issues. On one hand, participants expressed that mobile applications would provide adolescents with the privacy to seek SRH help confidentially: "That's really cool and it could be super useful, especially for like people who are in families that don't feel comfortable talking about those things are no" . On the other hand, some had concerns regarding the safety features of such applications and the possibility of receiving unreliable information: --- Oh yes, and mobile apps I tend to be a bit more wary because there is a lot of misinformation and bad apps, I guess, but the online websites if it looks credible than I do, listen to what they say and they have helped me get more information on productive health and, what is needed to ensure that I'm being healthy and safe. Supporting SRH through various strategies We explored various strategies and recommendations from participants that would help in developing targeted interventions to support adolescents' SRH. Participants talked about various features to facilitate the delivery of SRH resources and provide a holistic platform for adolescents to receive support. One participant indicated the importance of a simple interface design: Maybe simplifying the online website would be helpful, so making the interface easier to navigate and having more information summarizedbecause sometimes some aspects aren't very clearly explained so just summarizing where things and having an easier interface would be very helpful. Another participant recommended anonymous forums where adolescents with similar interests can freely discuss SRH with care providers: Just a mobile app that makes information easy to consume because some people are like shy to look things up and just some maybe anonymous forums like so we can have this community of adolescents, to discuss freely about their problems rather than being embarrassed by having their public identity revealed so just some forums and some information and maybe some clinicians, if possible. Service providers highlighted how they supported adolescents' SRH by implementing various strategies: I did a lot of texting with our youth patients during the pandemic. And I found that that was a really helpful way to sort of deliver a lot of information at once delivered in a way that's like immediate and accessible to the youth. Along with videos, Zoom, like virtual sessions were also effective, because again, I can reach more rural. And we made a frequently asked questions PDF because usually we get really similar questions... And so that was really helpful because they can just read that and get their questions answered. Many participants also expressed seeing various features to help them get quick access to resources and nearby SRH services: --- I think that an app or like a website with like STI testing centers or different clinics and areas that you can go and get you know condoms birth control, whatever we can get all of this help just one app or one website that has all of that information there with very minimal searching. I think that would be something that would really benefit society because it's very hard to find information. One service provider indicated how virtual strategies can target adolescents effectively and reduce barriers to their SRH and well-being: I think like it's the younger generations that are really doing well with like the virtual appointments. I think if they could find a way to make access to that more easier, I think kids would be more comfortable getting tested, instead of having to -you know, it's intimidating for a 15 or 16-year-old kid to walk into a health building all by themselves, try to register themselves, sign in, go to the right desk, answer all these medical questions. So, it'd be nice if there was a more comfortable environment for them, perhaps. --- Discussion This study evaluated the impact of the COVID-19 pandemic on the sexual and reproductive health of adolescents and explored the factors that supported and promoted their SRH. We found that the COVID-19 pandemic had significant impacts on various aspects of the SRH of adolescents, including but not limited to access to SRH education, services, and products. The pandemic disrupted major sectors, including preventive healthcare, leading to the loss of basic SRH rights [21]. Diverting resources to the acute crisis and restricting other services were necessary, but prioritizing the SRH needs of individuals was also important as the pandemic did not halt those needs. In fact, those needs increased due to the additional barriers introduced in the existing system [22,23]. We used Levesque's conceptual framework of access to health care to understand and categorize the barriers that adolescents encountered while accessing SRH care during the pandemic, both from the healthcare system's perspective and individual perspective. Figure 1 presents the adaptation of the conceptual framework to highlight those barriers and challenges in the dimensions of approachability, acceptability, availability and accommodation, affordability, and appropriateness along with the individual factors that impact access to care [18,19]. Adolescents in our study reported interruptions in SRH services during the pandemic which affected their physical, emotional, sexual and mental health. These findings are consistent with several other studies globally where SRH services were partially suspended and deemed non-essential during the pandemic [21,[24][25][26]. Adolescents in our study reported the challenges of securing SRH products timely , which was brought on by the closure of schools, Fig. 1 Adaptation of Levesque's conceptual framework for healthcare access [18] stores, and in-person clinics. The unavailability of contraceptives [7] and in-person clinics reduced their ability to access SRH products and services. Despite the barriers, adolescents were found to be sexually active during the pandemic and therefore engaged in risky and unprotected sexual behaviours, leading to potential unpredicted outcomes [23,24]. Maier et al. [25] reported that several states in the U.S. suspended essential SRH services during the pandemic, especially abortion services, failing to satisfy the basic SRHR of individuals. Although not evident from our findings, the literature from lowmiddle-income countries suggest an increase in the number of unintended pregnancies and STIs among adolescents during the pandemic due to these service disruptions [13,23]. A study in the U.K. highlighted the disproportionate impact of SRH service suspension on adolescents as contraception and STI screening became less accessible and available [27]. However, instead of completely suspending essential SRH services during lockdowns, such as abortion care, alternative approaches were used to deliver SRH care through remote appointments and medical abortion treatment packages delivered on request [27,28]. Adolescents faced a major setback when physical distancing requirements prevented them from meeting their social circles and receiving support from their peers, counsellors or service providers. Loss of support systems and educational and extracurricular activities increased anxiety among adolescents, leading to increased risky behaviours, including unprotected sexual activity [23]. A few adolescents in our study expressed how all these factors eventually impacted their mental health. Magson et al. [29] reported longitudinal declines in the mental health of adolescents during the pandemic which were associated with the loss of social connectedness and increased parental conflicts. The shift from in-person to online SRH services was well appreciated by many adolescents in this study, as it not only prevented longer waiting times in high-risk COVID-19 environments but also made services accessible at the fingertips. Service providers also reported that online services improved access and gave them an opportunity to connect with adolescents. Regardless of the ease of access, some participants expressed their discomfort and dissatisfaction with meeting healthcare providers in a virtual setting. Although telemedicine created ease for some SRH issues, many other reproductive health issues necessitated physical examinations or laboratory investigations [30]. Our findings align with another study where adolescents expressed a lack of confidence while communicating with care providers online independently and confidentially [24]. The pandemic caused a massive shift from in-person SRH care to self-management through telemedicine and digital resources [13]. Adolescents in our study utilized digital SRH resources to access SRH information. They also communicated the need for adolescent friendly SRH websites and mobile applications where they can receive comprehensive SRH resources altogether. Mobile phone supported SRH interventions are needed as they have immense potential to offer safe, confidential, accurate, and timely information and services to adolescents [31], and can increase accessibility by making services more approachable, affordable, and available [18]. Regardless, there are some limitations that may restrict the acceptability and appropriateness of SRH digital interventions [18]. Our future efforts will focus on co-designing an adolescent friendly SRH website and mobile application to improve SRH access to adolescents, especially those who are at more risk of SRH inequalities. --- Strengths and limitations This study is the first to report the experiences of adolescents regarding their SRH during the COVID-19 pandemic in Alberta, Canada. We collected and analyzed data from both adolescents and service providers, which yielded a comprehensive understanding of the challenges faced by adolescents. Also, we recruited participants from various backgrounds to ensure a better representation of the sample. Due to the sample size and geographical limitations, the results may not be generalizable to the entire population. Data were collected only from Alberta, and the results could vary in other provinces and territories. We did inquire about adolescents' current level of education but did not inquire about their school system as this could highlight the similarities or differences in the provision of SRH education across education systems. We did not collect data on participants' level of sexual activity or contraceptive methods; this can be beneficial in future to determine the impact of public health restrictions on sexual activity and/or unintended pregnancies. While our study focused on the SRH experiences of adolescents aged 15-19 years, we acknowledge the limitation that some participants might have referenced SRH services accessed as early as age 10, and we did not consistently specify the exact age or nature of the service at the time of their experiences. Studies in the future can address the nature of SRH services accessed in different age groups. Also, we did not analyze the impact of the pandemic in relation to the socio-demographic characteristics of the participants, which could provide multi-dimensional perspectives on SRH experiences. Future research can explore using the principles of intersectionality and health equity in determining the SRH experiences of adolescents during disruptive events. --- Conclusion In conclusion, the COVID-19 pandemic had a profound impact on adolescents' access to SRH education and services. School closures, limited store hours, and a lack of social support contributed to an overall decline in SRH well-being. Adolescents faced numerous barriers in obtaining essential SRH products and services and experienced a knowledge gap because of the pandemic. Service providers also experienced barriers in reaching adolescents for SRH education and services during the pandemic. Most participants supported the development of mobile applications as they can increase access to SRH resources. By incorporating these insights, targeted interventions can be developed to enhance SRH wellbeing and empower adolescents to navigate their SRH confidently during public health crises. Future research and policy efforts should focus on developing strategies that accommodate both in-person and remote services, ensuring that adolescents can access SRH services according to their needs and preferences, and fostering social support networks for adolescents in times of crisis. --- --- --- Additional file 1. Interview guide for adolescents. Additional file 2. Interview guide for service providers. Author contributions SM conceived and co-designed this study. MA, BA, and SK conducted interviews. SM, MA, AR, SK and BA led the analysis and drafted the manuscript. All authors commented on all drafts of the article and approved the final draft. --- --- --- Competing interests The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. ---

The COVID-19 pandemic led to major service disruptions in the healthcare sector, especially regarding sexual and reproductive health services. However, the impact of the pandemic on Canadian adolescents is relatively unknown. This study aimed to investigate the impacts of the COVID-19 pandemic and associated public health measures on the sexual and reproductive health (SRH) of adolescents in Alberta, Canada. Methods A qualitative study using an interpretive description (ID) approach and community-based participatory research principles was conducted to capture the subjective experience and perceptions of adolescents and service providers. With the collaboration of the Adolescent Advisory Group and community partners, 18 adolescents and 15 service providers were recruited for the study through purposive sampling. Findings from the qualitative interviews were analyzed using thematic analysis.Three major themes emerged from the analysis: (1) COVID-19 SRH experience, (2) barriers to SRH, and (3) adolescent SRH strategies. Our findings highlight numerous barriers and challenges that prevented adolescents from accessing SRH education, products, and services.The COVID-19 pandemic had a profound impact on the SRH and the well-being of adolescents. Our study reflects the need for diverse SRH strategies to maintain continued access to SRH resources during disruptive events, such as the pandemic.Access to sexual and reproductive health (SRH) services is a basic human right. All individuals require access to appropriate SRH services to maintain their optimal sexual and reproductive health. Adolescents require special guidance, support, and youth-friendly services in matters of SRH as they enter puberty and explore their sexual identity. However, during the COVID-19 pandemic, many health and SRH services were suspended. Access to SRH products and services became difficult due to public health restrictions, which has possible negative consequences for adolescents' SRH. The experiences of adolescents during the COVID-19 pandemic regarding their SRH are not reported in Alberta, Canada. Therefore, we explored the impacts of these public health restrictions on adolescents' SRH. We

INTRODUCTION Recent legal decisions inside and outside the sport system, as well as cases of prominent sport stars , highlight the issue of transgender and intersex athletes' inclusion in sports and especially in competitive structures of sports. In discussions on this topic, transgender athletes are often blamed for challenging the binary sex-segregated system of elite and recreational sports, having unfair physical advantages, and calling fairness and the level playing field of sports into question. Conversely, sports organizations are accused of implementing discriminatory policies, of adhering to binary sex-segregation instead of considering gender identity or other competitive categories, and of systematically excluding transgender athletes and preventing the positive effects of sport on transgender individuals . These arguments about systemic bias are supported by evidence of individual attitudes toward and personal experiences of transgender individuals in sports . For some time now, transgender issues have been discussed in Anglophone research, mostly under the LGBT umbrella . This approach of lumping transgender athletes together with other sexual minorities has been questioned, because these groups face different challenges and prejudices in sports . In the European context, transgender inclusion in sports, as well as experiences of and barriers to transgender sports participation, are rarely quantitatively examined on a broader level. For example, quantitative research in the Netherlands has focused more on sexuality and sexual orientation , while research in Scotland has considered both gender identity and sexual orientation . Without reference to sports and sports participation, the European Union Agency for Fundamental Rights found that almost half of the transgender respondents in a study reported experiences of discrimination and harassment in the year prior to the study and that transgender individuals are often exposed to physical attacks. Furthermore, in the year prior to the study, 38% of transgender respondents were discriminated against in places and situations other than workplaces, amongst which sports clubs were included . Although transnegativity is perceived as a problem in sports , the European research corpus is truly deficient. This was the initial starting point for the current research, which is the first to provide a comprehensive picture of the situation and to present experiences of LGBT+ people undertaking sports in Europe 1 . The current paper focuses on two entangled topics: situations and experiences of cisand non-cisgender individuals 2 within LGBT+ communities in 1 As part of the project OUTSPORT the research was funded by ERASMUS+. The aim of OUTSPORT was to develop innovative and educational approaches to prevent violence and tackle discrimination in sports, based on sexual orientation and gender identity. The European Commission support for the production of this publication does not constitute an endorsement of the contents which reflects the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein. 2 The term "non-cisgender individuals" is used as a catchall term for transgender women, transgender men, non-binary and non-identifying individuals, and thus goes beyond the umbrella term "transgender." It is not intended to denote otherness or deviation. countries of the European Union, based on a quantitative online survey; and organizational strategies and perceived barriers and facilitators for LGBT+ inclusion, based on qualitative interviews with stakeholders from sports organizations in five European countries. --- SEX BINARY Sex binary is based on opposite categories and ignores the spectrum of sex characteristics . According to such categorization, a person is either a man/male or a woman/female, or in the words of Krane and Symons "to be male and masculine is not to be female and feminine" and vice versa. Whilst sex refers to biological aspects of the body , gender is a socially and culturally constructed reference for being masculine or feminine according to bodies and behaviors . The term transgender describes an incongruence between sex and gender: namely, the sex assigned at birth does not fit the inner feelings of gender identity , defined as "one's sense of one's self as a gendered person" . Following Enke , transgender "may include a gender identity that differs from the sex assigned at birth; a gender expression that differs from that conventionally expected of people according to their bodily sex; and/or a desire for alteration of the body's sex/gender characteristics." Transgender serves as a category that includes transgender males and females, non-binary persons , and sometimes also non-identifying persons . Gender expression focuses on how "people express, wear, enact, and perform gender through behavior, mannerism, clothing, speech, physicality, and selective body modification" . Therefore, gender expression can take many forms, ranging from styling and self-presentation, to the changing of names and pronouns, to hormonal and medical treatments . Persons who do not conform to the assumed gender roles and expressions are known as gender non-conforming individuals. Sports and physical activities are a way to express one's gender by engaging in what is considered typically masculine, feminine, or gender-neutral sports; playing sports in specific ways; or meeting/rejecting sports-related gendered role expectations. Perceptions of adequate sports engagement and acceptable sporting behavior are strongly orientated toward the sex binary and differentiate between accepted and valued behavior for boys/men or for girls/women . Girls/women still face several challenges in the domain of sport, in which masculine characteristics are promoted, and the masculine body is the norm, reflecting a "hegemonic notion of athleticism as a masculine trait" . The normative presumption of male dominance in sport and the physical advantages of men substantiate sex-segregation in the sport system, which was theoretically established to provide equal opportunities and maintain fair competitions . The sport system is based on the sex binary and assumes that athletes can be unambiguously separated into the rigid categories of biological sex assigned at birth. To participate in competitive sports, individuals have to "align themselves as female or male and join the corresponding team, " which is an unconscious decision for the majority of athletes . Thus, transgender individuals are strongly challenged by having to align to the sex binary and adhere to sports policies that mostly require a match of biological sex and gender identity . One of these challenges is because adaption, the movement from assigned sex at birth to gender identity, happens at different levels and stages, ranging from personal and social recognition to hormonal/medical treatment to sex reassignment . As sex-segregation is based on the assumption of male bodies' superior physicality, transgender athletes are perceived as challenging gender boundaries and contradicting the level playing field in sports . In particular, female transgender athletes, who transitioned from male sex assigned at birth to female gender identity, are accused of having unfair advantages over so-called natural women. Griffin and Semerjian question this and emphasize that gaining competitive advantages over others is a fundamental feature of sports, and advantages can occur in different forms, among which physicality is only one of many advantages . --- THEORETICAL FRAMEWORK ON EXCLUSION, DISCRIMINATION, AND MINORITY STRESS Social theorists have developed different perspectives and discourses on social inclusion and exclusion, drawing on social inequality or systems theory . Based on Weber's category of social closure, the inequality discourse deals with social exclusion as missed opportunities for social participation and an intensification of social inequality . Following Durkheim's work, social inclusion is based on the achievement of social cohesion and organic solidarity, whereas Marshall sketches social inclusion as linked to the granting of social rights, among others, that provide equal access to institutions and an appropriate standard of living . Access to and participation in organized sport contexts can be understood as a social right, that should be granted to every individual independent of social status, ethnicity, sexual orientation, or gender identity. Apart from that, sports and sport engagement are seen as a means for social inclusion or at least as able to contribute to inclusive processes. According to Bailey , sports are theoretically assumed to contribute to inclusion on different dimensions: bringing people together from different backgrounds ; enabling people to share a sense of belonging ; providing opportunities to develop competencies and capabilities ; and increasing relevant social networks and community cohesion . However, just as the existence of these dimensions can lead to inclusion, their absence can foster exclusionary processes and discriminatory experiences in sports. For understanding discrimination based on gender identity, Cunningham's multilevel model appears to be a useful framework. It is embedded in sports and considers societal, organizational, and individual factors on the macro, meso, and micro levels for explaining "the attitudes toward and experiences and behaviors of sexual minorities" . The levels are closely interlaced and influence each other in either direction. On the macro level, cultural norms, and institutionalized practices in society-in which organizations are embeddedinfluence the situation for transgender persons in sports. The binary gender order and the male hegemony which are reproduced in sports organizations and the rigid sex-segregated competitive structures constitute barriers for transgender participation in organized sports. Transgender athletes are seen "as outsiders or others, because their gender identity did not match the institutionalized ways that sport has been traditionally organized" . Meso-level factors operate at an organizational level and impact on exclusion and discrimination via leader behavior, organizational culture, and group spirit. The levels of diversity surrounding activities are strongly promoted by the support of organizations' leaders and the allocation of time, resources, and money. Organizational culture reflects a pattern of values, assumptions, and beliefs that have developed over time and are widely shared by the members and serves as an orientation for appropriate behavior . Organizational cultures that assume transgender athletes challenge the superiority of male physicality, distinct gender binaries, and the level playing field impede or complicate inclusion. On the other hand, support within groups on an emotional, structural, and instrumental level fosters the establishment of safe spaces and thus positively impacts on the well-being of transgender athletes . On the micro level, transgender individuals' discrimination in sport can be affected by the personal traits of other participants, bystanders, and spectators. Prejudices against transgender athletes rest upon the widespread assumption of sex binary and heteronormativity and vary between different gender identities, according to how athletes fit into the maledominated sporting world and the implied physical advantages of mainly transgender females . Different perceptions of transmen and transwomen in sports are associated with evidence that transgressions of gender norms and gender non-conforming expressions are less socially accepted for and amongst men than women and thus are more likely to provoke discrimination and self-exclusion . Apart from this, different individual identities, such as gender, racial, cultural, or athletic identity have to be considered when analyzing negative attitudes and behaviors toward transgender individuals. Generally, men articulate more prejudices against sexual minorities than do women in order to ensure male privilege and construct masculinity in heteronormative structures , although a current study disproves differences in trans prejudices between men and women . Apart from Cunningham's multilevel model, Meyer's minority stress model is valuable for determining the effects of transnegativity. Meyer's model was originally used to explain the high prevalence of mental disorders amongst lesbian women, gay men, and bisexual individuals. Nevertheless, as transgender represent a minority group that is marginalized in dominant cultures and social structures, the model can also be applied for the present target group. In Meyer's theoretical framework, minority stress is related to environmental circumstances, status and identity of the minority group, different stressors, and social support. Minority stress culminates in stigma, prejudice, and discrimination and can be traced back to distal and proximal stressors. Whilst distal stressors reflect external events and conditions, in which prejudice, discrimination, and violence occur, proximal stressors are subjective "perceptions of the self as a stigmatized and devalued minority" and incorporate expectations of rejection, internalization of transnegativity, and concealment of one's gender identity. --- TRANSGENDER ATHLETES' EXPERIENCES Having laid the theoretical framework for the current study, the quantitative and qualitative research corpus on transgender experiences in sports will be outlined in the following section. Quantitative research on sports participation, along with the situations and experiences of transgender athletes in different sports and settings, is rare and sometimes inconsistent. Regarding the general sports participation of transgender individuals, Kulick et al. found fairly equal sports frequencies between transgender and cisgender students in USA, but a mediated, negative effect for sports participation of transgender individuals through feelings of unsafety in facilities . In contrast, lower participation rates in sports and physical activities were found for transgender people compared to cisgender people in USA and compared to lesbian women, gay men, and bisexual persons in Canada . Personal experiences of transgender athletes in sports, as well as sports-related social and psychological outcomes, are important to determine inclusive or exclusive processes in sports and operate as moderators for participation . In Scotland, eight out of 10 transgender individuals agree that homo-/transnegativity is a major barrier for sports participation and that homo-/transnegativity is a problem in sport . Furthermore, there is empirical evidence that most transgender individuals have witnessed or experienced transnegative episodes in sports . The numbers are alarming. In Canada, 85% of transgender individuals personally had negative experiences in sports contexts due to their gender identity and in Scotland, 80% of transgender individuals had personally witnessed and/or experienced homo-and transnegativity in sport . Referring to all LGBT athletes of the Scottish sample, the share who had personally experienced or witnessed homo-or transnegativity in sport is considerably smaller . In a large Anglo-American study , 82% of the LGB surveyed reported that they have witnessed or experienced homonegative episodes in different sport contexts . In the overall OUTSPORT data about 12% of the European LGBTI+ respondents reported personal negative experiences in the past 12 months, with a significantly higher share of non-cisgender people and a slightly increased prevalence in higher performance levels, but independent from the type of sport . The predominant form of experiencing transnegativity is verbal, as almost all transgender individuals experience verbal abuse, insults, disparagement, and offensive remarks in sports . Referring to homonegativity, Symons found that verbal abuse toward LGBT athletes most frequently comes from other participants and only a small amount comes from coaches, spectators, or officials. Apart from verbal discrimination, 16% of transgender athletes reported having experienced or witnessed physical abuse and 7% reported other discriminatory experiences related to their gender identity, such as sexual assault . After being asked what would help fight discrimination in sport, Scottish transgender respondents mentioned public campaigns, diversity training for different actors in the sport system, and role models . Transgender athletes receive less social support for their physical activities compared to cisgender athletes and share lower scores for several psychological dimensions, such as the perception of their physical self and self-efficacy for exercise behavior . Both social and psychological outcomes are stressors and risk-factors for physical and mental health problems or depression . To draw a comprehensive picture, transgender athletes' narratives about barriers and facilitators for sports participation, as captured in qualitative approaches, will now be discussed. Mainstream and organized sports activities are mostly experienced as unsafe spaces by transgender individuals, in which they face several distal and proximal stressors . There is strong academic agreement that the changing room is one of the most challenging barriers: feelings of shame, body incongruence, body dissatisfaction, and fears of others' reactions-all resembling major internal proximal stressors-generate distal stressors, such as abjectification and stigma, negotiations of gender conformity, and discriminatory behavior . The locker room is perceived as the most challenging situation, "entrenched in cisgenderism and heteronormativity" , but the entanglement of proximal and distal stressors generally leads to feelings of fear in public sporting spaces, which often impede sports engagement . Moreover, alienating negative experiences concerning physical education in school are major barriers for further sports participation of transgender individuals . Also, clothing regulations and norms, participation in sex-segregated team sports, and a lack of acceptance in teams pose further external barriers to transgender sports participation . Further exclusionary processes toward transgender athletes, discovered by Hargie et al. , provide supporting evidence for Bailey's dimensions of social exclusion. In regard to organizational sports settings, Hargie et al. found that hostility from facility staff and members of the public emerged as relational stressors; high social, and economic distances led to a lack of social networks; social capital acted as spatial stressors; missed opportunities for gaining knowledge, competencies, and capabilities, as well as refraining from specific sports activities, occurred as functional stressors; and internalized transnegative attitudes and reduced self-confidence arose on the power dimension. Despite the barriers outlined, transgender perspectives stress the positive consequences of sports participation. Being physically active and doing sports contribute to body awareness and satisfaction, gender identification, and recognition and accentuation of body changes . Positive effects from sports on mental well-being and social involvement in welcoming, safe, and comfortable sports settings strengthen transgender individuals' abilities to cope with challenges . Several studies have identified safe spaces in sport, in which transgender athletes were accepted, socially supported by the team, and enabled to compete without rigid restrictions . But as these conditions are rather rare in mainstream sports settings, establishing trans-only sporting environments might externally facilitate transgender sports participation . Reflecting on the findings regarding barriers and facilitators for transgender sports participation, it should be mentioned that experiences are diverse, individual, and often dependent on settings, sports, social environment, or personal traits . Nevertheless, most transgender athletes are confronted with prejudices and ignorance in sports and personally experience discrimination, harassment, and scarce access to safe, comfortable, and inclusive sports environments . Exclusionary mechanisms are based on the supposed incongruence of transgender individuals with the dominant assumption of sex binary and the sex-segregation in organized sports settings and culminate in the "overall effects of being denied the social, health and well-being aspects of sport" . Elling-Machartzki emphasizes that traditional sports clubs appear to be particularly challenging sports settings due to the changing room situation and discriminatory policies for engaging in the competitive system. --- ORGANIZATIONAL POLICIES ON TRANSGENDER ATHLETES' INCLUSION clubs are prototypes of self-purpose organizations with the main function to serve the interests of their members and trigger them to engage in specific activities . Apart from that, in the sense of Ferdinand Tönnies sport clubs can promote communitization in society as well as social change and are perceived as drivers for social inclusion. This has become evident in manifold sport for all campaigns or recent diversity management approaches in organized sports . However, as sports clubs have a considerably high "resistance to change" , inclusion of underrepresented groups is a tremendous challenge. Especially if these groups challenge the logic inherent in the sport system, such as sex binary or the level playing field. Therefore, barriers for transgender sports participation are not only situated on personal and interactional levels but on institutional and policy levels . In their systematic review of transgender persons' experiences of sports and organizational policies regarding transgender inclusion, Jones et al. emphasize that "the requirements that transgender competitive sport policies place on competitors were instrumental in transgender athletes' negative experiences, " even in recreational sports settings. To deal with transgender athletes in competitive sports, Griffin outlined four potential policy forms: use of mandatory sex-verification testing to maintain the gender binary; case-by-case decisions on transgender inclusion using sex-verification testing; implementation of categories other than sex [i.e., Kane's idea of a "continuum, whereby sport is organized based on ability and athletes compete in skill level groupings, regardless of gender" ]; and expansion of gender categories to include transgender athletes. Many current sports organizations' policies belong to Griffin's second category and are perceived as discriminatory for transgender athletes; more inclusive policies are mainly represented in the fourth category. The International Olympic Committee policy 2004, known as the Stockholm Consensus, was one of the first attempts to regulate transgender participation and has been fully or partly adopted by sports organizations worldwide . This policy requires of athletes a genderconfirming surgery, hormone treatment, and living in the gender identity, each for at least 2 years . Since hormones are supposed to have strong effects on physicality, some organizations' policies require different degrees of hormone treatment and different threshold levels besides the legal recognition of gender identity . Criticism of these rigid transgender policies centers on the exclusion of transgender athletes who are not undergoing medical treatment; who are just in the process of transitioning; whose gender identity lies outside the gender binary; and on the discrimination toward transgender males who have allegedly no physical advantages . Taking some of the reservations and criticisms into account, in 2015 the IOC implemented different policies for male and female transgender individuals. For female transgender athletes, the requirement demands 4 years of declared gender identity and a threshold for testosterone levels, whereas male transgender athletes can compete without any restrictions . The National College Athletics Association has already implemented such policies in 2011 and, beyond that, allows transgender athletes to stay on their teams, as long as they do not start medical, hormonal transformation . These policies were also implemented by the National Women's Hockey League for elite athletes. More inclusive policies only require legal recognition of gender identity or allow participation simply on the basis of gender identity, regardless of legal recognition, medical treatment, and anatomy . Travers and Deri articulate two barriers for inclusive transgender policies: first, the "powerful culture of sex binary logic in the organization of sport, " built on the assumed relevance of testosterone; and second, obstacles to "re-negotiate sexed boundaries, " related to the dissent amongst transgender athletes about appropriate inclusive policies. Jones et al. question the inclusiveness of the reviewed policies, concluding that the majority of the applied policies in different organized sports institutions discriminate against transgender individuals and most notably transgender males, mainly because no empirical evidence exists on the athletic advantages of female and male transgender athletes. This is empirically supported by Harper , who identified times for middle distance runs for transgender females that comparably change according to the bodily status of adaption. From a philosophical point of view, Gleaves and Lehrbach call for society to focus on "narrativity rather than physiological equivalency, " so as to include transgender athletes, instead of perceiving sport solely as a comparison of physical skills and competencies. --- METHODOLOGY As the theoretical framework and research corpus have shown, transgender athletes are a highly controversial topic, encompassing inherent sports values such as fairness, in-/exclusion, discrimination, and equal opportunities. Thus, inclusion of transgender athletes in organized sports often fails, which leads to discriminatory experiences, humiliation, exclusion, and stress. This paper has three objectives: to analyze the LGBT+ athletes' assessment of transnegativity in organized sports settings in Europe; to scrutinize negative experiences with regard to gender identity, performance level, and LGBT+ reference in organized sports settings in Europe; and to discuss inclusive strategies, as well as barriers and facilitators, for implementing diversity in sports organizations in Europe. The first and the second objective are explored through a quantitative online survey in all European Union countries, and the third objective is explored through a qualitative interview study in five European countries . Both studies were conducted as part of the European joint research project OUTSPORT, which was developed by the German Sport University Cologne in cooperation with the Italian Association for Culture and Sport , LEAP Sports Scotland , the Vienna Institute for International Dialogue and Cooperation , and the Organization for Fresh Ideas, Hungary . In contrast to further publications already issued within the framework of this project , the current paper focuses on: subsample of LGBT+ athletes within organized sports contexts, systematic differentiation between cisgender and non-cisgender athletes, and integration of qualitative interviews with representatives of sports organizations. --- Athletes' Experiences: Quantitative Online Survey --- Sample The target population of the OUTSPORT survey was defined as anyone who identifies as lesbian, gay, bisexual, transgender, and/or intersex; currently lives in one of the 28 member states of the European Union; and is at least 16 years old. Being currently active in sports was not set as a mandatory criterion. Due to stigmatization, discrimination, and the lack of knowledge about socio-structural parameters-particularly on a European level-LGBT+ populations can be characterized as hard-toreach, hidden, and vulnerable populations . In order to draw a systematic sample of this population, a combination of sampling strategies was applied. Primarily, various LGBT+ organizations in each of the 28 EU member states were contacted and asked for contact details from other LGBT+ organizations and umbrella sports organizations in their countries. Each of these organizations was asked for support and supplied with standardized packages of promotional material to advertise the web-based survey via their channels . International LGBT+ organizations such as the European Gay and Lesbian Sport Federation and the International Lesbian, Gay, Bisexual, Trans and Intersex Association also helped to promote the survey. To increase the reach, a snowball recruiting technique among participants was applied. Social media targeting was used in an adaptive process to balance out lower participation rates in specific countries. The survey was promoted in a neutral way. To avoid a negativity bias, words and expressions such as "negative experiences, " "discrimination, " or "harassment" were not used. The online survey was accessible in German, English, Italian, Hungarian, and French. All translations were conducted by ETC Europe, a professional translation agency. To improve validity and reliability, emphasis was placed on precise wording in each language. Sensitive wordings were thoroughly double-checked by native speakers with an LGBT+ background who also worked for the OUTSPORT project. Anonymity and confidentiality for the participants were secured, and the General Data Protection Regulation of the EU was applied. The research design was approved by the ethics committee of the German Sport University Cologne. The online survey was accessible between March and August 2018. The total sample contains 5,524 valid cases. The share of each country's respondents from the total sample roughly corresponds to the share of each country's inhabitants from the total EU population . Within the scope of this paper, the focus is placed on a subsample of active respondents who practice their sport in an organized sports context . This could either be organized sports clubs , forprofit organizations such as fitness centers , or other organizations . Participants from the chosen subsample are between 16 and 75 years old . More than half have a college or university degree, and more than a third have completed upper or post-secondary education. Almost two thirds live in urban areas with over 100 k inhabitants , one out of five lives in towns or smaller villages , and 17% live in rural areas . The assigned sex at birth was female in 60% of cases and male in 40% of cases. --- Measures Based on the relevant research corpus the questionnaire for the quantitative online survey has been developed by the authors in cooperation with experts from the OUTSPORT project team. To ensure validity and rigor, the questions have been further discussed with scientific experts inside and outside academia and a pre-test has been performed leading to further adjustments. The following section shows precisely how the variables used in this paper have been measured and operationalized. --- Gender identity A two-question gender status measure was used to retrieve data about the gender identity of respondents . In a first step, respondents were asked to report their sex assigned at birth, with the options3 "male" and "female, " before respondents were asked about their current gender identity: "female"; "male"; "transgender"; and "I do not identify myself as male, female or transgender." Cross-analysis of these two questions resulted in six distinct gender-identity categories: female cisgender, male cisgender, female transgender, male transgender, non-binary transgender, and non-identifying people . Within the scope of this paper, the main distinction made will be between cisgender and non-cisgender respondents. Non-cisgender respondents will be further subdivided into three groups: female transgender; male transgender; and non-binary respondents . --- Sports setting In addition to the organizational context of a sport , two more dimensions are used to describe the sports setting: performance level and LGBT+ reference. Performance level was assessed by a single choice question , offering the response options "recreational"; "competitive"; and "high performance." The LGBT+ reference of a sports activity was assessed by asking respondents to characterize their sports context as either "mainstream/not specified"; "explicitly LGBT+ friendly"; or "LGBT+ specific." --- General assessment of transnegativity in sports To assess the perception of transnegativity in sports on a general level, respondents were asked if they think that there is a problem with transnegativity in sports. Response options were provided on a five-point scale from 1 to 5 and a no-choice option . Transnegativity was defined in the questionnaire as "prejudice or discrimination on grounds of transgender identity." --- Helpful strategies and awareness of contact points Respondents were asked what they would consider appropriate measures to tackle discrimination or harassment in sports due to sexual orientation and/or gender identity. To find out how aware people are of whom to turn to in cases of homo-/transnegative incidents in their sport, respondents were asked whether they know any support contact points they can get in touch with if such an incident happens. --- Homo-/Transnegative incidents Homo-/transnegative incidents were assessed in a two-step approach. First, respondents were asked whether or not they had personal negative experiences in their sport during the previous 12 months: "Looking back at the last 12 months, did you personally have any negative experiences in this specific sports context as a result of your sexual orientation and/or gender identity?" The question was formulated such that the connection to a specific sports activity and its context, the causal attribution to the respondent's own sexual orientation and/or gender identity, and the topicality were emphasized. In the second step, respondents were asked how often they had personally experienced several forms of homo-/transnegative incidents in that time span: verbal insults and slurs ; verbal threats ; physical crossing of lines ; physical violence other . Respondents were provided with a five-point scale response option for each form. Besides the forms and frequencies, respondents were asked about the perpetrators. For every form of homo-/transnegative incidents that occurred, respondents were asked, "Who said or did this?, " offering seven different perpetrators via a multiple-choice option. --- Organizational Strategies: Qualitative Interview Study The qualitative study consists of 15 expert interviews with relevant stakeholders from sport systems in the OUTSPORT project partner countries Germany, Scotland, Italy, Austria, and Hungary. The interviewees represent responsible actors from umbrella sports organizations of the voluntary sector at national or regional levels , specific sports federations , and public sports organizations focusing on specific target groups . The semi-structured, guided interviews focused on the broader context of LGBT+ inclusion in sport and organizational strategies for diversity and inclusion. In order to identify facilitators for implementing diversity strategies and inclusive policies for LGBT+ minorities in sport, the sampling aimed at including good practice organizations. The qualitative data collection was carried out between September 2018 and April 2019. The Scottish, Italian, and Hungarian interviews were conducted and transcribed by the respective project partner; one Austrian interview and the German interviews were conducted by the authors. All interviews were coded by the authors using qualitative, category-led text analysis . Deductively and inductively developed categories structured the coding process and the interpretation of the findings; both approaches were critically discussed with colleagues to ensure trustworthiness of the data. For the current purpose, findings on barriers and facilitators for implementing inclusive diversity strategies in sports organizations and the parts on transgender issues are focused on. Due to methodological issues in conducting interviews and different stages concerning problem awareness and inclusiveness for LGBT+ athletes in the five project partner countries, the results on transgender issues focus mainly on interviews from Germany and Scotland. --- EMPIRICAL RESULTS Before the results are going to be presented in detail, the sports participation of the respondents will be briefly described. For findings on the general sports participation of different gender-identity groups, the overall OUTSPORT sample was used. It was found that cisgender and noncisgender respondents differ slightly in their general sports participation over the last 12 months. Namely, transgender males have the highest share of active respondents, followed by both cisgender males and females , followed by female and non-binary non-cisgender respondents . In the current sample of active respondents in organized sport , six out of 10 respondents practice their main sport activity in organized sports clubs, 27.3% practice in forprofit organizations, and 11.7% practice in other organizations such as university or company sporting groups. Cisgender individuals participate more often in organized sports clubs, while non-cisgender individuals engage more often in activities held by other organizations = 12.99, p < 0.01, V = 0.075). The performance level is predominantly recreational , while competitive and high-performance levels are scarcer. Eight out of 10 respondents practice their sport in mainstream or LGBT+ unrelated settings and 17.4% practice in LGBT+ related settings-either explicitly LGBT+ friendly or LGBT+ specific . With regard to the performance level, no significant differences were found between cisgender and non-cisgender respondents = 2.67, n.s.). The share of cisgender respondents in LGBT+ specific settings is slightly higher compared to non-cisgender respondents = 10.85, p < 0.01, V = 0.069). --- Transnegativity in Sport Transnegativity is perceived to be a "big problem" by the majority of respondents. On the 5-point scale , 88.9% of respondents chose either category four or five , causing a strongly left-skewed distribution 4 . Cisgender and noncisgender respondents show similar mean values, with no significant differences in mean ranks . Within the group of non-cisgender athletes, respondents with a non-binary gender identity perceive transnegativity to be a bigger problem than do female and male transgender athletes. Accordingly, a Kruskal-Wallis test shows significant differences of means = 9.47, p < 0.01), with mean rank scores of 144.11 for transgender females , 133.3 for transgender males and 166.99 for non-binary persons . Post-hoc tests of pairwise comparisons show significant group differences between male transgender and non-binary respondents , with a medium effect size. --- Personal Negative Experiences Of all active LGBT+ respondents in organized sports settings, 11.8% have had at least one negative personal experience that was associated with their sexual orientation or gender identity in their main sport in the last 12 months . The differences between cisgender and non-cisgender respondents are remarkable. The share of non-cisgender athletes who have become victims of homo-/transnegative incidents is three times higher compared to cisgender athletes = 118.22, p < 0.001, V = 0.228). This effect consistently occurs in all three performance levels and all LGBT+ reference types with weak to moderate effect sizes . Irrespective of respondents' gender identity, higher rates of 4 From all 2,282 cases, 7.5% chose the option "don't know, " and 0.1% did not respond. homo-/transnegative incidents are weakly associated with higher performance levels = 9.66, p < 0.01, V = 0.068) and more likely to occur in mainstream than in LGBT+ related settings = 6.67, p < 0.05, V = 0.053). Among non-cisgender respondents, differences with regard to the performance level = 4.01, n.s.) and LGBT+ reference = 1.64, n.s.) are not statistically significant. Among those who have personally experienced homo-/transnegative incidents in the last 12 months, verbal insults and structural discrimination such as unequal opportunities, unfair treatment, or exclusion were the most common forms . Moreover, verbal threats and intimidations occurred in 39.4% of the cases, and harassment via social media, messengers, or webpages occurred in 35.3% of cases. Physical types of homo-/transnegative incidents happened to 31.6% of respondents as lighter forms of crossing the lines and to 15.6% as severe forms of physical violence . For reasons of clarity, the 5-point scaled data was dichotomized first. Percentages are used to indicate the share of respondents who experienced specific types of incidents. Table 3 shows group comparisons regarding gender identity, performance levels, and LGBT+ reference of the sport. Results indicate that verbal insults are more frequently experienced among cisgender respondents = 13.10, p < 0.001, V = 0.221) and in higher performance levels = 6.85, p < 0.05, V = 0.160). Non-cisgender respondents more commonly experience structural discrimination = 3.53, p < 0.05, V = 0.114) and other negative incidents = 7.08, p < 0.01, V = 0.162), which are typically associated with misgendering . Database: Active sports participants in organized sports who have had personal negative experiences associated with their sexual orientation or gender identity in the last 12 months and who named at least one perpetrator . Test statistics and effect sizes are written in italic. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. pronouns and appropriate naming) or being looked at in a derogatory way. No statistically significant differences were found between mainstream and LGBT+ related sport settings, although a higher rate of physical violence is observed in LGBT+ related settings = 3.44, p = 0.064; V = 0.113). For each of the seven forms of homo-/transnegative incidents, respondents were asked about the perpetrators , with multiple-choice options. Table 4 shows the percentages for all seven forms combined . Members of respondents' own teams and other sport participants are the most frequently mentioned perpetrators, irrespective of the specific types of incidents. About a third of respondents with homo-/transnegative experiences identified members of opposition teams , coaches , and spectators as perpetrators, whereas about one out of six indicated that at least one negative incident was caused by other officials . Cisgender and non-cisgender respondents referred to roughly the same types of perpetrators except for coaches, who are more commonly indicated by non-cisgender respondents = 6.09, p < 0.05, V = 0.115). Highly significant differences can be found with regard to the performance levels. Perpetrators from respondents' own teams = 17.33, p < 0.001, V = 0.261) or opponent teams = 16.79, p < 0.001, V = 0.257) were more frequently indicated by athletes in higher performance and competitive levels, respectively. In contrast, other sports participants were more frequently referred to as perpetrators in recreational settings = 15.44, p < 0.001, V = 0.246). Spectators = 4.31, p < 0.051, V = 0.130) and other officials = 4.46, p < 0.05, V = 0.133) were mentioned significantly more often by respondents in LGBT+ related sports settings, whereas team members as perpetrators are more widespread in mainstream sports settings = 5.39, p < 0.05, V = 0.146). --- Tackling of Transnegativity Respondents who have had homo-/transnegative experiences at some point in their sport were asked whether they were aware of contact points for support . One out of three respondents indicated being aware of contact points in non-governmental organizations outside of the organized sport. One out of five respondents was aware of contact points in local sports organizations, and one out of seven was aware of contact points in umbrella sports organizations on a regional or national level . There are no statistically significant differences between cisgender and transgender respondents. The most frequently named measures to tackle discrimination and harassment on the grounds of sexual orientation and gender identity are to encourage more sports stars to come out and to run high-profile anti-homophobia/transnegativity campaigns, followed by diversity training and more inclusive policies 5 . More than two thirds of respondents considered tougher sanctions as appropriate measures, whereas almost 9% referred to other measures. The coming-out of sport stars is perceived to be helpful by a higher share of cisgender respondents = 7.32, p < 0.01, V = 0.100). Non-cisgender respondents consider inclusive policies = 22.35, p < 0.001, V = 0.175) and diversity training = 5.67, p < 0.05, V = 0.088) more important and also refer to other strategies twice as often as cisgender respondents = 9.76, p < 0.01, V = 0.115). Better education about gender diversity, genderneutral changing rooms, and enhancement of social acceptance were most frequently mentioned. --- Transgender Issues in Sports Organizations The issue of transgender inclusion in sports is an ongoing and complex topic for organizations in Germany and Scotland, 5 Due to an unintended permeability in the filter guidance, respondents who had not had negative experiences in the last 12 months also answered this question. As perceptions about helpful actions to tackle discrimination are probably independent from immediate experiences of discrimination and for the sake of a higher number of cases, the table includes all active sports participants in organized sports who made a valid statement in this question . while in the other countries, the organizational approach to the issue of transgender athletes appears relatively vague and sometimes unpopular. In relation to the engagement of several sports federations with transgender issues, the interviewee from the Austrian Center for Gender Competence pointed out that "it's rather an unpopular topic that they have to deal with, not want to deal with" . Thus, the German football associations interviewed, as well as organizations in Scotland, reported that transgender inclusion into competitive systems is mainly ruled by case-by-case decisions and individually by the responsible organization. Some positive examples of regulations were mentioned in the interviews, as the following quote from sportscotland indicates: Gymnastics made a regulation change because the rules stated that females had to wear leotards, a certain outfit, and there was a young transwoman who didn't feel comfortable with that. And she actually was getting deducted points because she wanted to wear different clothing. Now again that's something very simple but was creating a massive issue for her competition . The German and Scottish organizations highlight the need for implementing general inclusive policies on a broader level, and in most of the organizations, there are already efforts to develop generally binding regulations and guidelines. The ruling by the Federal Constitutional Court in Germany on the option of a third gender has increased the urgency for inclusive policies beyond the sex binary, as the German Olympic Sport Federation emphasized: "especially in the context of grassroots sport, national organizations are invited to formulate appropriate recommendations" . The organizations in Germany and Scotland perceive both cooperation within the sport system and low-threshold offers as beneficial. Workshops, events, and other opportunities enable different actors in the sport system to get in touch with each other and discuss various ways for transgender inclusion into competitive sport structures. Mutual learning and inspiring each other is beneficial, as the Football Association Baden stressed: "the topic is not about competition or who does what better or something like that, but it's simply about mutual stimulation and motivating people to move forward with the topic" . --- Facilitators for Implementing Diversity Strategies Apart from specific issues regarding transgender athletes, the 15 interviews focused on the general inclusion of LGBT+ athletes in sport organizations. With this broader focus, the barriers and facilitators for implementing diversity strategies will be discussed with reference to Cunningham's multilevel model. On the macro level, the majority of sporting organizations demand that politicians and the public lead the way to LGBT+ inclusion, address the situation, and raise awareness about the barriers for the inclusion of LGBT+. These external factors were perceived as beneficial for implementing the topic of diversity in sports organizations. Many interviewees perceived hardly any rejection of the topic but rather perceived a lack of knowledge and awareness, as highlighted by sportscotland: "If you're looking at resistance, the only thing we tend to come across is lack of awareness and education" . The Italian Culture Sport Agency perceived some areas of resentment, but only when this agency started to deal with LGBT+ issues: "And even those cultural resistances that existed at the beginning , slowly absorbed this new situation and have accepted it, and today frankly LGBT issue is not a sector we put a "different" effort in, it is a sector like all the others" . In contrast, however, the Hungarian Olympic Committee expected rejection of their member organizations when addressing the issues of LGBT+ inclusion: "But our member clubs are independent; we have no legal ways to interfere. What we can change is the mentality. But if we start to deal with these issues, there will be problems" . Therefore, communication, education, and role models serve as important tools for raising awareness and increasing sensitivity to the issue of diversity and inclusion of sexual minorities in sports. The interviewee from the Italian Sport Union For All emphasized this: "The fact that there are athletes who come out is very important as well as the fact that there are more and more moments where various subjects and institutions address LGBT issues and try to question themselves on this matter" . Moreover, male hegemony that is reproduced on various levels in sports is perceived as a barrier for LGBT+ inclusion. Individuals who are not representing heteronormative and male stereotypes do not fit into sporting realities as the interviewee from the Austrian Center for Gender Competence stressed: "I think anything that runs counter to the typical male ideal is simply insecure in sport. Women, lesbians, gays, transgender people, girls in headscarves, people with disabilities" . On the meso level, cooperation with governing bodies, along with relevant stakeholders inside and outside the sport systems, served as an important facilitator for the integration of LGBT+ issues into sports clubs and organizations. Furthermore, a welcoming organizational culture and robust inclusive values were mentioned as instrumental for the inclusion of LGBT+ athletes: "We definitely live and breathe our values we try and theme things around them. So, our values are at the core of everything we do" . The organizations believe that both top-down and bottomup strategies are needed to implement LGBT+ inclusion. The interviewee from the State Sports Confederation of Saxony-Anhalt considers umbrella organizations as important role models for implementing the topic: "It is also important, of course, that the umbrella organizations take the appropriate steps, because then, I believe, it will be easier for the smaller associations, organizations, and clubs to support the issues, if the umbrella organizations put them into practice" . However, the implementation of LGBT+ inclusion solely via top-down strategies in sport might ignore "barriers of volunteering, lack of time, and lack of personnel" that most sports clubs experience, as the Austrian Sport Organization mentioned. The organizations emphasized the need for a distinct commitment of the organizations' leadership to the topic of diversity and inclusion, along with the willingness to act and to allocate respective resources for initiating activities and implementing strategies. Conversely, a lack of commitment and a lack of financial, staffing, and time resources are viewed as major barriers for implementing strategies, campaigns, or other measures. "I suppose the only barrier that we have is the size [of the team/organization] in terms of our resource" . Official regulations, guidelines, and charters on the appreciation of diversity, diversity management, and prevention of violence are supposed to be beneficial. The Austrian Sport Organization stressed the importance of clearly naming the dimensions of diversity, such as sexual orientation, in the official guidelines and regulations. Furthermore, the implementation of the topic on diversity and inclusion of sexual minorities in the educational structures of sport-for coaches, officials, referees, employees, board members and so on-is seen in a positive light. To justify to higher management levels and potential funding bodies the need for inclusive strategies and measures and to draw a comprehensive picture of the situation of LGBT+ in sports, many organizations stress the relevance of empirical facts and figures. On the micro level, the organizations are highly aware of the significance of using sensitive and inclusive language in the context of discrimination in sports. People using homoand transnegative language are often not aware that it might be perceived as discriminatory by LGBT+ individuals. Instead, the word "gay" is commonly used in sport contexts to describe something in a negative way, but not with the intention to harm LGBT+ individuals. Raising awareness for the discriminatory effect of homo-and transnegative language in sports is an important concern of the organizations. "And I think particularly with LGBTI [inclusion], you want to have language that is inclusive because it does, it seems to change quite a lot" . Apart from that, it became obvious that the organizational engagement with the topic of diversity and inclusion often depends on highly dedicated individuals within the organizations , who are committed to advance the issue, implement initiatives and measures, involve relevant stakeholders, and cooperate with them. --- DISCUSSION Against the background of growing empirical research on LGBT+ people in sports from Anglophone countries, this paper enhances the research body on both the experiences of LGBT+ people in organized European sports in general and the various challenges with regard to gender identity in particular. Aside from that, the combination of quantitative findings on the experiences of LGBT+ athletes and qualitative findings on organizational strategies for diversity and inclusion represents a major contribution of this study. The current findings indicate fairly equal sports participation rates of cisgender and non-cisgender individuals in Europe . Transgender males, being the most active group, reflect structural conditions of the sport system. Compared to transgender females, transgender males challenge the level playing field less strongly because they are not accused of unfair physical advantages. Similarly, transgender males challenge the level playing field less than non-binary athletes because the former fit into the sex binary . These perceptions culminate in fewer barriers for and fewer prejudices against transgender males in the domain of sport, which in turn increase transgender males' participation . LGBT+ athletes in organized sports contexts in Europe perceive transnegativity as a major problem in sports and even as a bigger problem than the Scottish respondents perceived in the studies by Smith and colleagues some years ago . National differences in the prevalence of transnegativity amongst European countries, as indicated in the overall OUTSPORT results , as well as negative changes over time might have contributed to this difference. Within the group of non-cisgender respondents, those who reject the binary assess the problem to be the biggest, reflecting the sex-segregation in sports, which is an even higher barrier for non-binary athletes than for male or female transgender athletes. Despite the generally high awareness of transnegativity, the share of respondents with personal negative experiences is relatively low, compared to findings by Demers or Smith et al. . Different periods considered may partly account for the differences: while the current study looked at incidents in the last 12 months, Demers and Smith et al. focused on lifetime experiences. Apart from methodological issues, the spread of more inclusive sporting realities for LGBT+ individuals and self-exclusion processes based on proximal stressors serve as arguments for reduced transnegative episodes in sport. Elling-Machartzki and Hargie et al. stress that expectations of being harassed or discriminated in sport contexts and the "fear of the felt stigma" that is "related to transgenderism" led to self-exclusion and thus to LGBT+ sporting populations that are less vulnerable to harassment and discrimination. Within the group of LGBT+ athletes, non-cisgender athletes are identified as the most vulnerable group in organized sports in general and particularly with regard to structural discrimination . Non-cisgender athletes challenge the sex binary and sex-segregated sport systems and the alignment to either male or female teams that is required for participation in competitive structures . Even LGBT+ specific settings seem to be less safe for non-cisgender than for cisgender athletes , which emphasizes the importance of trans-only sporting environments as an external facilitator for sport participation of transgender individuals . Transgender athletes' negative experiences are reflected in calls, particularly from non-cisgender athletes, for diversity training and more inclusive policies. The disadvantages of non-cisgender athletes are also evident in the different status of transgender athletes' issues, lagging behind diversity strategies and inclusion of lesbian, gay and bisexual individuals, as the interviews indicated. But the will "to overcome resistance to change" is evident in many sports organizations, which stress the need for inclusive policies instead of the case-by-case decisions currently being taken. Concepts of hegemonic masculinity and heteronormativity have been identified as important determinants for LGBT+ discrimination. Verbal abuse, the most common form of discrimination is mainly anchored in expectations of heteronormative, male hegemony, which strongly devalues homosexual orientations . Moreover, the impact of performance level on discrimination experiences of LGBT+ athletes is closely related to the importance of hegemonic masculinity in competitive sport, which is more strongly anchored in higher performance levels than in recreational sports settings . Individuals who are embedded and socialized in sporting cultures identify more strongly as athletes, particularly if they are engaged in competitive sports, and therefore express more sexual prejudices than athletes with a lower athletic identity . The high prevalence of verbal abuse and severe physical incidents supports findings from Demers and Smith et al. , but the pervasiveness of incidents related to physically crossing lines is an alarming sign for the sport culture. Referring to gender identity, being looked at in a derogatory way, or being misgendered happen more often to transgender and gender nonconforming athletes . The interviewed organizations recognize these problems and emphasize the need to raise awareness and knowledge on the importance of a sensitive language in sports contexts and perceive the integration of LGBT+ into educational structures as an important measure. The need for this is strengthened by two aspects: first, the finding that coaches discriminate against non-cisgender athletes more often than cisgender athletes; and second, academic agreement on coaches playing a major role in raising awareness, establishing an inclusive climate, and addressing problems concerning correct pronouns and the changing room situation . Implementing the topic into educational structures and offering diversity training are not only perceived as important measures to tackle transnegativity by the organizations but also by the respondents themselves. Additionally, non-cisgender athletes express the need for more awareness and acceptance and the need for gender-neutral changing rooms, with changing rooms having been consistently identified as very unsafe spaces . --- CONCLUSION Before concluding thoughts will be presented, some limitations to the studies and future perspectives have to be mentioned. First, it should be noted that the generalizability of the study's results is limited, since the present sample consists of self-selected respondents from an international population with unknown socio-structural parameters. Second, the questionnaire was only available in five languages, leading to a considerable percentage of the European respondents who could not respond in their mother tongue. This might have resulted in problems in understanding the questions and therefore in biased results and/or a biased sample, accounting for the rather high education level of the respondents. Third, although the sample size is satisfactory, the subsamples for non-cisgender athletes with negative experiences in specific contexts is sometimes insufficient to draw reliable conclusions. Additional quantitative research on transgender and nonconforming athletes is needed to further examine some discriminatory experiences that have been indicated in the current study, such as LGBT+ settings as unsafe spaces for non-cisgender athletes . Fourth, the quantitative approach is unable to reflect the manifold, diverse, and different individual experiences of LGBT+ athletes in various sports contexts, settings, countries, and so on . But it is able to provide comparable insights into the experiences of LGBT+ athletes with regard to different dimensions such as gender identity, performance levels, or LGBTI+ reference of the sport setting. Thus, the necessary reduction of complexity in quantitative research highlights the need for further research and, preferably, mixed-method approaches. Fifth, the qualitative study is limited by some methodological issues in conducting the interviews, which complicated drawing differentiated and comprehensive pictures in all project partner countries. Moreover, due to the broader focus on LGBT+ inclusion and the rather elementary stages in non-cisgender inclusion of sports organizations in all countries, transgender policies could not have been adequately discussed. Integrating more countries in future research and providing scientifically based support for the inclusion of transgender individuals in sports organizations might be beneficial. The aim of the paper was to shed light on the often neglected inconsistencies between LGBT+ athletes in sports in Europe. Our findings strikingly support the necessity of disbanding the LGBT+ umbrella, as lesbian, gay, bisexual, transgender, and non-binary athletes are facing different challenges and different proximal and distal stressors in participating in organized sports contexts. Two realities seem to account for the different sporting realities of cisgender and non-cisgender athletes: first, transgender athletes thoroughly challenge people's assumptions of distinct gender binaries and heteronormativity, and second, they challenge the gender boundaries of the sex-segregated sports system. Transgender participation in sport "belies the myth of the level playing field and the myth of gender binary on which it rests" . Instead of discussing the implied unfair physical advantages of transgender athletes, Lucas-Carr and Krane call for a shift to discuss the "trans disadvantage" due to negative experiences and ongoing challenges in sports that complicate sports participation, practice routines, competition, and so on. Together with the positive outcomes of being physically active, this strengthens the need for establishing welcoming organizational cultures and inclusive structures enabling transgender athletes to participate in organized sports settings without the fear of being verbally insulted, misgendered, stigmatized, or discriminated against. These calls are being heard in sports organizations but are not yet sufficiently implemented as the interviews with the sport organizations indicate. The study emphasizes that change is needed on macro, meso, and micro levels : first and foremost, to raise awareness for transgender issues and implement inclusive policies that allow transgender athletes "to compete as they wish" ; secondly, to bring the coaches in to establish welcoming and safe sporting realities; and lastly, to provide gender-neutral changing/locker rooms in sports facilities and promote sensitive and inclusive language. --- DATA AVAILABILITY STATEMENT The datasets presented in this article are not readily available because data is part of a funded Project by EU Commission. --- ETHICS STATEMENT The studies involving human participants were reviewed and approved by Ethics commitee of the German Sport University Cologne. Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements. --- --- FUNDING As part of the OUTSPORT project the research was funded by the ERASMUS+ Programme. The European Commission support for the production of this publication does not constitute an endorsement of the contents which reflects the views only of the authors, and the Commission cannot held be responsible for any use which may be made of the information contained therein. --- Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

In relation to conceptualizing sports, beliefs about sex binary and male hegemony are dominant. To match these assumptions and provide level playing fields, sport systems are based on sex-segregation. Thus, people who do not fit into or reject fitting into sex categories are hindered from participating in sports, particularly organized sports. Studies on social exclusion of gender-identity minorities in sports mainly adopt a qualitative approach and focus on Anglophone countries. This research is the first to provide a comprehensive picture of the experiences of LGBT+ athletes in organized sports settings in Europe and is based on a quantitative online survey (n = 2,282). The current paper draws special attention to differences between cisgender and non-cisgender athletes (including transgender men, transgender women, non-binary, and non-identifying individuals). Besides athletes' experiences, organizational strategies of inclusion, derived from qualitative interviews with stakeholders from sport systems in five European countries (Germany, Scotland, Austria, Italy, and Hungary) are examined. Theoretically anchored in Cunningham's (2012) multilevel model for understanding the experiences of LGBT+ individuals and Meyer's (2003) minority stress model, the paper aims to (1) analyze the assessment of transnegativity and (2) examine negative experiences (prevalence, forms, perpetrators) of LGBT+ athletes from organized sport contexts in Europe; and (3) discuss inclusive strategies in sports organizations in Europe. Data reveal that transnegativity is perceived as a major problem in European sports, and non-cisgender athletes are the most vulnerable group, suffering particularly from structural discrimination. The implementation of inclusive strategies for non-cisgender athletes is perceived as a complex and essential task, but the sports organizations in the five countries differ substantially in terms of the status of implementation.

INTRODUCTION Cooperative in Indonesian and Malaysian economies have a very strategic position and role in growth and development economics potency. Ketaren states that having a cooperative role in macroeconomics is increase benefit social and economic for society and the environment, improve production, income, and welfare, and increase work opportunity. cooperatives as the third economic contributor after the public and private sector and their role as social entrepreneurs is emphasized in the 2018 National Entrepreneurship Framework by the Ministry of Entrepreneur Development . This also coincides with the national agenda through the National Entrepreneurship Policy 2030 which is a long-term strategy for the country in making Malaysia an entrepreneurial country nation) that is superior by the year 2030. Next based on data from the Ministry of Cooperatives and SMEs in BPS , cooperatives in Indonesia are experiencing an increase, where in 2008 the total cooperatives increased by 154 thousand units to 209 thousand units in 2014. Likewise, amount member cooperatives in 2008 amounted to 27.3 thousand members increased by 29% to 36.4 thousand members in 2014. Meanwhile, in 2019, the Malaysian cooperatives sector recorded revenue of 45.79 billion through six areas of success this main achievement is considered far from the DKN 2030 target set a higher revenue projection of RM60 billion in 2030. Focus literature about cooperatives in Indonesia and in Malaysia currently this is about the development of digital cooperatives, training as well as mentoring, and also providing good seminars on campus or at school. In Thing youth also became one thing to pay attention to, considering youth is the successor in the future come. The Youth Development Index is a tool measuring youth development in five basic domains that are education, health and welfare, field and opportunity work, participation and leadership, and gender and discrimination. Cooperative presents an opportunity potential for youth, to increase eye livelihood and income, and to provide a place for training for the youth in Skills organization, solidarity, cohesion social, and democratic processes. The research behind the article investigates cooperation could change standard youth life. Standard of living refers to the number of goods and services available that people bought with their money to have a Standard life this is interpreted as enough to finance necessary life like food, clothing, housing, recreation, education, and protection insurance. The standard idea could opposite of quality life, which takes into account not only standard material life but also factors subjective to others who donate for life someone, like entertainment, security, resources culture, life social, mental health, and others. Although the strategies for promoting youth cooperatives in the two countries are different, the fields of skills and develop the same person, the learning process, and its impact is identified in the third cooperative. Data shows that cooperatives provide space for extended learning ' for youth to shape their future personal them. Expanded properties from room study generated from provided network co-operatives, as well as from activities special in cooperatives and the formal and informal learning processes involved. Kindly specifically, this research attempted to find if can cooperative change the standard of living of youth. Sdrali social economy reflects a value system being different from that of the for-profit sector. The aim of social economy is initially to cover peoples' basic needs through an assessment of an equitable distribution of resources and allegiance. Α socialeconomy sector has a unique role in creating a strong, sustainable, prosperous andinclusive society. Cooperatives is one of social economy organization. Cooperatives are autonomous associations of people united voluntarily to meet economic, social, and cultural needs and aspirations through jointly owned and controlled enterprises . According to Kusnadi cooperative organizations exist in almost all industrialized countries and developing countries. At first, cooperatives grew in industrialized countries in western Europe, but then after colonialism in African, Asian, and South American countries, cooperatives also began to grow in colonial countries. After the colonial countries experienced independence, many countries used cooperatives as one of the government's tools in carrying out development. --- Youth and Coopratives The first cooperative principle was introduced in Rochdale, England in 1944. The principle this arranged as a guide for Rochdale at that time for Becomes guide in reaching the goal. However Thus, every country needs to adapt to conditions and circumstances cooperatively. In Indonesia, the cooperative principle consists of membership characteristics voluntary and open, management conducted in a manner democracy, division remainder results effort conducted in a manner fair comparable with magnitude service effort of each member, gift reply limited services to capital, and independence. More goes on, so you can develop so co-op is also necessarily doing education cooperation and cooperation between cooperatives. On cooperatives that violate principles, the Ministry of Cooperatives and SMEs provide nicknames as cooperative naughty. Cooperative mischievous in general, only activities for money laundering, investment illegal, and moneylender under the guise of cooperative . Article 33 Paragraph of Undang-Undang 1945 states that the Indonesian economy is structured as a joint venture based on the principle of kinship. In his explanation, it is stated, among other things, that the priority is given to the prosperity of the community, not the prosperity of individuals or companies, so what is in accordance with this is cooperatives. As for the first cooperative in Malaysia was established in 1907 and was introduced under the name "Syarikat Koperasi" to help farmers. This cooperative movement grew in the decade of the 1980s when the Ministry of National and Rural Development or KPNLB launched the "Era Baharu Koperasi ". This launch is an initiative and encouragement to make the cooperative movement in Malaysia more progressive, dynamic, and advanced . The development of the cooperative movement in Malaysia is not only undertaken by the government but also by researchers with the highest academic standards in Malaysia. The study conducted on cooperatives in Malaysia gives advantages to cooperative management to improve the management efficiency and performance of each cooperative. . Cooperative entrepreneurship plays a significant role in the effort of developing future entrepreneurs. This type of entrepreneurship increases the economic and social interests of its members and brings substantial advantages to the community and society. Due to capital sharing and low risks in starting a business, cooperative entrepreneurship is a prominent way to generate job opportunities and reduce unemployment without depending on wage employment, especially among youth. In cultivating the entrepreneurial spirit, the Malaysian Ministry of Education has introduced school cooperatives since 1968, which aim to nurture entrepreneurial experiences and orientations among youth. For statistical purposes, the United Nations defines 'youth' as persons aged between 15 and 24. However, this definition is not universal. As the experience of being youth can vary substantially across the world, between countries and regions, we consider 'youth' as a flexible category. As such, context is always an important guide in UNESCO's definition of youth. Ridzwan Youth is an important asset to the economic development of a country. This is because they have a strong vitality that is needed for economic development. In addition, at a young age, they also still have strength both mentally and physically Recently, the young generation especially youth constitutes a large segment, which is 43 percent of the total population in Malaysia. The involvement of youth in entrepreneurship has become a major discussion among policymakers, educators, and students throughout most developing countries. Accordingly, the youth had become one the important assets for many countries in sustaining the country's economy. Supandi mentioned, according to Badan Pusat Statistik it is stated that of the approximately 251 million Indonesian population currently, the productive age population dominates around 44.98% of the total current population of Indonesia. The proportion of the productive age population is expected to continue to increase until 2025, with this demographic bonus, the productive age population reaches 2/3 of the total population in Indonesia. From 2/3 of the productive age population, there is certainly the role of youth in it. This situation can certainly lead Indonesia to become a bigger and more advanced nation in various aspects. Based on that, the Malaysian government, in order to create interest in entrepreneurship among youth, provides initial capital to start a venture. There is also an initiative had been launched by the government to curb the unemployment rate among graduates because many people in Malaysia involvement in entrepreneurship at a young age where which is important for career development. Next whether in Indonesia or Malaysia, cooperatives moment this not yet fully become the choice of the main institution. this is seen from low participation residents to be member cooperative only 8.41%. this figure is still below the global average of 16.31 %. --- Standard of Living The three main, and sometimes conflicting, definitions of Standard of Living are: 1) the utility of life , 2) economic provision or 'opulence', which measures the amount and quality of commodities that the individual is free to use ; and 3) a type of freedom, ie to be free to do something, the capability to live well for a certain socio-economic class in a certain geographic area . Among these, the last one is the most complete and complex definition; however, in practice, governments and institutions tend to use the second approach since it is easier to obtain data for this one. Standard of living generally refers to wealth, comfort, material goods, and necessity of certain classes in certain areas-or more objective characteristics-whereas quality of life is more subjective and intangible, such as personal liberty or environmental quality. Characteristics that make up a good quality of life for one person may not necessarily be the same for someone else. According to Nur Baiti in her research entitled Theory of Quality and Human Living Standards, it states that the Standard of Living refers to the number of goods and services that people can buy with the money they have. clothing, housing, recreation, education, and insurance coverage. The quality and standard of human life play an important role in economic development. Equitable development in various sectors is the key to improving the quality of life. Without development in various sectors, the quality of life will not work as it should. It seems that no sector does not play a role. These sectors will have the best impact in terms of development. It is there that results will emerge that reinforce the quality of life indicators. --- METHODOLOGY This study uses a qualitative descriptive research method. When studying problems, research does not prove or reject hypotheses made prior to research but manages data and analyzes a non-numeric problem . This type of research satisfies the description of the data in the form of sentences that have deep meaning derived from the information and behavior observed. The data from this research are in the form of facts found by researchers in the field . In this study, the researcher tried to understand the subject from his own frame of mind, so that the most important thing was the participants' experiences, opinions, feelings and knowledge. Therefore, all perspectives are valuable for research. The research method according to Sugiyono is basically a scientific way to obtain data with specific purposes and uses. This study used several data collection techniques, including field studies or direct observation. Bungin mentions that an observation is made directly on the object being observed, in the sense that the observation does not use transparent media. This means that researchers directly see and observe what happens to the object of research. Field observations were made by visiting several cooperatives that were used in the research located in Subang and Selangor. In addition, data was also obtained through interviews with members as well as cooperative management. In this study, members who are also cooperative administrators aged between 15 to 25 years are used as the object of this research. The profiles of cooperatives used as research are as follows: The name Fajar Sidik was taken from a figure who played a major role in the founding of the cooperative, namely H. Fajar Sidik who was also the chairman of the first cooperative and was credited with bringing the cooperative to success. Currently registered with Legal Entity Number 3298/BH/PAD/KWK.10/VII/1997, dated 30 July 1997. Darul Ehsan Cooperative is a cooperative that stands on the campus of the Antarabangsa Islamic University of Selangor. This cooperative is engaged in trade and services, among the goods and services managed by the Darul Ehsan cooperative are retail shops, printing shops, van rentals, homestays, seminar room rentals, and camp equipment rental. The method research used in this study interviews. Data were obtained directly from the youth respondents' that work in cooperatives in Subang and Selangor. Respondent was chosen through observation initial, interaction, and discussion with the chairman cooperative about the suitability and availability of members, with attention special assigned to the type of position in a cooperative, duration of membership, and age of the member. Interview individual focused aim for obtaining experience study from youth member. To collect data, members are given a number of questions and then requested to tell a story about their history live them, and engage them in the cooperative. They are also asked to tell and identify various types of their learning and knowledge get through Becomes administrator at a time member co-op, how? The thing that happened and what can they do as a member cooperative. These youths were also questioned about the meaning of cooperative for they as well as hope they about cooperative in the future come. This article focuses on the narrative of youth from about involvement they are in a cooperative as well as the impact they had to get inside life them. Data is analyzed in a manner inductive to see Thing what comes up of youth working in cooperatives. Besides changes to standards of living, they were also questioned during the interview process about experience study what have they learned and related with Becomes part of cooperatives and what processes are they have involved inside. The data is also examined to identify an aspect of life that is valued by youth. Experience studying youth and how they use learning they also mapped on what they appreciate. --- RESULTS This section discusses the answers obtained in Interviews with administrators from cooperatives range ages 15 to 25 years that is, what is said by youth about standard life as well as what they learn through becoming a member cooperative and how they understand learning resources and process them and learning what is possible they for do it. In other words, the ability that has been created or improved through becoming cooperators and how the youth could use the learning to reach what they appreciate. The following are the results of interviews on matters relating to the standard of living of youth cooperative members: The data above is a summary of the results of interviews conducted with members as well as youth administrators in cooperatives which are used as research objects. From the results of the study, it can be concluded that the standard of living which consists of the utility of life, finance, and type of freedom is sufficient when youth work in the cooperatives, even some youths also show pride and enthusiasm for the development of cooperatives in the future. In this case, the cooperative has also become a place to change the standard of living of youth. In the utility of life, respondents convey about social security, welfare, and other guarantees that they get. They also mentioned the vision of the cooperative so that this did not make them worry about working in the cooperative, because with a vision, the cooperative would survive and develop. Furthermore, all employees show that the income they earn is sufficient and in accordance with minimum standards. Then the thing they like the most is when they can get positive things in the utility of life and finance is the type of freedom they get. The respondents mentioned that with their flexible time they got the opportunity to develop themselves, they also really liked it when their opinion was accepted and they also mentioned that they could get a very wide range of learning opportunities that they found in the cooperative where they worked. --- DISCUSSION Based on the results of interviews with cooperative managers in Subang and Selangor, it can be concluded that some of the skills they learned and acquired in cooperatives are as follows: Tabel 3. The data above is data obtained from the results of interviews conducted with cooperative youth managers. There are some skills or development they get either for themselves or for the cooperative as a business experience. This is not only felt by youth who are working in a cooperative for the first time but is also felt by youth who have previously worked. They even conveyed a comparison of the work they did at their previous workplace and at the cooperative where they currently work. Through this, the author realizes that in addition to an increased standard of living, the learning opportunities that exist in cooperatives are also a special attraction for youth to be able to survive in cooperatives. Cooperatives are socially oriented institutions and sometimes even the excess profits obtained by cooperatives are also used to help the surrounding environment, but this is not a problem for youth, it is a source of pride because their work can help the community. This is also one of the interesting things because their work not only provides an opportunity for them to learn but also provides an opportunity to develop a business and have a positive impact on all members and the environment in which they live. In modern society, none can overlook the fact that economic benefits are not a priority. Regardless of the economic benefits, there are also other forces, which significantly motivate employees. Sdrali What motivates employees varies and depends on their temperament. Many employees are motivated by financial rewards, while others are by a combination of recognition and financial rewards . The social economy sector is often seen as a workplace, where people work not only for money but also because they find satisfaction in their work . In the social economy sector, employees bring to their job a greater commitment and non-financial orientation, find more challenge, variety and autonomy, and more intrinsic rewards. Based on the results of the erespondent for youth cooperative employees, most of them were employees who felt sufficient and even helped by the existence of the cooperative. Their current orientation is not only to improve their standard of living, but also their opportunity to learn and the benefits that they spread to members of the cooperative which incidentally is the surrounding community, is a source of pride for them. --- CONCLUSION It is too early to say that all cooperatives can change the standard of living of youth, in this case, of course, further research is needed for various types of cooperatives in different places. However, according to the findings there is evidence from respondent in these three cooperatives which their standard of living is capable of meeting or exceeding their former standard of living, as well as equivalent to the minimum wage given in the region where they live. Beside that youth can also develop themselves in the cooperative where they work, this is not even felt by youth who have worked in other places apart from cooperatives. Cooperatives become places of learning and occur through the daily interactions between youth as they learn to run business on their cooperative. Many employees are motivated by financial rewards, while others are by a combination of recognition and financial rewards now day cooperatives and their networks provide an expanded learning space for youth that make youth feel recognized because as part of the cooperative their opinions are valued. They also feel proud when they can spread the benefits of fellow members and their community.

Typical of social economy employees focus on social good rather than economic gain. They prefer to work in a sector that prioritizes satisfaction and makes them feel valued. Besides that, youth prefer to work in large places that provide greater welfare for them too. So usually working in Socially oriented businesses is made the second choice or next choice by youth. Some countries have policies to engage youth in Cooperatives. Youth represent the most important group among the working-age population. But then can cooperatives enterprise, as socially-oriented businesses, can provide and change the standard of living of youth? What motivates youth to engage in the socially oriented businesses?. This article examines field data from youth employees at cooperative members in Subang and Selangor. A survey was conducted by interview, observation and documentation study. According to the findings, now day cooperatives and their networks can change standard living of youth, but not only about it, cooperative provide an expanded learning space for youth, although there is differentiation by age, education and type of cooperative. The results of this study can attract the attention of youth to choose cooperatives as a new way to motivate youth to change their standard of living through cooperatives and learn a lot about business attitudes from this socially oriented business.

Background According to a report from the World Health Organization, the average life expectancy in Japan was 81.5 years for men and 86.9 years for women in 2019, which represent the longest lifespans in the world. However, the healthy life expectancies are shorter, at 72.6 years for men and 75.5 years for women [1]. Healthy life expectancy is the average length of time during which there are no restrictions on daily life [2]. Over the past 15 years, the average life and healthy life expectancies have increased for both men and women [3]. However, the gap between the average life expectancy and healthy life expectancy is large, at 8.9 years for men and 11.4 years for women; in other words, people live with some restrictions on their daily lives for about 10 years. Extending healthy life expectancy is considered an important issue in advanced countries as we approach the era of 100-year lifespans. Social participation plays a crucial role in the health and well-being of older adults. Levasseur et al. define social participation as "a person's involvement in activities that provide interaction with others in society or the community" [4]. Research has shown that socially active older adults, such as those involved in neighborhood associations, have a significantly lower risk of prefrailty [5]. Additionally, engaging in exercise-based social participation has been linked to recovery from frailty [6]. Furthermore, social participation has been associated with the maintenance of daily activities, including instrumental activities of daily living, as well as cognitive function [7,8]. Given the positive impact of social participation on the physical and mental health of older adults, it is imperative to promote social engagement in this population to prevent the need for long-term care and enhance healthy life expectancy. However, despite the potential benefits, a substantial number of older adults do not participate in social activities. In a study of older adults in Japan, 26.4% of 65-74-year-olds participated in "local-government organizations" , while 4.2% participated in "community development and community safety", 18.2% in "volunteering and social service through hobbies and sports", 1.8% in "traditional performing arts and craft techniques", 1.9% in "life support and child-rearing support", and 3.7% in "other" activities. However, 59.8% of respondents did not engage in any particular activity [9]. Furthermore, a study of 51,302 physically and cognitively independent persons aged ≥ 65 years living in 12 municipalities in Japan found that 54.1% had never belonged to a hobby activities group, 72.8% had never belonged to a sports group or club, and 82.2% had never belonged to a volunteer group [10]. These findings suggest that a significant proportion of Japanese older adults, estimated to be around 60%-80%, do not actively participate in any social activities. Similar trends have been observed in studies conducted in the United States and Europe, indicating that low social participation is not limited to Japan but is instead an issue among older individuals worldwide [11,12]. To extend healthy life expectancy, it is crucial to address the factors contributing to a lack of social participation among independent older adults. One potential reason for this phenomenon is a lack of self-efficacy. Social participation requires older individuals to possess various self-efficacies in different domains. Moreover, access to public transportation and related information is crucial to facilitate social participation [13][14][15]. Positive perceptions of the user-friendliness of the walking environment have also been associated with higher levels of social participation [16]. In addition, from an interpersonal perspective, individuals who are willing to communicate with their neighbors or friends are more likely to engage in social activities [17]. Moreover, a desire for social connectedness and enjoyment of group activities play significant roles in motivating older adults to participate therein [18]. Being personally invited, and feeling valued and acknowledged, further promotes social participation [15]. At the managerial level, health concerns and lack of confidence in physical strength were commonly cited by older adults as reasons for not engaging in social activities [19]. Flexibility in activity scheduling, including volunteering, has been identified as important for promoting social participation [15]. From a cultural perspective, access to information about available activities and services is crucial for encouraging older adults to participate in society [15]. Additionally, factors such as motivation to learn, personal interests, and the desire for personal growth can influence the willingness to engage in new interactions [20]. Lastly, older adults with greater social orientation tend to be more involved in social activities [21]. Considering these findings, it is evident that social participation among older adults requires the development of self-efficacy in the instrumental, interpersonal, managerial, and cultural domains. According to Bandura, self-efficacy refers to recognition of the possibility that an individual can effectively perform the necessary actions in a certain situation [22]. Bandura also distinguished two types of self-efficacy: general and specific self-efficacy. General selfefficacy relates to an individual's belief in their overall ability to handle various challenging situations and achieve desired outcomes. It refers to a broad and generalized sense of confidence in one's own capabilities to succeed in different life domains. People with high levels of general self-efficacy tend to have confidence in their problem-solving skills, resilience, and adaptability across a wide range of situations. They view challenges as opportunities for growth and are more likely to take on new tasks and persist in the face of obstacles. Specific self-efficacy refers to an individual's belief in their ability to perform particular tasks or activities successfully within a specific domain or context. This relates to the confidence an individual has in their capacity to accomplish specific goals, acquire specific skills, or execute specific behaviors. Specific self-efficacy beliefs are context-dependent and can vary across life domains, such as academic tasks, athletic performance, social interactions, and specific health behaviors. In summary, general self-efficacy reflects an individual's overall belief in their own competence and adaptability across various life domains, while specific self-efficacy pertains to the belief in one's capability to perform well in specific, context-dependent tasks or activities. Both types of self-efficacy influence individuals' motivation, choices, and behaviors and are important considerations when seeking to understand how people approach and navigate various challenges and opportunities in life. However, specific self-efficacy is thought to play a particularly important role in social participation among older adults, enabling them to lead healthy and fulfilling lives by enhancing their confidence in specific social participation-related skills and abilities, as well as their ability to cope with health challenges and adapt to changes in the environment. However, to date, studies have been limited to general self-efficacy [23][24][25][26], or to specific self-efficacy in the context of volunteering [27,28] or social participation by people with mental illness [29]. Specific self-efficacy has not yet been fully explored in the context of older adults' social participation. Developing a self-efficacy scale for social participation among older adults will be crucial for planning, implementing, and evaluating public health policies aimed at promoting healthy longevity among the increasingly large worldwide population of older adults. First, self-efficacy plays a vital role in motivating older adults to participate socially. By developing a self-efficacy scale for social participation, policymakers and implementers can assess the levels of self-efficacy among older adults and plan appropriate support and interventions. For instance, programs and resources could be provided for older adults with low self-efficacy in social participation to support healthy longevity. Second, a self-efficacy scale for social participation would be a valuable tool for quantitative assessment of the effectiveness of public health policies and interventions, including those aimed at improving older adults' self-efficacy in social participation. On the basis of evaluation results, strategic decisions could then be made to maximize the effectiveness of public health policies, again including those aimed at improving social participation and healthy longevity among older adults. Third, developing a self-efficacy scale for social participation would facilitate knowledge sharing related to the effects of policy on social participation among older adults, across regions and countries, by allowing the collection of standardized data. In this manner, policymakers and researchers would be able learn from each other's best practices. Improving policies on the basis of information sharing and comparisons at the international level will be highly beneficial in promoting social participation and healthy longevity among older adults. Therefore, we developed the Self-efficacy for Social Participation scale to assess the self-efficacy of community-dwelling older adults for social activities, and examined its reliability and validity in this study. Herein, self-efficacy for social participation is operationally defined as "the perceived ability to participate in activities shared in time and space with others". --- Methods --- Phase 1: instrument development First, to generate a pool of potential items for the SOSA, we conducted a literature review using PubMed and the following search terms: older people OR older adults OR community-dwelling older adults OR community-dwelling elderly OR elderly OR aged AND social engagement OR social activity OR volunteer activity OR volunteering OR social participation OR civic participation OR civic engagement. All fields were searched and the searches yielded 136 articles. Four papers dealing with elements of self-efficacy relevant to social participation were also included [17,21,30,31]. The criteria for selecting potential items for the SOSA were as follows: relevant to self-efficacy in the context of social participation; clear, logical, and easily understandable for older adults; relates to cognition or action. Finally, 32 items satisfying these criteria were identified. The 32 items were reviewed by four expert academic researchers and four independent older adults who did not require long-term care insurance. The researchers specialized in public and community health nursing, and also had experience in the development of scales for independent older adults. The independent older adults were volunteers residing in Hokkaido. The content validity, face validity, and practical utility of the 32 items were assessed, and the wording and necessity of each item were evaluated according to the opinions of the researchers and older adults. Through this process, the number of items was reduced from 32 to 28. The items were scored on a four-point Likert scale . --- Phase 2: instrument validation --- Participants and setting To validate the instrument, we aimed to recruit approximately 5,000 independent, community-dwelling older adults . To achieve this, all 5,351 community general support centers nationwide in Japan were contacted and asked to provide the questionnaire to one randomly selected person who met the inclusion criteria for the study, which were as follows: aged ≥ 65 years; living in the community ; and living as an independent older adult, i.e., no requirement for long-term care or support. These criteria were in accordance with the Certified Level of Need for Long-Term Care National Insurance of Japan . Data were collected between July and August 2022 via anonymous self-administered paper questionnaires. Completed questionnaires were returned to the academic office by the older adults. A total of 1,336 participants completed the questionnaire, and 1,292 were included in the final analysis after excluding thawed aged < 65 years. --- Measures The demographic characteristics of interest included age, sex, residential status, number of years spent residing in the area, education level, employment status, living situation, and disease treatment status . Two measures were used to assess the construct validity of the SOSA: the General Self-Efficacy Scale [32], which classifies perceived self-efficacy as high or low , and a single question about perceived health answered on a fourpoint scale . These scores were reversed using IBM SPSS Statistics , such that higher scores indicate better health. --- Statistical analyses IBM SPSS Statistics and Amos software were used for all analyses. Item analysis was used to investigate the reliability of the SOSA, and the factor structure was investigated by exploratory factor analysis. Items were excluded if the proportion of "Completely confident" responses was ≥ 80%, or if the non-response rate was ≥ 5%, there was evidence of multicollinearity , the correlation coefficient for the "item-total analysis" was ≤ 0.6 , or there was no significant difference between the highest-and lowest-scoring groups in the "good-poor analysis". Items remaining after the item analysis were subjected to exploratory generalized least squares factor analysis [33]. According to the eigenvalues and scree plots, a four-factor solution was obtained. We then repeated the factor analysis after excluding items with loadings < 0.4, and assuming a four-factor structure. Factors with a Cronbach's alpha ≥ 0.7 were considered reliable and construct validity was verified by confirmatory factor analysis. We examined model fit using the goodness-of-fit index , adjusted GFI , comparative fit index , and root mean square error of approximation . The model was accepted if the GFI and AGFI were ≥ 0.90, and the CFI was ≥ 0.95. When the RMSEA is < 0.05 the model fit is considered good, while values > 0.1 are undesirable. We also examined criterion-related validity by correlating the total SOSA score with those of the GSES and the subjective health status scores. --- Results --- Demographic characteristics Table 1 shows the demographic characteristics of the participants. The mean age was 72.6 years . In total, 55.7% of the participants were female, 63.4% lived with their family , 47.1% had resided in the same area for > 40 years, 47.6% had a high school level of education, 39.5% were employed , 81.3% indicated that they were not "in trouble" with respect to their living situation, and 82.6% had a medical condition for which they were currently being treated. --- Item analysis As shown in Table 2, following the item analysis of the initial 28-item version of the SOSA, none of the items were excluded on the basis of the criteria delineated above, except for four that exhibited multicollinearity . First, we compared items 12 and 13, and items 25 and 26, and decided to retain items 12 and 25 given their importance to the scale. Items 13 and 26 were excluded such that 26 items were retained for the factor analysis. --- Factor structure The results of the exploratory generalized least squares factor analysis of the 26-item scale are shown in Table 3. The eigenvalues for the one-two, three-, four-and fivefactor solutions were 12.242, 1.863, 1.368, 1.181, and 0.795, respectively. We repeated the exploratory factor analysis with promax rotation until the item factor loadings exceeded 0.4; items 2, 4, 5, 7-11, 14-18 and 28 were excluded because their factor loadings did not exceed 0.4 in any analysis. After excluding those items, differences in factor loadings between factors became apparent. Excluding items with loadings of < 0.4 yielded a fourfactor solution. In the final version of the scale, 12 items were loaded onto the four factors. Factor 1 included three items and was labeled "instrumental self-efficacy" . Factor 2 included three items and was labeled "managerial self-efficacy" . Factor 3 included three items and was labeled "interpersonal self-efficacy" . Factor 4 included three items and was labeled "cultural selfefficacy" . Together, the four factors explained 63.21% of the variance in SOSA scores. --- Internal consistency and validity The fit indices for the four-factor model were as follows: GFI = 0.948, AGFI = 0.915, CFI = 0.952, and RMSEA = 0.078 . The values indicated good construct validity for the SOSA. The Cronbach's alpha coefficients were 0.84, 0.81, 0.80, 0.81 and 0.90 for factors 1-4 and the entire scale, respectively . Pearson's correlation analysis showed a correlation between the total SOSA score and GSES and subjective health status scores. The SOSA score showed a highly positive correlation with the GSES score and a moderate positive correlation with the subjective health status score . --- Discussion We developed the SOSA to evaluate the self-efficacy of community-dwelling older adults for participation in social activities, and examined its reliability and validity. The final scale comprised 12 items and four factors: instrumental self-efficacy, managerial self-efficacy, interpersonal self-efficacy, and cultural self-efficacy. The Cronbach's alpha coefficient for the scale indicated satisfactory internal consistency; this was also the case for criterion-related validity according to the correlations between the SOSA score and those for the GSES and subjective health status. Exploratory and confirmatory factor analyses confirmed the construct validity of the SOSA, which has sufficient reliability and validity to measure self-efficacy for participation in social activities among community-dwelling older adults. The first factor of the SOSA, instrumental self-efficacy, represents older adults' self-efficacy for independent living and mobility . Lawton distinguished seven levels of human abilities, ranging from basic life maintenance to advanced and complex social abilities [34]. The items of the SOSA relate to instrumental self-maintenance in a manner that represents the various stages proposed by Lawton. Factor 1 relates to basic self-efficacy for social participation. In a previous study, a decline in the ability to perform instrumental activities of daily living was negatively related to social participation among older adults [35]. Other studies have shown that neighborhood walkability [36] and the use of public transportation [14] promote social participation among older adults. Thus, higher instrumental self-efficacy may promote social participation by providing the means to engage in basic outings and interactions. The second factor of the SOSA, managerial self-efficacy, represents self-efficacy for independent management Fig. 1 Confirmatory factor analysis of the Self-efficacy for Social Participation scale of one's life . After retirement, older adults have to consciously reconstruct many parts of their lives that were previously managed by their employer, which requires the setting of new goals consistent with their values and interests, and management of their lives, while also adapting to the health status changes associated with aging [37]; these health status changes are directly associated with social participation [38]. In addition, in today's society, the use of personal computers, cell phones, and smartphones has become widespread, and the number of older people who own such devices is increasing [39]. Information retrieval from the Internet is becoming more widespread among older adults [40]. Therefore, managerial self-efficacy as it pertains to the appropriate application of information technology to daily life is also essential for older adults to participate in society. The third factor of the SOSA, interpersonal self-efficacy, represents self-efficacy for socializing and interacting with others . Previous qualitative research has identified factors that promote social participation in older adults, such as staying active [41]. In another study, older adults who actively interacted with their neighbors were more likely to participate in social activities [21]. Other factors related to social participation and volunteer activities include good relationships with other people in the community, reflected in having a large number of close friends and receiving invitations to participate in activities [42]. In summary, interpersonal self-efficacy may promote social participation by expanding one's interests and horizons as they relate to other people. The fourth factor of the SOSA, cultural self-efficacy, represents self-efficacy for forming new habits and engaging in socially oriented behaviors . Culture has been described as "all of the behavioral patterns, arts, beliefs, customs, ways of life, and other products of a group of people that are handed down in society and guide worldviews and decision-making" [43]. Aoo et al. argued that, in Japan, older adults are often viewed as "beneficiaries" or people in need of support [44]. They also stated that, in many countries experiencing a rapid population decline, including Japan, older adults support those in need, rather than being exclusively the beneficiaries of support, and there is a need to cultivate the view that at least some older people can play a more active and responsible role in society. In a study of the requirements for social participation among older adults [15], the availability of information relevant to activities was the most important factor for engaging in a new social activity, although willingness to learn was also important [17]. Greater cultural self-efficacy may lead to a new culture of proactive social participation among older adults. A novel aspect of the SOSA is its focus on self-efficacy for social participation. Previous measures of social participation among older adults assessed the frequency of social participation [45], motivation to engage in activities [46], attitudes toward activities [47], and confidence and beliefs relating to participation in a particular activity [27]. The SOSA is a broad measure of social participation that assesses the role of instrumental, managerial, interpersonal, and cultural self-efficacy. There are three main potential applications of the SOSA. First, it can be used to evaluate the self-efficacy of older adults for social participation and inform interventions designed for this population. Second, it can identify older adults that provide, rather than merely benefit from, social support. Third, it could lead to the creation of new community systems and policies fostering social participation among older adults, especially through comparison of a given community with others. This study had some limitations. First, because it used a cross-sectional design, the predictive validity of the SOSA remains to be established. In the future, the longitudinal association between SOSA and social participation should be examined. Specifically, it is necessary to determine whether older people with higher SOSA scores can promote their behavioralization of social participation. Second, the response rate was moderate in this study . Third, although questionnaires were distributed to community general support centers nationwide, regional bias in the respondents cannot be When responses are limited to specific individuals or regions, questions arise about whether the results can be accurately applied to the entire population. In such cases, the skewed representation of respondents can potentially distort results or overly reflect certain trends. Moreover, the geographical bias among respondents is an important factor. Different socioeconomic, cultural, and environmental factors exist across regions both within and outside Japan, which can influence SOSA and social participation. There are several approaches to addressing these limitations. First and foremost, it's crucial to devise survey methods that enhance response rates by making surveys accessible to the target population in older people-friendly formats. Furthermore, to mitigate regional bias, securing an adequate sample from various regions and considering methods to appropriately adjust for regional differences both within and outside Japan are essential. --- Conclusion The SOSA consists of instrumental self-efficacy, managerial self-efficacy, interpersonal self-efficacy and cultural self-efficacy showed sufficient reliability and validity to assess self-efficacy for social participation among older adults. This scale has the potential to help create interventions and systems, policies that could promote social participation by assessing the self-efficacy of community-dwelling older adults for social participation. Furthermore, the scale could help in efforts to improve the physical and mental health and longevity of older adults through increased social participation. --- --- Abbreviations --- AGFI Adjusted goodness-of-fit index CFI Comparative fit index GFI Goodness-of-fit index GSES General Self-Efficacy Scale RMSEA Root mean square error of approximation SOSA Self-efficacy for Social Participation scale --- --- --- Funding The sources of funding for our study were Grants-in-Aid for Scientific Research Type A, 17H01614of the Japan Society for the Promotion of Science . The funders had no role in data collection and analysis, decision to publish. --- --- --- Competing interests The authors declare no competing interests. ---

Background Social participation is important for the health of older adults and super-aging societies. However, relatively few independent older adults in advanced countries actually participate in society, even though many of them have the capacity to do so. One possible reason for this could be a lack of self-efficacy for social participation. However, few scales have been developed to measure self-efficacy for social participation among community-dwelling independent older adults. Therefore, we developed the "Self-efficacy for Social Participation" scale (SOSA) to assess the self-efficacy of community-dwelling independent older adults, and examined the scale's reliability and validity.We distributed a self-administered mail survey to approximately 5,000 randomly selected independent older adults throughout Japan. The construct validity of the SOSA was determined using exploratory and confirmatory factor analyses. Criterion-related validity was assessed using the General Self-Efficacy Scale (GSES) and according to subjective health status.In total, 1,336 older adults responded to the survey. Exploratory and confirmatory factor analyses identified 12 items distributed among four factors: instrumental self-efficacy, managerial self-efficacy, interpersonal self-efficacy and cultural self-efficacy. The final model had a Cronbach's alpha of 0.90, goodness-of-fit index of 0.948, adjusted goodness-of-fit index of 0.915, comparative fit index of 0.952, and root mean square error of approximation of 0.078. Significant correlations existed between the SOSA score and GSES (r = 0.550, p < 0.01) and subjective health status (r = 0.384, p < 0.01) scores.The SOSA showed sufficient reliability and validity to assess self-efficacy for social participation among older adults. This scale could aid efforts to improve the physical and mental health, and longevity, of older adults through increased behavioralizing social participation.

of the iceberg. Equally important are the covert manifestations of racism and bias in society and in institutions, such as white silence, denial of institutional racism, fear of people of color, and microaggressions. Indeed, these latter realities negatively affect education in the health professions, learning and work environments, the well-being of students, trainees, and colleagues, and the health of the community. Disparities in the infection and death rates due to coronavirus disease 2019 provide a current example of the interaction between structural racism, social risk factors, and inequities in health care access and quality that affect the health of Black people in the US. Structural racism encompasses the sum of the ways in which societies perpetuate racial discrimination through mutually reinforcing systems. Although racial injustice and discrimination are not new, after recent egregious incidents, many organizations issued calls to action with an eye toward addressing racism and discriminatory practices in medicine and science. Antiracism practice and training are recommended to address racism and bias and advance health equity. It is essential to respond to the moral imperative of mitigating the effects of racism, discrimination, and bias in medicine. The latter compromises the ability of health care professionals to fulfill their oaths and ethical responsibilities to treat patients, student, trainees, and colleagues with respect, justice, and equity. It is time to take bold, but measured, steps to bring antiracism training to fruition and begin to unravel the pernicious effects of systemic racism in the nation's medical schools, academic medical centers, and teaching hospitals. There also must be a commitment to move the needle by making real advances and not be satisfied with just undertaking a few time-limited, box-checking training activities that result in no improvement in antiracism competencies or in institutional measures of inclusion, equity, or health outcomes. There are several recommended strategies that medical schools, academic medical centers, and is required, not optional. Because antiracist training will be required for stakeholders from all mission areas, it must be designed with broad relevance in mind. Second, establishing a steering committee composed of institutional leaders, faculty, staff, and students and trainees to oversee and direct antiracist training initiatives is essential for success. Institutional transformation requires leadership and stakeholder engagement, sustained institutional commitment, as well as resource and incentive allocation. The diverse composition of the committee blends strengths from leadership and grassroots perspectives. It must also rely on data-driven methods to identify and monitor evidence of disparities, inequities, and bias across the institution. In so doing, the steering committee will be in a position to recommend appropriate administrative, process, or training interventions to drive improvement. Third, building an antiracism vocabulary with a glossary of terms and references is needed to create a common language with an eye toward promoting precision and consistency in communication. A common vocabulary will also promote the cultivation of a shared cognitive model by providing stakeholders with common terminology and constructs. A fourth strategy of developing a longitudinal antiracist training curriculum marked by an iterative approach supports continuous quality improvement of curricula as well as meeting stakeholder needs. Using a longitudinal approach also permits the periodic reinforcement of core A sixth strategy involves medical schools, academic medical centers, and teaching hospitals. These institutions should develop external partnerships to build trust with vulnerable communities and community organizations. Establishing an external review process will further enhance the effectiveness of antiracist training initiatives and responsiveness to community needs and concerns. Partnerships with professional associations and societies can also facilitate advocacy for the standardization of antiracism competencies and the incorporation thereof into Entrustable Professional Activities-activities that all medical students should be able to perform upon entering residency-or accreditation standards. The ability of physicians to fulfill their oaths, create inclusive learning and work environments, provide bias-free education, and deliver quality care that reduces health care disparities is undermined by bias, discrimination, racism, and systems of oppression in medicine and science. Commitment to advancing antiracism training is essential for deconstructing the effects of these pernicious elements and improve health equity. --- ARTICLE INFORMATION Open Access: This is an open access article distributed under the terms of the CC-BY License. Corresponding Author: Eli Y. Adashi, MD, MS, Brown University, 272 George St, Providence, RI 02906 . Author Affiliations: College of Medicine, Baylor University, Houston, Texas ; Cullen Trust for Higher Education, Houston, Texas ; Medical College of Georgia, Augusta University, Augusta ; Warren Alpert Medical School, Brown University, Providence, Rhode Island . --- Conflict of Interest Disclosures: Dr Adashi reported receiving personal fees from Ohana Biosciences Inc. No other disclosures were reported.

Racism and bias are antithetical to the oaths, moral commitments, and ethical responsibilities upheld by health professionals who are dedicated to the missions of medicine. Accordingly, educators, healers, and role models for the next generation of physicians must redouble their efforts to work alongside leaders at medical schools, academic health centers, and teaching hospitals to address racism, bias, and other factors that negatively affect the health and well-being of students and trainees, colleagues, patients, and communities. The horrific deaths of numerous unarmed Black people-such as the killing of George Floyd in May 2020-highlight the twin traumas of racism and police brutality that affect the lives of Black people in the US. Overt acts of racism such as violence and public harassment constitute only the tip

La Psicología Sociocultural como Teoría Postformal del Rendimiento Académico: Interrogando la Educación Formal Este artículo se cuestiona la dominación de la educación formal. Dado que la educación formal se basa en la habilidad de rendimiento académico como objetivo, la exploración de enfoques post-formales, como la noción sociocultural de rendimiento académico es el aspecto central de este trabajo. Se pretende interrogar la dominancia cultural existente en la educación formal en lo que se refiere a la necesidad de contar con grupos estereotipados por cuestión de habilidad, sin descartar totalmente la educación formal. A través del camino marcado por la experiencia socio-cultural de las niñas y niños, este artículo también explora su proceso de identificación social con el sistema educativo actual. La forma cómo la identificación de estudiantes se construye y co-construye o reconoce o rechaza el aspecto de rendimiento de la educación es un aspecto de máxima disputa. En general, el artículo intenta responder a la siguiente pregunta psicológica básica: "¿Por qué una forma particular de educación y rendimiento bajo el discurso dominante en educación es legitimada y valorada en las representaciones psicológicas sociales?" academic performance?' The answer to this question has been explored through various approaches via, cognitive , motivational , contextual and cultural . Cognitive approach insisted on development of mental structure where various information processing activities happens pertaining to the task given, for example, intellectual ability of the student in mathematical tasks. Motivational approach conceptualized factors that influence learning, such as factors that directs or limits choice of action and factors that affect intensity of engagement with the task . Contextual approach emphasized the immediate social context of learner which either hamper or promote learning, for example, social class or socioeconomic status. And cultural approach emphasized the use of artifacts in a social space through which children are socialized. Many theorist visualized the role of context and culture as observed to be the real causal factors where the chances of explaining psychological processes was expected to be better. However, it was observed that these two factors have been interrogated as a separate entity rather than as macro level forces shaping the individual level phenomenon. Contrary to this, the mainstream psychology separated its form and structure from social experiences and history of people from diverse background . In this regard, present work pondered on sociocultural position as a critical postformalist catapult aiming at understanding and interrogating formalist legacies dominating the discourse of academic achievement and education. Recent direction for introducing continuous and comprehensive evaluation system in Indian schools is an attempt to relieve students from the burden and stress of exam has important policy implication. This is done by more uniform and comprehensive patterns in education for the children all over the nation. Though CCE is expected to improve students' classroom performance by identifying students' learning difficulties by regularly employing suitable remedial measures for enhancing their learning performance, still the pedagogy T he most basic question often repeated in the educational discourse is that 'why some students get difficulty in adjusting with school environment resulting in either dropout or low IJEP -International Journal ofEducational Psychology, 2 223 The opposites ofgreat truths may also be true; it is only the opposites ofsmall truths that are false and curriculum is regulated by the formalists' agenda of mainstream education. This drive towards the homogenization and nation building in contemporary society through an effort to control diversity in a school seems to be regulated by prevailing societal values representing the dominant identities in power. However, there is a prospect in revisiting to the failed attempts of improving students' performance, by interrogating existing structure and history of society. The issues of identity and sociocultural experiences of students from disadvantaged background have not been addressed openly, thus limiting education to increase in literacy rate only. The scheme of continuous and comprehensive evaluation depends upon the discretion of schools to plan their own academic schedules as per specified guidelines on CCE. However, the possibility of negligence of low performing students who do not fit into the school value system can't be denied. Academic performance of students, as represented by the dominant society, preferred to value cognitive ability as it appeared legitimate in the technocratic world motivated by the values of economic gains. The third world countries were the colony of the western power that came to understand orients for their political and economical expansion. Due to their powerful and structured economy the notions of the methodologies were observed to be superior for controlling the colonial nations . Thus, the western interlocutor prioritized and legitimized western values to be better than the indigenous. This took historical turn by recognizing new divide in the social structure as elite and non elite. The mode which was creating this divide was based on the various metaphors of intelligent quotient , where privileged were considered as moral, intelligent, hardworking and gracious as compared with the underprivileged. Any other opinion how much empirically validated was rejected unless accepted by the scientific community. This formalist approach highlighted individualistic agency responsible for prevailing inequality in ability and academic achievement. In other words, it gave importance to the permanent aspects of one's ability rather than shaping it through adequate environment . It also goes beyond the entity and malleable notions of ability to the reproduction of inequality legitimizing dominant identity as genuine. The criticism of formalist approach is not new . The formalist approach represented a paradigm of thought that ruled the educational system and was the powerful criterion under which various facets of human agencies were judged. The other approaches whose metatheory derived its aspects from the experience of oppression in the history by dominant culture were either not prioritized or abandoned. The Postformalists' perspective got space in academic circle via the dominant discipline like sociology and political science under the umbrella of postmodern thoughts. However, when the call for intervention arises, the formalist conception of ability became more prominent. This defines the power structure of society which is regulated by political dynamics driven by power ideology. The sociocultural apparatus shaped through the diversity of experience has different metatheory of assessment when compared with the universal model of academic achievement. The social constructivist viewpoint of Vygotsky and his contemporaries acknowledged the social and political dynamics of micro human behaviour. It was based on the assumption that human activities are embedded in cultural contexts, mediated by artifacts like language and other symbolic systems, and can be better understood by exploring the history of development . Emphasizing the sociocultural facets of children may offer better insight into the problems of formalist approach and create a platform for understanding postformalist agenda. The formalist approach to ability and academic achievement became universal phenomenon and took the form of grand truth worldwide. The problem of the democratic education as expressed by the formalist may be derived from the dominant value system considered as legitimate. Any opposite patterns of thought is not valued until its grandiosity gets fixated by the other truth emerging out from actors' viewpoint and experience . The challenge of sociocultural psychology towards the psychometric tools of formalist created alternative inputs to the politics of psychology and its philosophy. Psychologists have long been interested in knowing the causal factors behind high and low performance of students in the classroom. These causal factors dominated the construct academic achievement positioning them in the dominant worldviews in many ways, two of which were quite prominent representing traditional and realist epistemology, namely, the organismic and mechanistic nature of human agency . The Organismic view holds Piagetian or schema-driven brand of constructivism in which self organization was an inherent feature of the organism, a tendency most evident in the activity of the human mind which was nurtured under the paradigm of rationalism. Themechanistic world view was tailored under the academic regime of realism which was philosophical antithesis of Piagetian constructivism. These worldviews were observed to be more individualistic rather than social in orientation and was placed under the deficit model of achievement . Apart from the traditional and realist worldviews, the alternative worldviews comprised sociocultural model, symbolic interactionist model and 'mind in society' model. These alternative worldviews were more context driven, and were positioned under the postmodernist paradigm of social constructivism rejecting the formalism completely . Social constructivist perspectives focused on the interdependence of social and individual processes in the co-construction of knowledge . The social constructivist comprises mainly of Piagetian and Vygotskian explanation. However, presently the focus shall be on Vygotskian notion of academic achievement and also an effort will be made to interlink and differentiate it from other perspectives. Apart from these perspectives of education, rest shall be presumed to be inherently the area of formalist agenda of mainstream educational psychology .The reason behind this categorization as formalist and post-formalist educational psychology is manifold. One of the reasons which impelled the present discussion in this direction is not universal but more or less based on sociocultural understanding of academic achievement. The formalist forms of education, though, fiercely debated under diverse disciplinary circle compared the students of ability stereotyped group under the same mainstream and middle class educational value system. It was obliquely stated that those who were not fitting under the formalist system of education were enough to be projected as deficit in ability, thus strengthening the existing legitimizing myths portrayed by the traditional class and culture . These formalist approach dominant in educational system due to colonial impacts demeans the cultural and linguistic diversity of historically marginalized students and became reified as common sense knowledge. This representation of education in the form of academic achievement disregarded other aspects and paradigms of education. For example, category of students involved in proper education, their achievement as compared to underachiever or low achiever, their cultural representation in schools, their social identification never had became part of people's understanding of academic achievement. Urgent need to understand other aspects of education and their representation is the need of present hour. A very practical illustration of present educational system is its classroom effect which has sustained the authority of past educational system in its discourses. In this sense, the representations of formalist education weakens the position of students' from marginalized and low socioeconomic status background and labeled their under performance in the school as deficit and not as different from the children from un-marginalized and high-SES background . Some cultural arguments problematically define certain ethno-racial identities and cultures as subtractive from the goal of academic mobility while defining the ethnic cultures and identities of others as additive and oriented toward this goal . This has shown that dominant formalist force accepted the superiority of the students coming from privileged socioeconomic background . In the process of judging students' academic achievement, larger educational context was never interrogated. Meacham argued from Vygotskian perspective that 'a culturally diverse learning environment, in contrast to the tradition of deficit, may embody important advantages in higher-order conceptual development ' . The exploration of factors beyond the individual and structural dimensions in terms of children viewpoint about themselves in particular social context and situations has been major concern of sociocultural theory. The understanding of these dimensions may hold the possibility of mitigating the gap in terms of cultural practices of marginalized communities and those assumptions of the school regarding learning, which were expected to be beneficial for literacy achievement . Thus, totally rejecting formalism, as supported in the postformalist formulation of education may overlook the link between policy and practice. In this context, Sharma posited that children who belong to extremely marginalized communities may get certain sense of empowerment through the knowledge of letters and limited access to any kind of formal education. At the outset, it seems that people of minority and disadvantaged background when come in contact with the outgroup context justify their present status as legitimate . This justification of underachievement by people of disadvantaged background undermines their sociocultural experience as deficit and not equally important. This may project marginalized members as uncultured and bastion them with imposition of education that is not representative. This demerit of formal education doesn't reduce its charm among policy makers and educators. However, linking of several aspects of formal and post formal education enrich the substance of education which are the fundamental right of every child. On the other hand, Govinda and Bandopadhyay recently pointed towards the multifold expansion of educational infrastructure for the improvement of accessibility and availability of education, the way system has grown seems to be contributing to further social divisions in the country. The nature of social division attribute to discriminatory factors causing more psychological harms rather than perception of equality. Therefore, varied paradigms comprising implicit processes of self and cognition due to one's experiences with the contexts and practices of artifacts also need to be vigorously debated. --- Factual understanding of sociocultural theory and literacy acquisition Literacy acquisition has been the central concern of sociocultural theory . Scribner and Cole in their analysis of relation between literacy and cognitive development of a child expressed possibility that literacy acquisition can be independent of schooling and have contextual implication in the development of cognitive competencies. Sociocultural approaches emphasize the interdependence of social and individual processes in the coconstruction of knowledge . One reason attributed was that, "children from working-class and lowersocioeconomic-class homes do not ascribe the same importance to the mental functions required by intelligence tests or achievement tests and academic work in the same way as do middle-and upper-middle-class students" . Studies showed that school failure resulted from the cultural inferiority of the poor or the marginalized and teaches us that power relations between groups must be reconsidered when students' performance is studied . Also, it was posited that, working-class and poor students often see academic work as unreal, as a series of short-term tasks rather than something with a long-term relationship to their lives . The social context and power relations of the culture at large and the school culture in particular may be essential in understanding the class and cultural dynamics of student performance . Kincheloe emphasized socio-political cognitive theory which tried to understand the way consciousness and subjectivity is shaped by the society. This emphasis on socio-political theory rejects the Cartesian-Newtonian mechanistic world view that is embedded in the cause-effect, hypothetical-deductive system of reasoning. Lev Vygotsky theorized in the 1930s that individuals do not develop in isolation but in a series of interconnected social matrices in which cognition is viewed as a social function . In a socio-psychological context, Vygotsky's work creates a space where integration between macro social forces and micro psychological forces takes place. Analysis of these integrated spaces becomes a central activity for a democratic post-formal educational psychology concerned with the way identity is formed by large social forces and mediated by individuals operating in specific environment . Such understanding allows us to imagine pedagogies that move individuals to greater understanding of themselves and their relation to the world, to higher orders of thinking, previously unimagined . The most fundamental concept of sociocultural theory is that the human mind is mediated . Sociocultural revolution focused on learning in out-of-the school context and on acquisition of skills through social interaction . Failure of educational system has resulted into new revolutions which very much deviated from the established framework of looking at education. Vygotsky argued that human being do not act directly on the physical world but with the help of cultural tools and labor activities. This gives us the freedom of self to operate on its ecology and systems and to change it. The use of symbolic/cultural tools or signs, to mediate and regulate our interaction and operation with the others is the major characteristic of sociocultural model of human experience. --- Recent development in sociocultural theory in postformalist context Child's mind is, as pointed by sociocultural theorists, culturally shaped and has the flexibility to grasp the utility of the artifacts in the social settings in the form of experiences, thus developing new identities. In the process of understanding children in their school contexts, Vygotsky reasoned that adequate approach to the study of higher mental abilities is through genetic analysis . Sociocultural theory recognized four genetic domains viz., phylogenetic domain, sociocultural domain, ontogenetic domain and microgenetic domain, though, most of the research has been carried out in the ontogenetic domain . For example, focusing on exploring the ways in which abilities such as voluntary memory are formed in children through the integration of meditational means into the thinking process . However, these four aspects were found to be interwoven together in the development analysis from Vygotskian perspective . Hence, it was with the application of ontogenetic analysis that the complex interplay of meditational tools, the individual, and the social world is explored to understand learning and development and the transformation of tools, practices, and institutions . According to Lantolf their mental system had been reformed as a result of their participation in a culturally specified activity known as schooling . A well established fact of child cognitive development fragmented in the stages were challenged by the notion that learning is not the result of pre-established stage of certain form of maturation but rather as result of social interactions and socially learned phenomenon giving impetus to the inner development of child. The social context ascribes varied meaning to the individual or group performing the task due differences in motives and goals underlying the behaviour . Activities in different settings do not seems to unfold smoothly but there may be chain of one activity reshaping itself into another activity in the course of its unfolding . Shift in activity may increase the need to discover different meditational tools for carrying out new activities with the help of identifies group or peer. In this regard, Palincsar pointed that "the peer collaboration resembled interactions between teachers and children, resulting in the generation of new story elements and more mature forms of activities. Thus facilitative aspects of peer interactions in the form of shared perspectives imparted more meaning to sociocultural psychology of children. --- Social class and sociocultural experience Behaviorist and latter constructivist agenda was limited to discourse of teaching and learning, pontificating the framework of individual agency based on maturation and rewards, thus ignoring cultural-historicalpolitical forces. In the classroom discourses, students form a shared identity with each other which can be very effective factors to be utilized for practical learning through dialogues and discussion. Gee suggested that as researcher and teacher we must go beyond mere recognition of discourses' role in producing or potentially challenging hierarchies of power. Therefore, it becomes foremost to IJEP -International Journal ofEducational Psychology, 2 231 look into the basic tenets of child which had its genesis in the sociocultural configuration and experiences . This sociocultural format has been dominantly synchronized by the Childs' SES whose definition has became more contextual rather than unequivocal as in earlier formalists rudiments. It was observed that despite expansion in educational reforms and access to education, the subtle form of discrimination still continues. The exclusion and blatant sort of discrimination faced by children depends upon their position in the social ladder both because of their social identity and their role in a domain. Burkit pointed toward social class as a fit for certain category of capitals essential in ones understanding of social selves. Categorization of SES as objective criterion for measuring ones hierarchical position is based on set of variables which is clustered and complementary. Thus, for the French sociologist Pierre Bourdieu, the social class differences and distinctions between individuals that influence their biographical trajectories and identities were not just based in the ownership or non-ownership of material capital, or in the person's relation to the division of labour, but also depend upon the possession of cultural, social and symbolic capital . These capitals can be associated with Vygotsky's sociocultural and postformal theoretical assumptions given long before Bourdieu's thesis. However, these associations of capitals decide the social position of the individual in any social situations such as classroom. According to Bourdieu each individual occupies a position in a multidimensional social space or fieldwhere he or she is not defined only by social class membership, but by every single kind of capital he or she can articulate through social relations. These invisible and visible accumulations of capital include the value of social networks, which Bourdieu showed could be used to produce or reproduce inequality. The argument tried to differentiate the cultural control from the sociocultural experiences where social class as a cultural perception and practice seems to have its genesis in the history of legitimate ideology. To simplify it further, the perception of one's objective position on the socioeconomic ladder can be a derivative of one's self concept, values and beliefs depending upon the reciprocal interaction of cultural variations or social environments with one's individualistic self. The notion of individualistic self used here draw from phylogenetic reality of individual which at one hand categorize as prototype of human being and at other as a individual having experience gained from own human agency. This complex structure of human and society impels Snibbe& Markus to remark that, "Cultural models are sets of assumptions that are widely shared by a group of people, existing both in individual minds and in public artefacts, institutions, and practices. At the individual level, these cultural models provide implicit blueprints of how to think, feel, and act. When people act according to these blueprints, they reproduce the public models, thereby perpetuating the cultural context from which both were derived.". Above definition of cultural model can also add to its three major forms, namely, religion, SES and region , where SES has been seen as of major practical importance. The American Psychological Association's Task Force on Socioeconomic Status recently noted that differences in socioeconomic status and social class have important implications for human development, wellbeing, and physical health. In research on socioeconomic status and social class, these are commonly operationalized as combinations of variables such as income, education, and occupational prestige. When investigating social class and socioeconomic status, many investigators also probe subjective social class, or individuals' estimation of their own social class . People may perceive their social class to be different from what objective indicators might suggest . Thus, socioeconomic and class inequality may be perceived not only in terms of tangible resources such as income but also in terms of structural aspects such as power, privilege, and social capital . Cohen highlighted that, "whereas much attention has been paid to the effects that socioeconomic status and social class have on domains such as health, development, and well-being, psychologists have not often taken a culturally informed approach or considered the rich culturally textured beliefs, values, and practices of higher versus lower social class individuals" . Snibbe & Markus through various experiments had shown how people of low and high socioeconomic status differ in their views of agency. It was found that high socioeconomic status people are more able to control their environments and influence others whereas those of low socioeconomic status are more likely to have to adapt to their surroundings and maintain their integrity because of their inability to directly control their environments . Thus, Snibbe and Markus claimed that the culture of high socioeconomic status valued control and agency, whereas the culture of low socioeconomic status valued flexibility, integrity, and resilience . Thus, it can be concluded that children of different socioeconomic status are enculturated to have different values . Providing meaningful education for all children sets the agenda for more diverse form of education to the child . In this context, Moll asserted that "in studying human beings dynamically, within their social circumstances, in their full complexity, we gain a more complete and a much more valid understanding of them . Failure of the school to serve children from all diverse background have been explained through the following sociocultural explanations viz., a) discontinuities between the culture of the home and school , b) mismatches in the communicative practices between children of lower class and SES and mainstream teachers who represent monolithic value system of middle social class that lead to miscommunication and misjudgment , c) the internalization of negative stereotypes by minority groups or people of working class who have been marginalized and may see school as a site for opposition and resistance , and, d) relational issues, such as the failure to attain mutual trust between teachers and students and a shared sense of identification between the teacher and the learner . Adding to the above sociocultural explanations of mismatches between value assumption of child and the school, the children co-construct their knowledge system in the social processes with their use and familiarity with the artifacts. Thus, we may call for alternative views that reconsider tradition and scheme of schools and provide major overhauling through awareness. This is required to have a shift in the perceptions of an observer and to value the agency of the child which is actor and bearer of the oppressive situations. Therefore it becomes important in understanding child's appropriation of his/her cultural values and to provide better education from the diverse perspective. Sociocultural experience in text: Reconsidering tool for literacy and pedagogy According to Giroux , critical pedagogy is situated as a political and moral project. Its proponents recognize that pedagogy is always political because it is connected to the formation and acquisition of agency. As a political project, it illuminates the relationships among knowledge, authority, and power drawing attention to questions concerning who has control over the conditions for the production of knowledge, values, and skills. Moreover, it sheds light on the ways in which knowledge, identities, and authority are constructed within particular circuits of power. Most importantly, it draws attention to the fact that pedagogy is a deliberate attempt on the part of educators to influence how and what knowledge and subjectivities are produced within particular sets of social relations. Ethically, critical pedagogy stresses the importance of understanding what actually happens in classrooms and other educational settings. This was done through raising questions regarding the choice, direction and desirability of knowledge. It also takes seriously the important relationship between how we learned and acted as individual and social agents. In this instance, critical pedagogy was concerned with teaching students not only how to think but also how to assume a measure of individual and social responsibility-that is, what it means to be responsible for one's actions as part of a broader attempt to be an collectively engaged citizen. --- Prospects and conclusion We discussed about the role of sociocultural psychology as a postformal approach. The challenges associated with the formal education in terms of increasing achievement gap is not new and more serious attempt is required to understand the existing reform policies in education. The present work, however, highlighted the nuances and merits associated with formal and post formal viewpoints only, and highlighted the need IJEP -International Journal ofEducational Psychology, 2 235 for understanding sociocultural aspects of human psychology. The necessity to understand the problems and prospects of both the perspectives may provide better picture of educational system. In recent times, lot of researchers have attempted to look into the arguments presented in this paper through different cultural contexts, yet many questions still remained to be answered. As it is evident from the review presented here, this topic is one with manifold aspects to its ranging from broad ones such as cultural issues, government policies and plans to subtle nuances such as teaching strategies and curricula. Hence future researchers may consider the employment of collaborative effort from social scientists belonging to various disciplines so that the different issues associated with the subject maybe dealt appropriately. The need also arises to understand the tenets of social class and SES as structure under which various other identities gets represented and constructed depending upon the volatility of social context and situation. The universal aspect of social class may not only mutually constitute the individual and structural factors but also convey about the construction of selves depending upon the situation of the domain, that is system of education and classroom affects. Giroux pointed in migratory context of America that it is time for Americans to take note of the fundamental importance of retaining educational theories and pedagogical practices that produce the knowledge, values, and formative culture necessary for young people to believe that democracy is worth fighting for. Taking the recourse from Giroux and Portes , intentions are to develop awareness program to reinvent the society, so that its education system may understand real meaning of democracy and stay away from its myths sidelining itself from corrosive and oppressive corridors. The generation of empowerment among marginalized both in perceptions and objectivity may then reflect the possibilities of diversity inclusion. Thus, the need is to respect diverse form of education suitable for everyone's sociocultural experience without legitimizing one form of educational culture and methodology. It may be more justifiable to acknowledge the promises associated with both formal and post formal educational system. That may create more democratic framework for education where no child is neglected for being not fitting into the systems and values of other identities.

The present paper interrogates the dominance of formal education. As formal education system relies on ability based academic achievement as a goal, exploring post-formal approaches, such as sociocultural notion of academic achievement is the hallmark of present paper. An attempt is made to interrogate the existing cultural dominance in formal education referring to the need of ability stereotyped groups, not discarding formal education totally. Taking the route from sociocultural experience of children, paper also explores their process of social identification with the present educational system. The way students identification gets constructed and co-constructed, either acknowledges or discards the achievement domain of education, is the major point of contention. Overall, the paper tries to answer the basic psychological question that "Why particular form of education and achievement under the mainstream discourse of education is legitimized and valued in the social psychological representations?

BACKGROUND Until recently, health and health services research in Nepal has been dominated by quantitative research, often using surveys and questionnaires and providing statistical data; e. g.an overview in numbers as highlighted in a recent review of health and medical research in Nepal. 1 There are many examples of good quantitative articles in the Kathmandu University Medical Journal , one example from the most recent issue would be Aryal's survey analysis of the determinants of post-partum amenorrhea among mothers in rural Nepal. 2 Another common example of quantitative studies would be an institution-based health survey such as the one in KMUJ by Pandey and colleagues of pupils at schools in Bhaktapur. 3 A further example of quantitative research would be a community-based survey to determine the proportion of pregnant women in a certain district attending antenatal care or the proportion of Nepalese women aged 16 to 24 who used a condom during their first sexual intercourse. Quantitative research can inform us thereby of what is happening in health education and health care delivery in Nepal. It addresses the 'how many' questions, i.e. what proportion of the population does Y, likes X or prefers Z, but not the 'why' questions. Qualitative methods, on the other hand, address the 'how' and 'why' questions, for example, "why do people use too much alcohol, tobacco or other drugs?" or "Why do some doctors prescribe drugs to patients with certain symptoms whilst others are more likely to operate on people with the same conditions?" or "How do doctors cope with making medical mistakes?" 4 Qualitative methods can also help to explain quantitative data, and to refine quantitative measures. 5 Qualitative research is a highly specialized field with its own methods, ways of sampling, analysis and dissemination of findings. 'Qualitative Methods' here refers to the three most commonly used techniques in health and social care, namely: observation; in-depth interviews; and focus group discussions . Before we discuss these key qualitative methods in detail, we outline why they should be used and when it is most appropriate to use them. --- Structured or unstructured Each of these popular techniques allows the respondent to 'speak for themselves' either during an interview or focus groups or through their actions being carefully observed. They provide insight into why people do the things they do and the reasoning behind their actions. Additionally, observation may illustrate what people actually do, rather than what they say they do. Instead of ticking a box on a questionnaire which may or may not conform to their situation, they describe their feelings or actions for themselves. This enables a greater depth and quality of information to be collected, leading to a greater understanding of why people do the things they do. Qualitative methods are especially suitable for exploring new topics and obtaining insightful data on complex issues. For example, Regmi and colleagues studied the relatively new phenomenon of dating among young people in Nepal. 6,7 Since no one else has done this kind of research before there is no existing questionnaire that can be used, therefore explorative qualitative research offers insights not otherwise available. --- THE QUALITATIVE RESEARCHER The role of the researcher is quite different in qualitative research compared to quantitative research. The researcher plays a more active role in data collection and the interaction between researcher and participant can often be crucial. For example, as interviewers come into contact with a participant, they need to be aware of their own role and the effect they may have had on the participant and their answers. This is called being reflexive and is an important part of qualitative research from study design to data interpretation. The next section outlines the three key qualitative methods, each with one example. --- INTERVIEWS Face-to-face interviews are the most commonly used method, but telephone and internet interviews are increasing. 8 Interviews range from structured, through semi-structured to unstructured format, depending how strictly the schedule is followed. Structured interviews are conducted by interviewers using a questionnaire with fixed questions and sometimes pre-coded responses. Qualitative researchers typically use semi-structured or unstructured interviews. In semi-structured interviews the researcher wants to make sure a number of key questions or topics are addressed, but the order in which they are covered is determined by the interviewee, not the interviewer. The schedule or aide-mémoire may change as interviews progress in order to accommodate new topics emerging from earlier interviews. The term 'unstructured' refers to interviews in which the interviewer sets out to address a few key issues, but leaves it largely to the interviewee what is discussed and in what order. --- Example interview studies A study which lends itself to face-to-face interviews is being conducted by one of the authors at present in Nepal. She is interested in the decision made by women who recently had a baby about whether or not to attend for antenatal care. As the literature suggests that the decision to attend antenatal services is not taken by the woman alone, BS also interviewed the husband and mother-in-law where possible. She recruited women who did and did not attend through the local hospital and health post. She met the local health worker and attended a training day for health volunteers in order to gain their confidence and support. 9 Our second example is an interview study of trekking guides on sexual health issues conducted in Kathmandu, Pokhara and trekking areas in Annapurna, Langtang and Everest. 10 The in-depth interviews were conducted by PB who is a native Nepali speaker and explored the participants' attitudes towards sexual health, their reasons for having sex with tourists/trekkers, patterns of risky sexual behaviours and knowledge of sexually transmitted infections and HIV . This --- Review Article topic is particularly suitable to this method as a sensitive issue such as sexual behaviour is more likely to be disclosed in the more confidential private setting of individual faceto-face interviews. --- FOCUS GROUPS 'Focus groups' refer to group interviews with a researcher who facilitates the discussion. Focus groups are widely used to seek opinions and experiences of service users and health care providers. The underlying notion is that focus group participants interact with each other in ways that the facilitator could not predict beforehand. Usually the group members trigger off ideas amongst themselves, ask each other questions, deliberate over issues and hence reconsider their own understandings of their specific experiences or opinions. The dynamics of the focus group and interaction among participants is a key feature of the data. 11 Generally, focus groups work well for topics which people "could talk about to each other in their everyday lives-but don't". 12 Focus groups should not be seen as an opportunity to interview several people at once, as people may express different views outside the group. They often present organisational challenges in getting six to 12 or 15 participants together at one time. 11 --- Example focus group study A non-governmental organisation keen to set up health clinics in remote parts of Nepal might want to know the views of different groups in the locality -adult males, adult females, young women, young men -and hence hold separate focus groups to address such issues as perceived needs, current barriers to health care, appropriate opening hours or attitudes of health care staff. As part of a study of the influence of user demand on prescription behaviour, the Britain Nepal Medical Trust, an international NGO, conducted focus groups with patients attending health facilities to explore patients' attitudes towards drugs. 13 Another example is a focus groups run with young people to illuminate the changing attitudes in Nepal toward love, marriage, and childbearing. 14 --- OBSERVATION The traditional anthropologist would study a community or group of people by living with them for some time, asking questions about their everyday life, their religious festivals, their myths and so on. For example, a medical anthropologist might live with a community in a remote part of Nepal to study people's everyday lives, their decisionmaking around accessing health care , or the organisation of their health care system. As part of an ethnographic approach, observations are typically combined with informal conversations with those being observed and with more formal interviews. This allows the researcher to directly observe, but also seek greater explanation and understanding around events being observed. Observation can provide additional information that is not obtained through other qualitative methods. For example, an interviewer can observe facial expressions and body posture of an interviewee that can trigger doubt about what the person is stating. The interviewer can ask for clarification, or suggest that the interview is continued at a different place, away from the prying ears of partners or neighbours. BS used health workers to gain access to interviewees in Nepal, but quickly found that the positive things which interviewees said about the health worker were belied by their facial expressions and body posture. Observations can be unstructured and unobtrusive, with a researcher walking around and making occasional notes, or they can be fairly structured. For example, an educational researcher attending an epidemiology lecture for medical students at a university might note every five minutes who is paying attention, who is asleep and who making notes, etc. The qualitative aspect of such a structured set of observations lies in the fact that the observer often has to make a qualitative judgement on how to code each observation. Observers can be outsiders such as sociologists who come in to directly observe an existing situation, or participants such as patients or staff who actively participate in the situation. 4 Participant observation has the advantage that those observed are less conscious of the observation and less likely to change their behaviour. For example, a Nepalese midwifery student might observe a training event as a group member, of course taking considerations about research ethics into account. --- THE NEED FOR INTERPRETERS Research in some parts of Nepal may need interpreters. 15,16 Using translation in any kind of research raises issues of quality and accuracy. Techniques such as back translation of research tools such as questionnaires should be common practice. Qualitative research raises particular issues that need careful consideration. When an interpreter is used, this effectively adds another layer to the research interaction and the researcher needs then to reflect not only on their own effect on the research process, but the added effect of the interpreter. 15 When using semi or unstructured interviews, qualitative researchers often aim for an interview to flow like a conversation. Using an interpreter, means the flow is inevitably disrupted to allow time for translation back and forth. An alternative model of practice is to train the interpreter to carry out much of the interview themselves and only translate main points to the researcher during the interview. This allows greater flow in the interview, but removes the researcher from the interaction and depending on the particular study, may be impractical. It is important always to consider best use of interpreters in qualitative research, and the effect they have on the research process. 16 --- SAMPLING Unlike quantitative methods where the gold standard is random sampling, qualitative studies use purposive, snowball or opportunistic sampling. 8 Purposive sampling selects participants according to different characteristics such as age, gender, social class, ethnicity, risk status and so on, in order to elicit the widest range of responses. 17 Snowball sampling is a technique whereby one research participant gives the researcher the name of another, who in turn provides the next respondent. 18 Snowball sampling is useful when there is no sampling frame or the topic of study occurs infrequently. For example studying the health care needs of illegal drug users in Pokhara, a researcher may have access to one or two drug users through the police or local health clinic. If these drug users participate in the study, they are asked for contact details of other drug users who might be willing to participate. Opportunistic sampling is often used for small explorative pilot studies where easily accessible populations are targeted. The chosen method of sampling in any study will depend on the particular research question being addressed or the ease of access to a particular population at a certain time. It should be noted however, that because the sample is not randomly selected, results cannot usually be generalised to a wider population. Perhaps the solution is to use mixed methods, combining the valuable insights which typically emerge from a qualitative study, with the generalisability achieved through a quantitative study of a representative sample. 19 --- QUALITATIVE ANALYSIS Qualitative data are usually transcribed audio recordings. Transcribing is surprisingly time consuming; one hour of recording can take a skilled secretary four to five hours to transcribe word for word. Transcriptions could be accompanied by notes describing the mood of the interviewee, the setting where the interview took place, any incidents, or any other information which may help to understand the particular interview. The transcripts can then be analysed by hand or by one of a number of computer software packages. For smaller studies and novice researchers, it is advisable to do the coding by hand through reading and re-reading the transcripts using what is called an inductive thematic analysis. 17,20,21 As Subedi recently outlined: "The researcher assesses, analyses, and interprets the empirical materials … moves from field text to a research text, notes and interpretations." 22 A qualitative analysis usually leads to a number of key themes, each supported by one of more sub-themes. --- DISSEMINATION OF QUALITATIVE RESEARCH The findings of qualitative research are usually presented in text format. First the key themes are listed and explained. Usually verbatim quotes from focus group discussions or from interviews are used to illustrate the key themes distilled in the analysis. These provide rich descriptive data which generally answer the question of 'why' people act as they do. 23 When transcribing qualitative data it is obvious that many people do not speak in grammatically correct sentences. The researchers have to decide whether to present quotes verbatim or to add words which make the quote and its contents easier to understand. Table 2 provides an example of disseminating qualitative research; this particular quote relates to a sub-theme of 'safe sex' in the above mentioned study of Nepalese trekking guides. 10 This example also illustrates the common practice of labelling quotes to allow the reader to understand the context a little better without identifying the interviewee. --- CONCLUSION In this paper we have highlighted the role of qualitative research in the growing field of health and health services research in Nepal. We introduced the three main qualitative methods and discussed sampling, analysis, dissemination and the use of interpreters. Conducting rigorous qualitative research is challenging and can be resource-intensive but we hope to have shown the benefits of doing so. Qualitative research can help to develop a deeper understanding of the complex issues influencing health and health service use and give insight into the design of intervention programmes. We would like to finish off with the warning that qualitative research is not as easy as some novice researchers think it is. A doctor or nurse who conducts many clinical interviews every day might be tempted into thinking: "I do interviews with patients all the time; research interviews can't be that different." The difficulty lies in being able to step out of that health professional role, letting interviewees talk without directing the conversation and using the language of the interviewees or focus group participants rather than medical jargon.

There has been a steady growth in recent decades in Nepal in health and health services research, much of it based on quantitative research methods. Over the same period international medical journals such as The Lancet, the British Medical Journal (BMJ), The Journal of the American Medical Association (JAMA) and the Journal of Family Planning & Reproductive Health Care and many more have published methods papers outlining and promoting qualitative methods. This paper argues in favour of more high-quality qualitative research in Nepal, either on its own or as part of a mixed-methods approach, to help strengthen the country's research capacity. After outlining the reasons for using qualitative methods, we discuss the strengths and weaknesses of the three main approaches: (a) observation; (b) in-depth interviews; and (c) focus groups. We also discuss issues around sampling, analysis, presentation of findings, reflexivity of the qualitative researcher and theory building, and highlight some misconceptions about qualitative research and mistakes commonly made.

INTRODUCTION The Community Popular Opinion Leader multisite intervention trial for HIV/ sexually transmitted infection prevention was a Phase II, five-country collaboration conducted by local government health officials and researchers in China, India, Peru, Russia and Zimbabwe. 1 The intervention aimed to modify community level social norms for safer sex behaviours using peer C-POLs as change agents to endorse and diffuse safe behaviour norms. The C-POL intervention targeted at-risk communities in each country and recruited large community based samples that included thousands of individual participants with a range of risk levels and STI histories. The recently published main study results that examined both aggregated data across the five countries and within country data showed no significant intervention effects for STI and behavioural outcome measures. 2 The C-POL intervention's positive outcomes in previous successful trials 34 and the recent negative findings 2 may be accounted for by differential impacts among the highest risk community members compared to the much larger proportions of less risky participants in the samples. In addition, to ensure that only the impact of the C-POL intervention was evaluated and to meet ethical standards for conduct of research, both the control and intervention conditions received a substantial intervention comprised of STI diagnosis and antibiotic treatment services, access to condoms, STI/HIV educational materials and presentations, and ongoing monitoring through study assessments. This article presents a sub-analysis from the China site by the China site's investigator team examining whether the C-POL intervention was efficacious in reducing STIs and behavioural risks among the highest risk participants: those with a STI or reporting recent unprotected extramarital sex compared to participants without non-spousal partners or a STI. --- Hypotheses 1. We hypothesise that persons with STIs in the C-POL intervention will have significantly fewer STIs at the next follow-up compared to persons with STI in the control markets. --- 2. We hypothesise that persons reporting unprotected extramarital sex in the C-POL intervention previously will have significantly lower probability of unprotected extramarital sex at follow-ups compared to those in the control markets. --- METHODS --- Study background The National Institute of Mental Health Collaborative HIV/STD Prevention Trial was conducted from 2002-2006 in China, India, Peru, Russia and Zimbabwe. 1 The C-POL intervention is a community-level intervention that seeks to reduce STI and HIV risk by changing social norms at the community level based on Roger's theory of the diffusion of innovations. 5 The communities were selected to have sufficient baseline STI or rates of unprotected extramarital sex; population and network stability; organisational capacities ; sufficient numbers of sites for conducting a randomised trial; and sufficient distance between venues within country sites to minimise contamination. 6 In China, these factors were explicitly considered when selecting migrant market vendors over other potential study populations, including construction workers, truck drivers, injection drug users, sex workers, men who have sex with men or factory workers. 6 --- Participant recruitment and randomisation Migrant food market vendors in a large Eastern coastal city in China were selected through a two-stage enumeration and random selection process. Markets in the site typically have 80-200 stalls and 150-300 owners and/or employees. Market venues were selected based on size and sufficient distances to prevent contamination across the intervention and control markets . In total, 40 markets were selected from 95 possible markets. 6 Box 1 describes the pre-trial activities to establish venue selection and study procedures. Randomisation proceeded in two phases. First, markets were paired based on a preliminary study of STI and behavioural risk conducted prior to intervention implementation. 7 Sites were then randomly assigned to either the C-POL intervention or control condition by a Data Coordinating Center at the Research Triangle Institute with no investigator involvement. Second, a census of all stalls and employees was documented in the 40 markets. Potential participants were aged 18-49 years old and were briefly screened for eligibility based on whether they had engaged in sexual behaviour unprotected by condoms in the past 3 months. Randomisation schedules from the Data Coordinating Center identified the randomly selected stalls and a single eligible employee within each stall. Figure 1 shows the movement of participants through the trial. Box 1 Pre-trial activities to establish venue selection and study procedures. C-POL --- 1. Identify potential venues with observations and brief surveys : construction sites, factories, karaoke bars, rural villages, markets. --- 2. Select city and market venues; check for contamination possibilities across venues. --- 3. Map karaoke bars, sex work and beauty parlour establishments surrounding each market. --- 4. Survey and train pharmacists on current treatment strategies for STIs in city site. --- 5. Ethnographic studies on selection processes for community popular opinion leader ; adapt HIV educational materials; risk situations; healthy norms regarding sexuality; segments of social groups. 6. Pilot C-POL training and reunion sessions. --- 7. Establish base rates of STI in venues with large epidemiological study in 40 markets; STI testing and treatment. --- Match venues based on STI rates. --- Study procedures A detailed description of the intervention design, methodology, protocol and procedures are presented elsewhere. 189 Briefly, all participants provided written voluntary informed consent and completed baseline, 12-month and 24-month assessments that included a physical exam; STI/HIV testing from blood, urine and vaginal swabs; and syndromic STI diagnosis. The refusal rate for the initial recruitment was less than 8%. Interviewers collected self-report assessment data. Physicians conducted STI diagnosis and treatment at each assessment at a site separate from the markets. The interviewers and physicians did not know who was in the intervention or control conditions until after the 24-month assessment. Of 4512 eligible participants contacted and assessed at baseline, a total of 3914 was retained. About 20% of the population was selected to be C-POLs who were eligible to be randomly selected for study assessments at the intervention markets. --- C-POL intervention procedures C-POLs were identified through brief interviews conducted in intervention markets to identify social networks and to solicit nominations for C-POLs by market employees, managers, key informants and self-nominations. In addition, intervention trainers made repeated observations of the markets to identify popular employees. Based on Diffusion of Innovations theory, 5 approximately 20% of the market vendors regularly present in each site were selected as C-POLs and invited to attend a series of four small-group training sessions. The sessions taught skills in delivering and diffusing theory based HIV/STI prevention messages to friends and acquaintances during everyday conversations. Each C-POL practiced diffusing prevention messages daily to peers and reported the frequency of conversations weekly. After the training sessions were completed, the C-POLs attended bimonthly reunion sessions to support sustained diffusion of prevention messages and to report on diffusion of messages to peers. --- Study outcomes The primary biological outcome was incidence of any new STI, including chlamydia, gonorrhoea, syphilis, trichomonas , herpes 2) or HIV. All participants with a bacterial STI at baseline were treated. All STI diagnoses were based on laboratory confirmed testing . Two composite binary variables were constructed: one indicating new infection with at least one of these six STIs and a second indicating a new bacterial STI infection . The composite variables were set to missing if more than one-third of a participant's tests were either indeterminate or not done and there was no new positive test. The primary sexual behaviour risk outcome was defined as any unprotected sex with non-spousal partners in the past 3 months. --- Power calculations The five country C-POL trial was designed so that each country would have at least 80% power with a type I error rate of 5% to detect a 33% lower STI incidence in the intervention versus control, a participation rate of 95% and a follow-up rate of 84%. For the China site, we estimated 40 markets and 124 participants per market for statistical power. A detailed description of the power calculations for the multisite trial was published previously. 1 --- Statistical analysis We used the Transition Model 10 -one of the three commonly used longitudinal models in which the outcome variables are modelled as a function of immediate past outcome measures and explanatory variables-to test the hypothesis that intervention effects on STI incidence would be significantly greater for participants who had a STI at baseline or an incident STI during the study. We conducted the same analyses examining behavioural risk reports at prior assessments instead of STI. We used a generalised estimating equation model to test the hypothesis that participants with a STI at baseline would report significantly greater reductions in unprotected extramarital sex in the C-POL intervention compared to control. Baseline differences between intervention and control samples were tested using χ 2 and t tests for categorical and continuous variables, respectively. ORs and 95% CIs for longitudinal models are presented. All analyses were performed using SAS version 9.2. --- RESULTS Demographic and baseline characteristics are summarised in table 1. The majority of participants were women aged 31-40 years old with a mean age of 36 years. Men were 1 year older in the intervention compared to control condition . Almost 88% had less than a high school education and 9% had no education. Almost all women were married , but fewer men were married . About 87% had a regular income. No differences between the two study groups were found in education, marital status or regularly earning money. Table 2 shows a summary of behavioural and biological outcomes over time by gender and intervention condition. At baseline, only 6.5% of participants reported unprotected extramarital sex in the prior 3 months but about 20% had a STI. The disparity in baseline behavioural risk reports and STI infection rates was greatest among women: about 24% had an STI but only about 2.4% reported unprotected extramarital sex. In stark contrast, 14.4% of men had STIs of whom 11.5% reported unprotected extramarital sex. Unprotected extramarital sex and laboratory confirmed STIs decreased significantly in both intervention and control conditions over time. Results from unadjusted analysis show that the women in the C-POL intervention had a significantly lower rate of STI at 24 months compared to the control . Figure 2 presents the percentage of any new STIs by condition over time for each gender. A similar trend was found when the same analytic method was applied to the behavioural outcome of self-reported unprotected extramarital sex. At baseline, participants in both conditions had similar rates of recent unprotected extramarital sex , which was associated with STI at baseline . Participants reporting behavioural risk at the 12-month follow-up reported lower rates of risk in the intervention compared to the control at the 24-month follow-up. Finally, we examined the intervention effect on unprotected extramarital sex at the followup assessments for low versus high risk participants using model 2, which compares intervention and control participants stratified by their baseline STI status. Figure 3 presents the percentage reporting unprotected extramarital sex by study condition and baseline STI status. Among participants who had not experienced any STIs at baseline, the percentage having unprotected extramarital sex dropped in both conditions from baseline to 24-month follow-up from 5.5% to 3.8% in the intervention and from 6.5% to 3.9% in the control condition with no statistically significant differences in trends between conditions. However, among those who experienced one or more STIs at baseline, the percentage reporting unprotected extramarital sex dropped significantly from 10.7% to 4.5% in the intervention markets and non-significantly from 6.9% to 5.7% in the control markets . The control participants were more likely to engage in any unprotected extramarital sex than those in the intervention condition at 24-month follow-up. --- DISCUSSION The C-POL intervention had previously demonstrated efficacy when implemented with high-risk communities in the USA. 34 A large, multi-site intervention trial in five countries was recently completed and the main outcome analyses did not show statistically significant intervention effects on STI or condom use with non-spousal partners. 2 This paper presents results from the China site by demonstrating that the C-POL intervention did have significant impacts on STI incidence and behavioural risk among those most at risk for transmitting and acquiring STI-that is, those with diagnosed STI or unprotected sex with non-spousal partners. Investigators from the other C-POL sites may find similar results for their study samples using similar methods. Previously, this team demonstrated that the C-POL intervention implemented in China shifted attitudes and reduced stigma around STI and HIV among the migrant market vendors in the intervention markets compared to control regardless of their risk level. 11 The results suggest that the C-POL intervention succeeded in shifting social norms by diffusion of health messages through social networks. Community members in the intervention, but not the control communities, were significantly more likely to identify the intervention logo; to endorse routine clinic check-ups and condom use with non-marital partners; to talk to partners regarding sexual relationships; and to value the preservation of the health of one's community. 11 Although the results show a significant community shift in beliefs, attitudes and actions, public health officials are not impressed unless significant changes in biological outcomes are demonstrated. One of the most important findings of this study is the specific populations with which community-level interventions are most efficacious-that is, among only the highest risk community members. Most efficacy trials select the highest risk members of a population to intervene with and include in study samples, which increases the likelihood of detecting statistically significant intervention effects. Since this efficacy trial examined a communitylevel intervention, at-risk communities were selected and not solely the highest risk individuals within those communities. The community level focus of the C-POL intervention required much larger sample sizes to enable examination of intervention impacts through diffusion among the entire community. Furthermore, intensive intervention of STI clinical and education services were provided to the control condition to meet ethical standards of research conduct and to isolate the effects of the C-POL intervention compared to clinical and educational services alone. The relatively large samples in this study, and intensive intervention for controls, more closely resemble large scale effectiveness trials, which have had consistently disappointing results. 12 Although the larger samples increase statistical power, statistically significant intervention effects are often masked by the relatively large proportion of low or moderate risk population members included. In addition, many major randomised controlled trials for HIV prevention have demonstrated significant improvement in their control conditions resulting from repeated monitoring over time in addition to significant intervention for controls. Effect sizes in the control conditions can range from 15-30% reduction in risk. [13][14][15] Thus, we must both reduce risk with lower risk participants and show reductions of at least 60% with the most risky participants-the initial 30% to match the changes in the control condition and at least another 30% change to be significantly greater in the intervention compared to the control. A similar challenge to demonstrating efficacy is presented by the spontaneous clearance of some STIs without treatment . [16][17][18] Spontaneous STI clearance is particularly problematic when there are long periods between study assessments such as in the C-POL trial in which there were 12-month intervals between assessments. --- Generalisability These findings should be generalisable to other populations in China and other countries, as well as other large scale, community-wide efficacy and effectiveness trials. Lessons learnt during the implementation of the intervention and assessment could be relevant to other randomised controlled trials and to evaluations of the C-POL intervention. The sampling and assessment procedures, monitoring of the quality of implementation and the outcomes were carefully documented and the findings were quite robust. --- CONCLUSION Community-level interventions are likely to be most relevant when all members of a population are in need of intervention. When a new disease is discovered , knowledge and attitudes are the first targets of public health providers. National social marketing campaigns are examples of community-level interventions that have been highly successful in increasing knowledge and building positive attitudes. 19 Yet, for behaviour change and biological outcomes, researchers are likely to be unable to demonstrate community level change except under conditions of a generalised epidemic. 20 In this multisite trial of the C-POL intervention, intervention and control groups had similar STI incidence and behavioural risks over time when outcomes were analysed among all community members. 2 Only when we examined differential intervention impacts for the highest risk participants compared to less risky participants did we find significant intervention effects. Participant flow through the trial at each major point. Percentage of any new STIs by condition and by gender across time . Percentage of unprotected sex by treatment and by baseline selected sexually transmitted infections. --- Summary of behavioral and biological outcomes by condition

Objectives-A community level randomised controlled trial of a Community Popular Opinion Leader (C-POL) intervention to reduce bacterial and viral sexually transmitted infections (STIs) and unprotected extramarital sex was carried out over 2 years in five countries. The main study results did not find significant intervention effects. This paper presents a sub-analysis examining the differential intervention impacts among high-risk and low-risk participants in the China site. -From 2002 -2006, 3912 migrant market vendors aged 18 and 49 years were recruited at an urban site in China. Markets were randomly assigned to the C-POL intervention (N=20

Introduction Structural racism may play a role in observed disparities in adverse birth outcomes by different racial groups in the United States . Structural racism, described as "the larger system of policies, practices, ideologies, and institutions that foster racial inequality by creating differential access to resources and opportunities" , is reported to be a major driver of the uneven distribution of air pollution across different racial and ethnic minoritized groups in the U.S in both the short and long term . The impact of these disparities may be even greater when consideration is given to the historical context of structural racism and corresponding data are incorporated in study designs. A comprehensive approach to understanding factors that may contribute to disparities in adverse birth outcomes will be crucial for the development of effective intervention strategies in the U.S. Race, or the division of the population by racially defined subgroups, has been described as a socially constructed undertaking that does not have a biological basis and may not be causally associated with disparities observed in key health outcomes such as adverse birth outcomes. Instead, policies and practices such as redlining, a government-backed practice in the 1930s that led to the classification of neighborhoods with a high percentage of Black individuals and immigrants as "hazardous" and consequently, considered a low investment priority by lenders , may be a key component driving observed differences in birth outcomes. Maps created by the Home Owners' Loan Corporation , a government-sponsored entity intended to assist homeowners at risk of foreclosure, informed this practice. The HOLC utilized the following system to classify neighborhoods: Green = "Best"; Blue = "Still Desirable"; Yellow = "Declining"; Red = "Hazardous" . This practice led to disinvestment in neighborhoods that were redlined or classified as "red", which subsequently led to major declines in neighborhood quality and lost opportunities for home ownership and wealth accumulation . With neighborhood decline, land value may have been affected and this may have created conditions that were favorable to industries that produce pollution . In addition, surrounding areas near redlined neighborhoods were zoned as "industrial", thereby allowing for the siting of major pollution producing sources . Despite officially ending in 1968 , redlining's impact has persisted and links exist between redlined neighborhoods and the siting of present-day pollution sources such as roadways, industrial facilities and power plants . Additionally, environmental justice scholars and activists in the United States have documented that exposure to air pollution and other adverse environmental conditions are often disproportionately higher among racially minoritized groups , despite lower consumption of goods and services associated with the production of air pollution . In addition, associations have been examined between redlining and current day air pollution levels in cities across the U.S . Although air pollution concentrations have declined over time, disparities have persisted . Exposure remains a problem from an environmental justice point of view because even levels below health-based regulatory standards are associated with adverse health outcomes . Despite increasing evidence that redlining and other important historical policies are associated with disparities in health, educational and economic outcomes, particularly given the pervasive and persistent nature of structural racism, very few studies focus on intergenerational associations . In environmental epidemiology, the lack of widespread use of methodological approaches that incorporate intergenerational exposures and measures of structural racism may partly be due to logistical complexity in finding and linking datasets across generations. The first goal of this article is to provide context about studying racial disparities, structural racism and environmental epidemiology. Then, we review evidence on air pollution associations with adverse birth outcomes. In addition, we discuss a conceptual framework that informs the rationale behind our recommendation to evaluate associations between intergenerational exposure to air pollution and adverse birth outcomes. We also identify examples of existing historical data from multiple sources to evaluate these intergenerational associations and to identify factors, including structurally racist policies that differ across communities, which may provide opportunities for intervention. In this manuscript, we focus on non-Hispanic Black and non-Hispanic White persons because the largest disparities in adverse birth outcomes such as preterm birth, low birth weight and infant mortality are found among non-Hispanic Black individuals compared to non-Hispanic White individuals. However, the methods here would be applicable and equally important in studying health disparities affecting other racially minoritized groups. Ultimately, the intended purpose of this manuscript is to provide an overview of the current state of structural racism in environmental epidemiology, introduce a conceptual framework of intergenerational exposures, and offer data considerations when taking an intergenerational approach in environmental epidemiological research. --- Studying race and structural racism and current practice in environmental epidemiology Many studies evaluating differences in exposure to environmental pollution or adverse birth outcomes compare a racially minoritized group or groups to non-Hispanic White persons as the reference group . Although useful for highlighting differences by racially minoritized groups with nonminoritized groups, continued comparisons of health outcomes by race without a discussion of the myriad of factors that race represents may lead to incorrect conclusions or perpetuation of the idea that observed disparities may be due to biological differences and thus limit potential opportunities for intervention. Instead, health disparities by racially classified categories are likely in many cases to result from longstanding exclusionary and racist systems and practices . Therefore, using rigorous study designs that illuminate the impact of these systems will ultimately inform policies needed to change these systems and will likely advance the field in meaningful ways . The use of multilevel and multidimensional models ; mixed data ; and life course and longitudinal analyses, have been recommended as a way to move health disparities research forward . Specific to environmental epidemiology, Hicken and colleagues recently reviewed literature on race variables as an effect modifier of the air pollution and mortality association in the U.S. as a way of illustrating current practice and areas for improvement on the conceptualization of race in the environmental health literature . They found a pattern in which the reviewed studies provided little discussion conceptualizing and justifying the use of race in the study, and few discussions of reasons behind any observed effect modification by race. Indeed, the authors suggest that frameworks that include social meaning of race and positioning of race within structural and cultural systems are lacking in the air pollution and mortality literature, and call for greater "rigor and intention" in the way race is considered in such studies. One example they provide is looking at the role of "place within race categories" as a way of identifying features that differ by place and their role in exposure/outcome differences. They also call for framing of research questions in a way that is grounded in the humanities literature and has clear policy and intervention implications. Similar calls were given in a recent editorial emphasizing interest in publishing research on environmental racism in environmental health sciences research and introducing new guidelines to enhance the rigor of such research . --- Air pollution and adverse birth outcomes-Conceptual framework for intergenerational studies Several individual epidemiologic studies and meta-analyses support associations between higher levels of air pollution and adverse birth outcomes . Associations are hypothesized to occur via multiple mechanisms, including oxidative stress, inflammation, and endothelial dysfunction . Air pollution is a risk factor for adverse birth outcomes, such as low birth weight and preterm birth . Such birth outcomes occur at higher rates among non-Hispanic Black compared to non-Hispanic White women , and persist even after accounting for important factors such as education and socioeconomic status . Reasons for this finding remain unclear, but long-term air pollution exposure, either historical or both historical and ongoing exposure to air pollution, may partly contribute. It is thus crucial for the underlying factors, including systemic factors, which affect exposure to air pollution to be investigated in any serious attempts to address health disparities in the U.S. Previous and ongoing stressors may operate via the "weathering hypothesis", which posits that combined and repeated exposure to social and environmental stressors over the life course may lead to accelerated declines in health and may help explain the mechanisms involved in the cumulative potential of harmful exposures. Although weathering can affect anyone who is continually exposed to stressors, including non-Hispanic White individuals, the effects are more pronounced among non-Hispanic Black individuals . An important consideration for understanding the broader systems and processes that may contribute to persistent disparities in adverse health outcomes between non-Hispanic White and non-Hispanic Black individuals is that weathering does not account for environmental exposures or other stressors across generations, which may be important in understanding cumulative risks. Extending the weathering hypothesis by looking beyond an individual's life course to include intergenerational exposures may be important in understanding risks in subsequent generations among non-Hispanic Black families. Potential links between air pollution exposure and adverse birth outcomes may lead to the perpetuation of increased risks among subsequent generations and contribute to sustained disparities across generations. For example, low birth weight and preterm birth rates are higher among individuals who had lower than normal weight at birth or who were born prior to 37 weeks of gestation , and intergenerational associations have been reported . Numerous studies conducted to date have evaluated long-term or pregnancy-wide exposure to air pollution and birth outcomes within a single generation . Others have evaluated pre-and postnatal exposures on child health outcomes . However, exposure to air pollution may have health impacts across generations that may not be fully accounted for using only prenatal exposures or exposures within a given generation. Examining intergenerational air pollution exposure may be helpful for understanding racial and ethnic health disparities in subsequent generations and help account for yet unexplained racial and ethnic differences in adverse birth outcomes. Associations may operate via multiple pathways that include increased potential for morbidity during the life course, and reduced socioeconomic advancement opportunities among families. For example, Sharkey found that individuals whose childhood residence was in the lowest quartile of income, which are disproportionately Black individuals, tended to live in similar neighborhoods as adults, suggesting that neighborhood disadvantage can be "inherited" across generations . Neighborhood income and racial composition, as well, are associated with higher levels of air pollution as well as potentially toxic components of pollution , and these disparities can persist over time and contribute to illness. Morbidity during the life course may subsequently contribute to lower socio-economic status of the mother and other caregivers at the birth of the next generation, and increased risks of adverse birth and child health outcomes in subsequent generations . In addition, multiple pathways may involve direct and indirect effects, including epigenetic changes in germline cells that have been reported for air pollution and other environmental Examples of data sources in the U.S. for examining intergenerational associations --- Health outcomes Researchers interested in conducting studies to examine intergenerational associations can benefit from using existing data resources, such as linked birth registries data or existing cohorts. In other instances, opportunities may exist to work with state level or federal agencies to obtain restricted identifiable data. To date, a number of studies have reported analyses using intergenerational datasets; these datasets may provide opportunities to investigate the role of structural racism on environmental exposures and adverse health outcomes. One example of such a dataset that provides potential opportunities to investigate intergenerational associations is the South Carolina birth outcome data and multigenerational dataset. This is an integrative dataset that includes data obtained from singleton birth certificates for the state of South Carolina from 1989 to 2014 . Birth data are geocoded at the census tract level and linked along the maternal line and span three generations. In the U.S., census tracts are statistical subdivisions of varying geographical area that include an average population size of 4,000 inhabitants, range 1,200-8,000 inhabitants . The resulting structure of the data includes grandmothers , mothers , and G2's children within the same family. Birth outcomes are available in this dataset as this information is routinely collected at birth and recorded on birth certificates. A description of this multigenerational dataset for 1989-2014 has been previously reported and the dataset includes 60,258 linked families , with 32,046 for non-Hispanic Black infants and 28,212 for non-Hispanic White infants. In addition to the South Carolina birth outcome data and multigenerational dataset, other states have produced similar multigenerational datasets. Two of those intergenerational datasets cover periods in the early 1980s for which census tract level air pollution estimates are available for at least a part of the older generation life span. First, using data from the State of Washington, a database was created of vital statistics data that were subsequently linked to birth certificate data of mothers born in Washington State . The resulting dataset included 46,000 births that occurred between 1987 and 1995 to mothers who were born between 1949 and 79 and who were of the following racial and ethnic groups: African American, Hispanic, Native American and White individuals. Second, an intergenerational dataset was generated by linking the State of Virginia live birth data from 2005 to 2009 to mothers born during 1960-1997. This overall dataset included 170,624 records, but a subset that was restricted to non-Hispanic Black and non-Hispanic White firstborn singleton infants included 69,702 records . In addition, a number of multigenerational studies , have been established and may serve as either potential data sources or methodological inspiration for future studies. Furthermore, future studies exploring the logistical complexity of assembling a cohort of three generations could incorporate the design of the Nurses' Health Study II . In the Nurses' Health Study II, both historical and current data were collected, thereby enabling the utilization of data on three generations. Specifically, enrolled participants represent the second generation or G2 and information was gathered on G1 and G3. These datasets point to existing opportunities and highlight methodology that may be applied to different states with other characteristics of interest. The intergenerational data format described here offers an advantage over single point-in-time measures of structural racism or environmental exposures and may elucidate pathways leading to adverse health outcomes. Using intergenerational information may inform whether maternal exposure to structural racism impacts health outcomes through shared exposure or through biological pathways . An important consideration is the statistical approach that would be appropriate for intergenerational studies and effectively account for both historical and ongoing exposures. Mediation analyses have been previously used to understand direct and indirect effects in studies examining smoking transmission across three generations , and such an approach would be useful for studies that might be conducted with the data described here. For example, in one study that examined smoking transmission across three generations over a 45-year period, the study team utilized a multistep approach . First, logistic regression models were utilized to evaluate direct relationships for each of the following pairs: associations between the first generation and second generation smoking, G2 and the third generation smoking, and G1 and G3 smoking. Second, the investigators examined whether transmission of smoking between the grandparents and grandchildren was mediated by G2 smoking using a Sobel test that incorporated standardized coefficients resulting from the logistic regression models. Another approach could be the use of multilevel logistic regression to account for clustering resulting from the use of census tract air pollution data. Pollutant may be represented by typologies that could be used to create a composite measure of air pollution for G1 and G2. Based on the distribution of the data, a cutoff value could be used to create high and low categories for G1 and G2, separately. Combining the resulting data would form the following categories G1 high/G2 high, G1 high/G2 low, G1 low/G2 high, and G1 low/G2 low . In addition, the large dataset afforded by using available birth data over a long period would allow for examination of the joint effects of G1 and G2 prenatal exposure to air pollution on adverse birth outcomes among G3. --- Air pollution The United States Environmental Protection Agency routinely collects data on air pollution. The U.S EPA compiles data from states reporting hourly air pollution concentrations at approximately 5,000 monitors across the country and these historical data are publicly available to be linked to existing or even newly created intergenerational datasets. Additionally, other groups have compiled EPA and other data using modeling. For example, the Center for Air, Climate, and Energy Solutions is a research center staffed by researchers from multiple disciplines and institutions. The CACES website has air pollution data that can be easily requested in a few steps, the process does not require a lengthy processing time, and the data timeframe supports examination of intergenerational exposures. Data availability from CACES varies across pollutants, but the earliest year is 1979, which would make this data source compatible with studies interested in at least two generations. This data source is particularly useful for studies interested in average exposure as the data are available as annual estimates. CACES focuses on key themes about air pollution, including regional differences of air pollution. Additionally, CACES incorporates measurement and modeling approaches that are helpful for investigating both spatial and temporal differences in exposure and associated outcomes. Information on CACES and their air pollution data may be accessed at the following website . Data available at CACES can be used to estimate annual air pollution exposure for individual participants at the census tract level. CACES produces air pollution estimates for criteria air pollutants, the air pollutants for which the U.S. EPA has established standards to protect human health. Criteria pollutants include carbon monoxide , ozone , particulate matter ≤10, ≤2.5 μm in aerodynamic diameter , nitrogen dioxide , and sulfur dioxide . According to CACES, average annual estimates, except for ozone, were derived using models that incorporate data from U.S. EPA regulatory monitors, satellite-derived air pollution estimates of air pollution to account for locations with missing data, and information on land use. For ozone, estimates are based on the daily maximum 8-hour moving average concentrations for the period May through September , which corresponds to the warmer months of the year . Figure 2 shows a sample of air pollution data obtained from CACES using PM 10 from 1989 to 2015 for two census tracts in South Carolina. The census tracts shown are in Allendale [Federal Information Processing Standards code: 45005970200] and Oconee . These census tracts were selected to compare a census tract with a high percent of non-Hispanic White population to a high percent non-Hispanic Black population. According to 2010 census data available on Social Explorer , which provides access to current and historical U.S census data, the population for the census tract in Allendale was 78.4% Black and the population for the census tract in Oconee was 88.4% White. These data can be generated for any census tract in the United States and annual concentrations of air pollutants can be compared over time. Data are available for O 3 , SO 2 , NO 2 , for 1979-2017, PM 10 for 1988-2017, CO 1990-2017 and PM 2.5 from 1999 to 2017. Another source of historical air pollution data is from the Atmospheric Composition Analysis Group at Washington University in St. Louis. This source has monthly and annual measures for fine particulate matter which are available beginning from 1981 . This data source is particularly useful for studies examining time varying exposures and outcomes that might be impacted by windows of susceptibility that vary over time because it includes data at a finer temporal scale compared to the CACES data . The spatial resolution is 1 km by 1 km grid squares. Other sources of historical air pollution data that allow for evaluating time varying exposures and windows of susceptibility over the gestational period are also publicly available. For example, using data from the U.S EPA Air Quality System and other publicly available data sources including meteorological and land use data, daily air pollution data for different periods of gestation can be estimated at the census tract level based on methods used in long term studies of air pollution . Data from the EPA AQS can be used to estimate daily prenatal exposure . This provides a distinct advantage in that it accounts for variations in pollutant levels at different points in the year as well as potential windows of susceptibility during pregnancy, such as trimester specific exposure, or exposures from shorter times frames, such as weeks or days before delivery . Methods to assign exposure estimates of census tract-level concentrations are described in Di et al. . In addition to recommendations to use multilevel data and incorporate the life course and time, and, as we have discussed here, across generations, a major advantage of using such datasets, especially birth registry data, is that they are sufficiently large and adequately powered to allow for the evaluation of effect modification and subgroup analyses. Although the historical air pollution data sources listed above are, in some cases, readily available, they are not without limitation. First, attempting to document air pollution exposure over long periods of time is a challenge. Challenges include availability of data to cover the entire timeframe of interest or if using data from multiple sources, potential differences in measurement methodology might be an issue. Second, an important limitation is that air pollution estimates using only data from the U.S EPA may be based on monitors that are placed in a central location and the same estimates may be assigned to individuals who live near a sited monitor and to individuals who live miles away from a monitor if using census tract level estimates. Modelling, which takes into account other information such as land use and satellite data , is used in some cases and may improve estimates. However, historical data do not take into account daily activity patterns as has been done in prospective studies or adjust for time that individuals may be at locations other than at home . In other cases, kriging, an interpolation method that takes into account available data, including address information and distance to available monitors, to improve air pollution estimates, may be used . Third, these limitations may influence study results particularly if the misclassification varies by groups and caution should be taken when interpreting and discussing results. Nevertheless, utilizing the above data sources will provide unique opportunities to examine factors that may contribute to health disparities. Equally important, studies using data sources listed above may be important in identifying gaps for future prospective studies where air pollution exposure assessments may be improved. --- Summary Combining long-term air pollution and other environmental data and linked multi-generational health outcome data is essential for environmental epidemiologists interested in estimating associations between air pollution and intergenerational social, health and economic disparities. Further, evaluating pathways by which environmental and structural racism could play a role in these associations, and incorporating available metrics that reflect the historical underpinnings of the systems in our society that result in unfair exposures and health burdens, has been called for on multiple fronts. Insights obtained from this kind of study may be useful for understanding persistently observed disparities in adverse birth outcomes, and intervening to eliminate them. --- Discussion A fundamental component to advancing health equity is a need to design rigorous research studies that quantify structural racism such that it accurately captures the multi-dimensional way it affects the lives of minoritized individuals. The data sources presented here provide opportunities to examine associations between exposure to air pollution, which may be driven by structurally racist policies and practices, and adverse birth outcomes. These opportunities can be expanded to include maternal outcomes as well as other outcomes that have a persistent disparity between minoritized and non-minoritized groups. In addition, these steps highlight a need to focus less on individual-level factors associated with adverse outcomes and more on systematic attempts at correcting the systems that enable structurally racist practices to flourish. Consistent with recent calls to enhance the rigor and care with which hypotheses around race as an effect modifier in environmental epidemiology are developed and findings explained , we suggest that researchers can tap into existing data and frame their research in a way that points toward policy and interventions to reduce health disparities. This suggestion is consistent with a recent exhortation to move beyond characterizing race/ethnic disparities in air pollution to addressing them . Others have called for combining antiracist principles with those of community-based participatory research, an approach that emphasizes community knowledge and equal partnerships with academics . Collaborating in research with affected communities and co-analysis of collected data will lead to enhanced insights into causal mechanisms being explored . Thus, in addition to tapping into existing data sources and applying methods such as those we illustrate here, we advocate for engaging community members and practitioners in evaluating and framing research questions. Such approaches may be useful for providing a comprehensive understanding of the cumulative nature of environmental exposures and provide actionable information on how public health policy and action can be tailored. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Structurally racist policies and practices of the past are likely to be a driving factor in current day differences in exposure to air pollution and may contribute to observed racial and ethnic disparities in adverse birth outcomes in the United States (U.S.). Non-Hispanic Black women in the U.S. experience poorer health outcomes during pregnancy and throughout the life course compared to non-Hispanic White women. This disparity holds even among non-Hispanic Black women with higher socioeconomic status. Reasons for this finding remain unclear, but long-term environmental exposure, either historical exposure or both historical and ongoing exposure, may contribute. Structural racism likely contributes to differences in social and environmental exposures by race in the U.S. context, and these differences can affect health and wellbeing across multiple generations. In this paper, we briefly review current knowledge and recommendations on the study of race and structural racism in environmental epidemiology, specifically focused on air pollution. We describe a conceptual framework and opportunities to use existing historical data from multiple sources to evaluate multi-generational influences of air pollution and structurally racist policies on birth and other relevant health outcomes. Increased analysis of this kind of data is critical for our understanding of structural racism's impact on multiple factors, including environmental exposures and adverse health outcomes, and identifying how past policies can have enduring legacies in shaping health and well-being in the present day. The intended purpose of this manuscript is to provide an overview of the widespread reach of structural racism, its potential association with health disparities and a comprehensive approach in environmental health research that may be required to study and address these problems in the U.S. The collaborative and methodological approaches we highlight have the potential to identify modifiable factors that can lead to effective interventions for health equity.

Background Each year, there are approximately 350 new cases of human immunodeficiency virus infections within members of the U.S. military; with those most affected being younger, Black, and men who have sex with men [1][2][3][4][5][6]. When taken as prescribed, pre-exposure prophylaxis effectively prevents HIV infection [7][8][9][10]. However, the percentage of at-risk military members taking PrEP measures is only 16%, and even lower in members of color [2]. Factors that contribute to the low PrEP engagement rate include access pathways, delivery mechanisms, and dosing methods impacted by geographic, psychological, and adherence barriers [11][12][13][14][15][16][17][18]. For military members, PrEP availability largely depends on geographic proximity to large medical facilities with specialty services, as evidenced by 41% of all military PrEP prescriptions originating from military medical centers in three locations in the U.S. Additionally, 60% of all military PrEP prescriptions also occur only after consultation with an infectious disease specialist [2]. Features of PrEP delivery programs suitable for military MSM remain unknown. Health services designed around preferences in terms of product type, delivery method, and location settings have been shown to produce improved outcomes and retention in care [19][20][21]. Stated preference methods, such as conjoint analysis, quantify preference data of new market entrants and product attributes [16,17,[22][23][24][25][26]. In conjoint analysis, products/programs are viewed as a composition of various attributes that possess a certain amount of value determined by preference. By quantifying these part-worth utility scores for preferred attributes, these scores can then be entered into market simulation models to predict how respondents might respond to any potential combination of attribute levels [16,17,[22][23][24][25][26]. We used conjoint analysis to identify preferred attributes that are most influential to at-risk U.S. military MSM's decision to take PrEP within the military healthcare system. --- Methods --- Data collection A convenience sample of self-reported HIV-negative, U.S. military MSM and trans-individuals were recruited between March and April 2020 through a closed Facebook group with over 7000 LGBT U.S. military members. The group administrators placed monthly advertisements describing the study on the group's main forum. Those interested could click on a link to access an online study disclosure form with a 'click to consent' procedure. An option to provide an e-mail address that was not linked to survey responses was offered to participants who opted to receive $5 compensation for questionnaire completion. The study was approved by the Yale University Institutional Review Board. To collect and quantify respondent preference data, an adaptive choice-based conjoint survey instrument was developed based on a starting set of PrEP program attributes resulting from review of the literature of previous PrEP preference conjoint experiments, and then refined through in-depth, qualitative interviews between PrEP experts and U.S. military MSM [2-5, 11, 12, 14, 15, 27-39]. The final survey instrument focuses on modifiable PrEP program characteristics, and consists of the following five attributes: dosing method [daily oral tablet, on-demand tablet regimen , rectal douche , injection , implant ], provider type , visit location , dispensing venue , and lab evaluation . The survey was piloted by the author using a convenience sample of eleven military MSM members within the targeted social media group for concept testing, and the descriptions and wording of three attribute categories and two attribute level choices were revised for clarification based on feedback. Additional figure shows a sample item of the conjoint survey, and Table 1 describes the program attributes in the survey. Additional information that we collected include age, race, ethnicity, rank type , military branch, geographic region, PrEP experience ["Have you ever used PrEP ?"], depressive symptoms with the Patient Health Questionnaire-2 [40,41], and the HIV Incidence Risk Index for MSM , which identifies MSM at high risk [42]. Measures to explore levels of satisfaction with a current level of HIV protection and disclosure discomfort within interactions with a primary care provider were also collected. --- Data analysis The final survey was loaded into Lighthouse Studio 9, and an experimental design module was used to pretest the design with 500 simulated respondents for optimal choice task configuration. The final design produced a survey where each level within an attribute was seen at least three times per respondent; achieving a high degree of precision at the individual level with a standard of error of < 0.03 and all efficiencies reporting at 1.00 [43]. Respondent enrolment and exclusion are shown in CONSORT style in Fig. 1. To ensure data integrity and eliminate random or duplicate responders, security features within the Sawtooth software and servers recognize returning study participants through the use of internet browser cookies and IP addresses. It also prevents repeated or duplicate attempts to retake the survey [44]. Since extensive pilot testing required at least 10 to 15 min, responses completed in less than 10 min were excluded. Furthermore, the root likelihood fit statistic for each respondent was analyzed to evaluate within-respondent choice consistency. RLH, which has a probability value from 0 to 1.0, was used to discriminate between respondents who answered choice-questions consistently or randomly [45]. The survey design was tested by 1000 computer-generated mock respondents to determine the median RLH for "random responders" at the 95% percentile . Survey respondents with an RLH below this score were excluded, as the inclusion of "random responders" can affect the calculation of preference scores and participation rates [45]. For conjoint analyses, a hierarchical bayes procedure was used to estimate part-worth utility scores on an individual level for accuracy and efficiency [46,47], and was used to analyze the PWUS of the aggregated sample across all 16 attribute levels. The resulting PWUS of the levels under each attribute category are zero-centered; meaning that the sum of the level scores under each attribute category equal to zero. Scores that are further away from zero indicate a Home-based mail-in kit means that you prefer to receive a home-based lab testing kit in the mail. You will provide selfcollected, small samples of blood and urine and mail the kit back to the laboratory for evaluation. Your PrEP provider would then see the lab results after processing PrEP dispensing venue stronger positive or negative preference for the level choice in relation to the other level choices under the same attribute [39,43,47]. After identifying each attribute level PWUS, the attribute relative importance scores was calculated to characterize the magnitude of influence that each attribute category has on the respondent preference decision-making. The RIS for this study was calculated by dividing the range of PWUS for levels under each attribute by the sum of the ranges, and then multiplying by 100 [48,49]. For example, an attribute RIS at 45% denotes that 45% of an individual's decision making for program engagement is influenced by preferences within that attribute category. The PWUS were then used to predict the share of preference among eight hypothetical PrEP program scenarios. PrEP program scenarios were configured after a variety of currently available or hypothetical PrEP program models. For this study, participation rates for these PrEP scenarios were generated using the randomized first choice model, in which PWUS are summed across the levels corresponding to each option, and then exponentiated and rescaled, so they sum to 100 [48,49]. This approach is based on the assumption that respondents or consumers will prefer a product with the highest composite utility adjusting for both attribute and program variability [48]. The randomized choice model accounts for variation in each participant's total utility for each option and error in point estimates of the utility, and has been shown to have better predictive ability than other shares of preference models [49]. All data analyses were performed using XLSTAT and Sawtooth Lighthouse Studio 9.0. --- Results --- --- Relative importance and part-worth utility scores Table 3 shows the relative importance scores of the five attributes and Table 4 shows the part-worth utility scores for each attribute level. Overall, the dosing method was the most important attribute, regardless of PrEP experience. For participants reporting PrEP experience, the daily tablet was the most preferred dosing method option, followed by the on-demand tablet regimen. For those with no previous experience with PrEP, the bi-monthly PrEP injection was the most preferred dosing method option, with the yearly implant and daily tablet preferred but to a lesser degree. Among the remaining attributes, both groups generally preferred the option to see a military healthcare provider, to use a smartphone to conduct the PrEP visit, and to utilize an on-base location for laboratory evaluation and receipt of medication. Thus, respondents with no PrEP experience are less likely to select "None" compared to PrEPexperienced individuals, and are more likely to initiate PrEP regardless of program configuration. The "None" parameter represents the positive or negative magnitude in which a respondent is likely to select "None" PrEP pre-exposure prophylaxis --- Preferences for PrEP program scenarios --- Discussion The current study demonstrated an overall preference for daily tablet PrEP services at an on-base location vs. civilian and off-base settings, yet those with no previous PrEP experience have a stronger preference for longeracting injectables and implants. Additionally, over half of all respondents screened positive for depressive symptoms, the majority of respondents engage in risk behaviors that categorize them as having a high risk for acquiring HIV. Over one-third of respondents reported discomfort in discussing their sex life with PCP. With a growing body of literature suggesting a link between depression and sexual risk behaviors among MSM [50][51][52], it may be beneficial for PrEP-prescribing providers to provide PrEP clinics that are sensitive and inclusive to sexual identity and to remain vigilant to address factors related to mental and sexual health specific to MSM. In the conjoint experiment, dosing method attribute was the most critical and influential preference factor within a PrEP delivery program, with a strong overall preference for a daily tablet among the total sample. Among those with no previous PrEP experience, a dominant preference for PrEP injectables and implants suggests that a demand for these longer-acting PrEP methods exists within this population if these alternatives become available. The long-acting injectable cabotegravir for PrEP has demonstrated superiority over oral tablet PrEP [53,54], and could become further prioritized by future users as efficacy data becomes more widely known. The MK-8591-eluting PrEP implants have also achieved promising preliminary results [55], with clear benefit over daily tablet for individuals with adherence concerns or an unpredictable work schedule [56]. The expeditionary nature often entails military members to relocate, deploy, or miss regular follow-up appointments [57,58]. Therefore, availability of longer-acting PrEP modalities should be an important component of the future PrEP programs for military members. The current study confirmed a preference to see a military provider for PrEP services, highlighting the important role of military healthcare providers. However, a survey of military health care providers regarding PrEP knowledge and prescription habits revealed that 49% rated their knowledge as poor and only 29% had ever prescribed it [2]. Additionally, most military members receive their PrEP prescription only after seeing an infectious disease specialist [2], suggesting that military PCP may feel uncomfortable prescribing PrEP. This lower level of PrEP knowledge and prescription practice may contribute to the heterogenous nature of PrEP availability that currently seems dependent on a military member's geographic location [2], and could explain the statistically significant difference in respondents' regions of station within this study when stratified by PrEP experience. An increase in PrEP knowledge has been associated with an increase in prescribing habits [59], again suggesting military PCP with necessary training and resources to comfortably prescribe PrEP may help military members engage in a wider availability of PrEP services without the extra step of a referral to an infectious 5 PrEP pre-exposure prophylaxis disease specialist. Further research is needed to explore the preference for a military provider within this context, and how this preference can best be leveraged to improve PrEP implementation. In the current study, respondents with no previous PrEP experience reported fewer male sexual partners and fewer HIV-positive partners within the last six months, and more likely to report being unsatisfied with their current level of HIV protection. Given that sexual contact with men and condomless sex have been found to be the most common indications for initiating PrEP among military MSM [2], further studies will need to explore what type of PrEP programs could best accommodate the desire for more intimacy with male partners without using condoms. This study has limitations. First, this study utilized self-report from a convenience sample recruited from an online social media group comprised of U.S. military members who identify themselves as LGBT. As a result, we could not verify actual eligibility . Having said that, existing literature that compared MSM recruitment via online methods versus in-person had found similar samples of HIV/STI prevalence and HIV-testing patterns among MSM [60,61]. Also, our findings may not be generalizable to atrisk military members who have sex with men but do not identify themselves as being MSM or LGBT. We also excluded a large number of respondents based on RLH cut-offs. There is a rising trend of "random responders" and "survey-bots" that attempt duplicate submission of surveys that provide financial compensation and can impact preference data if included within final the analyses [45]. These RLH standards follow evidencebased consistency cut-offs to eliminate these "random responders", but could nonetheless introduce sample selection bias. Finally, reporting preferences is distinct from actual behavior. The final evidence will come only from future studies when the preference data are implemented into practice. --- Conclusion In military members with a high risk of acquiring HIV, PrEP programs with the following features are preferred: daily tablet, injection or implant, medical visit provided by a military healthcare provider through a telehealth smartphone app, and on-base locations to provide laboratory samples and to receive PrEP medication. --- --- Abbreviations ACBC, adaptive choice-based conjoint; AD, active-duty; APO, Army Post Office; Cis, cisgender; CONSORT, consolidated standards of reporting trials; CRAS, condomless receptive anal sex; HB, hierarchical bayes; HIRI-MSM, HIV incidence risk index for men who have sex with men; HIV, human immunodeficiency virus; IP, internet protocol; LGBT, lesbiangaybi & transgender; MSM, men who have sex with men; OCONUS, Outside Continental United States; PHQ2, patient health questionnaire-2; PrEP, pre-exposure prophylaxis; PCP, primary care provider; PWUS, part-worth utility scores; RIS, relative importance score; RLH, root likelihood --- --- Additional file 1. Authors' contributions JIG, AD, and FLA developed the attributes and analyzed and interpreted the preference data regarding PrEP delivery program preferences, as well as demographical descriptive statistics. JIG and DV were major contributors in writing the manuscript. All authors read and approved the final manuscript. --- --- --- --- Competing interests The authors declare that they have no competing interests.

Background: Pre-exposure prophylaxis (PrEP) prevents human immunodeficiency virus (HIV) infection, but its use remains low among U.S. military men who have sex with men (MSM), likely due to mis-matching with personal preferences. We conducted a study to characterize preferences to PrEP measures within this population. Methods: HIV-negative military MSM were recruited through a closed, Lesbian, Gay, Bisexual, and Transgendered (LGBT) military social media group. The survey was anonymous, and consisted of five experimentally varied attributes in service delivery: dosing method, provider type, visit location, lab work evaluation location, and dispensing venue. Relative importance and part-worth utility scores were generated using hierarchical bayes (HB) estimation, and the randomized first choice model was used to examine participation interest across eight possible PrEP program scenarios. Results: A total of 429 participants completed the survey. Among the eight scenarios with varying attributes, the most preferred scenario featured a daily tablet, PrEP injection or implant, along with a military provider, smartphone/telehealth visit, and on-base locations for lab evaluation and medication pick-up. The results also emphasized the importance for providers to be familiar with PrEP prescription knowledge, and to provide interactions sensitive to sexual identity and mental health.A PrEP program consisting of daily tablet is preferred in military healthcare settings is preferred. Longacting implants and injections are also desired.

Introduction Population ageing within global demography is one of the most significant social transformations of the twenty-first century. Projections show that by 2050, the proportion of people over 65 years will increase from 1 in 11 to 1 in 6 [1]. The diversity of the ageing population is a significant factor in understanding any challenges and opportunities in the way in which we respond to demographic changes [2]. While lesbian, gay, bisexual, transgender and gender diverse people face similar challenges to their heterosexual and cisgender peers in later life, an established evidence base demonstrates specific health, social and structural inequalities for LGBT+ older people [3][4][5]. sign to signal inclusion of the wide diversity of sexual and gender identities, unless the paper being cited is focused on specific identities). Inequalities are compounded by the cumulative effects of lifelong exposure to prejudice, discrimination, criminalisation [6] and environmental factors nuanced by a wide range of intersecting identities, including socio-economic status, culture, race and ethnicity, disability and religion [7]. These unique circumstances impacting on ageing experience for LGBT+ people pose risks linked to minority stress [8] and stress adaptation in later life [9]. Minority stress posits that in comparison to the heterosexual and cisgender community the stigma and resulting discrimination experiencd by sexual and gender minority people creates a multitude of stressors that heighten their risk of negative physial and mental health outcomes. Some older LGBT+ people are more successful than others in adapting and coping with ageing: those with strong psychological and social resources are likely to enjoy better health and practice more health promotion behaviours [3,10]. However, the design of effective interventions to promote such positive adaptations is not yet well understood. In addition, with the increased documented exposure of LGBT+ people to social and culturally embedded discrimination, there has been growing research interest in the role of resilience in promoting wellbeing in LGBT+ individuals, communities and populations. A wider focus on resilience in later life has led to burgeoning research alongside debates about the ambiguities and methodological limitations of the research itself [11]. There is no universal definition of resilience [12] which is often described as a dynamic concept that may be researched as a trait, a process or an outcome [13]. Resilience has been defined as 'both the capacity of individuals to navigate their way to the psychological, social, cultural and physical resources that build and sustain their wellbeing, and their individual and collective capacity to negotiate for these resources to be provided and experienced in cultural meaningful ways' [14, p. 17]. It is also considered something that can be taught or learnt, as an individual characteristic or trait, and as a coping process in response to one's changing physical and social environment [11]. Resilience as a personal trait helps individuals cope, adjust and develop, inoculating them against the impact of adversity or traumatic events [13,15]. Resilience can be regarded as a function or behavioural outcome that can conquer and help individuals to recover from adversity [16,17]. It can be a process in which individuals actively adapt to and recover from major adversities [18,19]. In older adult research, resilience is often described or examined through a life course lens where the potential to adapt to the challenges, changes and disruption to adversity associated with normative ageing [20,21] can lead to the use of positive coping practices [22]. Allen et al's [11] definition of resilience speaks to the processes of being mindful of and prioritising behaviours, thoughts, and feelings that facilitate contentment within one's specific developmental, physical, emotional, and spiritual context. Angevaare et al's [23] concept review in ageing research identified three common features of descriptions of resilience: a stressor, a response and a mechanism, all of which are dynamic and emphasise the importance of the context in achieving resilience. More critical commentators [24,25] take a theoretical and political critique of how the concept of resilience has been applied in the social science literature and the implications of the resilience discourse. They argue that trait, process and outcome perspectives are focusing on our 'subjectivity' and as a consequence discussions about the 'outside world' [25, p. 40] and the radical transformation needed to challenge established social 'systems' [24, p. 254] are closed off or subjugated. In other words the emphasises on 'individual responsibility' [24] is a form of neoliberal governance that places onus on individuals and communities as consumers to become resilient and adaptable to external stressors, and in so doing the inequalities and oppressive social structures which create the need for resilience in the first place go unquestioned and depoliticised. Looking specifically at LGBTQ health research, Colpitts and Gahagan [26] found that although the concept of resilience emerged as a key conceptual framework to advance a strengths-based approach and suggested ways in which resilience is defined and measured in relation to LGBTQ populations, it remains a contested concept. De Lira and de Moreis in their review of the LGB literature [27] note that in the light of the simultaneous interaction between the individual, family, and social contexts, and their contributions to the process of resilience the conceptual dimensions of resilience must be further integrated to provide a more accurate description of its relational and systemic nature. In relation to LGBT+ ageing, some of the research conducted with populations experiencing physical or mental health difficulties has identified different factors and processes, which form pathways to resilience [28,29]. It is also important to take account of the intersectionality of social, cultural, economic and other factors that shape resilience in later life [29]. Higgins et al. [30] caution that while negative experiences can adversely affect LGBT+ peoples mental health and emotional wellbeing, this perspective may also unwittingly lead to LGBT+ identities being viewed as pathological. Herek et al's [31] social psychological framework for understanding stigma in sexual minority adults articulates how individual personal acceptance of sexual stigma as a part of ones own value system is internalised by adapting one's self-concept to be congruent with the stigmatizing responses of society. This has the potential to obscure and silences any potential for the development of unique strengths and skills that can be characterised as resilience [32] which in turn go unrecognised or undervalued. As a rapidly growing field of enquiry, knowing more about the degree to which theories and the nature of theories can shape the integration of knowledge will promote understanding and support for LGBT + lives [33,34]. How resilience is conceptualised and defined in research could help improve definition and methods used to study it and help deepen our understanding about what interventions might be effective in promoting resilience and quality of later life [29]. --- Aims of the review The aim of this systematic review was to examine how resilience is defined in LGBT+ ageing research and how it is studied. --- Methodology Given the aims and objectives of the review, we conducted a systematic mixed study review as described by Pluye and Hong [35]. Mixed study reviews have the advantage of allowing for a more complete analysis of the available evidence drawing from qualitative, quantitative and mixed methods studies [36]. The complexity and lack of consensus within the phenomenon of interest informed the need for integrating qualitative and quantitative papers and this emerged when articulating the research question [37]. A data based convergent synthesis design was adhered to [35,37] and we present the findings of the systematic searches using PRISMA guidelines [38]. --- Search strategy and study selection A comprehensive and systematic search strategy was constructed by adapting strategies in previous studies coupled with input from the project team. This was trialled in a sample database, edited based on comments from the team, and implemented by the information specialist [GS]. The search string explored the three main concepts of 'LGBT+' [e.g. Lesbian; Gay; Bisexual; Transgender], 'older' and 'resilience' using synonyms, controlled vocabulary and Boolean operators as appropriate ). CINAHL , Embase , Medline , PsycInfo , Social Science Database , and Web of Science were all searched from inception until 17 June 2022. The protocol for conducting this review was registered with PROSPERO . --- Eligibility criteria Given the aims of the review, the focus was on peer-reviewed papers in English only. Studies which reported primary research with older LGBT+ individuals who were aged 50 and above were included. Depending on the type of review people at the outset may define their concept or phenomenon of interest very tightly and use that definition as an inclusion criterion. However, as our review was focused on how researchers conceptualised and defined resilience within their research, we did not have a strict definition of resilience, although our reading and knowledge of the concept did inform our thinking on the search terms used. Studies included could be qualitative, quantitative or mixed methods as long as there was a clear mention of resilience within the paper. Resilience could be a primary or secondary focus of the research study or have emerged from participant's narratives in qualitative studies. Systematic reviews, case studies, randomised controlled trials, intervention studies and descriptive/discussion papers were excluded (see S1 --- Screening and selection Following the database searches, all citations were uploaded into Covidence screening software and any duplicates removed. Each title and abstract were independently assessed against the inclusion and exclusion criteria for eligibility by two members of the review team. Where there was difference in assessment, a third member of the team reviewed the title and abstract and made the final decision to include or exclude. Out of the 7101 papers reviewed at title and abstract level, 254 were put forward for full text review. These papers where once again assessed against the inclusion and exclusion criteria for eligibility by two members of the review team. Out of the 254 citations in total 27 papers were eligible for inclusion. --- Quality assessment Quality appraisal in a mixed study systematic review has been described as complex [39] and different approaches have been used [40]. We used the Mixed Method Appraisal Tool or MMAT [41], specifically designed for quality appraisal in mixed study reviews. The MMAT includes two general screening questions and a further five methodological quality criteria for common types of methodologies which permits the critical appraisal of mixed study reviews within a single tool and specific quality criteria for mixed methods studies [42]. An affirmative response to both the initial screening questions indicate that the paper meets initial criteria for screening. As per the MMAT user instructions, two of the authors independently assessed the quality of the papers and met to discuss and reach consensus about the quality of the individual papers. Excluding papers that are of low methodological quality is discouraged by the authors of the MMAT. All papers met the baseline for empirical studies following initial screening. Of the 27 papers included there were 13 qualitative and 14 quantitative papers. Collectively the quality of the papers ranged from average to good using the MMAT. All quantitative papers were appraised as cross-sectional descriptive studies based on their methodology. For the quantitative studies, only one study met all the quality indicators [43] with the others failing to meet at least one of the MMAT criteria. For the most part, studies were not representative of the target population or used non-probability convenient sampling. In addition, it was not clear in eight of the quantitative papers if the risk of non-response bias had been addressed [44][45][46][47][48][49][50][51]. Only four qualitative studies met the full criteria following initial screening [52][53][54][55]. All papers were subsequently included in the extraction and analysis phase as advocated by Hong et al. [42]. --- Data extraction and analysis Two authors designed and piloted the data extraction template, which included the aim of the research, sample characteristics, and details on the study methodology. Any data on resilience, particularly how resilience was defined and conceptualised, if resilience was the primary or secondary focus of the study, and whether it was an entry or an outcome of the study was extracted. For the purpose of this review, studies that had resilience as a stated or inferred objective were classified as entry point studies and those where resilience was a topic inductively emerging from the data were classified as outcome studies. Details on the tools to measure resilience and conceptual frameworks used, were also extracted . In the qualitative papers, data findings relating to the review aims were extracted verbatim, whereas in the quantitative papers, data was transformed in line with the convergent synthesis design that was adopted [37]. Data was extracted manually from each paper by two authors who then met to discuss the extraction and finalise the included data. Once the templates were populated with data, meetings were held with all authors who tabulated and summarised the data given the aims of this review, using a descriptive approach. This process included establishing an audit trail which articulated the process from paper selection through to data extraction, tabulation and data visualization which was checked by an author who acted as moderator who critiqued the process. In addition, regular meetings with all team members ensured that the team were clear on issues such as inclusion criteria, data selection, analysis and presentation of findings. --- Findings --- Overview of studies The 27 papers included represented 22 studies as four papers came from the 'Caring and Aging with Pride' study [47,48,68,69] and three papers further analysed data from the 'Aging with Pride' study sample [10,43,75], which collected data in 2010 and 2014. The majority of the papers came from the USA , three from Canada, two from Portugal, one from the Republic of Ireland and one from Australia. Three papers involved more than one country: USA/Canada , USA/UK and USA/UK/Canada . The papers were almost equally divided between quantitative and qualitative study designs. Quantitative studies focused on measuring variables that the researchers of the included papers considered represented resilience and the designs generally used descriptive, comparative and crosssectional designs , descriptive and cross-sectional designs and cross-sectional designs . The emphasis of qualitative studies was on exploring participants' perspectives or narratives on a variety of issues using qualitative descriptive , grounded theory , narrative methodology , phenomenological approaches and secondary analysis of previously collected data. Data were collected mostly using one approach, namely individual interviews , although in Bower et al's [65] study participants were also asked to share artefacts they felt enriched their narrative interview. Other methods used were focus groups , online survey with open-ended questions and one mixed methods study using interviews, focus groups, and a town hall meeting . Data analysis in four studies were informed by Charmaz's [76] grounded theory approach [52,53,57,66], guidelines on phenomenological research [56], thematic analysis [58,59,61], process coding [60], comparative analysis [54] and Merriam and Tisdfell's [77] method for data analysis using open and analytical coding [55]. Some studies used theoretical frameworks to support the analysis, such as ambiguous loss theory and the model of minority stress [52], generativity theory with a critical feminist perspective and queer theory [65], the Model of --- Commentary on focus: Resilience is neither a primary or a secondary focus of the study but an outcome of the research and is included by the authors in the discussion. Resilience emerges as an outcome of the study as a theme in relation to how participants demonstrated resilience against discrimination and oppression, which was interpreted as activism and is briefly addressed in discussion. Global and Situational Meaning [57], the Anderson's Health Care Utilization Model [61] and the Resilience properties in Aging Framework [62]. Quantitative studies applied methods of analysis correspondent to the research designs, such as t-tests, chi-square test, correlations tests, linear and multivariate logistic regression among others. Most studies reported the participant's demographics, including ethnicity, age, gender identity, and sexual orientation, while education and income were less commonly reported. In terms of demographics, a variety of terms were used such as gender identity, gender, sex, sexual orientation and sexuality. When reporting gender identities besides male man) or female woman) some papers used terms such as transgender [46,59,60,65,68,69], trans [66], transwoman and transman [45], while others additionally reported a variety of identities such as female with trans histories, pangendered, queer femme and trans, and male to female and back [54,58]. To avoid misrepresenting or misinterpreting the original authors, the terminology used in the data extraction table is what is reported within the papers. In those studies that reported ethnicity, the majority of participants were white and/or white non-Hispanic ethnicities, and the mean age varied from 52.5 to 67 years. While six of the quantitative studies mention gender identities besides men and women, five studies [43,44,51,71,75] focused exclusively on gay and bisexual men but none of the studies specifically state if all identified as cisgender. Six of the quantitative studies reported including people who identified as transgender or were reported as 'other' than female or male [10,45,46,48,68,69]. These constituted between 0.6% [48] to 39.8% [45] of the sample. While Fredriksen-Goldsen et al in one paper [48] reported that 0.61% of their sample identified as other than female or male in earlier papers on the same study sample [68,69] they report on a sample of 4.1% and 7% of transgender people. One paper by Fredriksen-Goldsen et al. [68] focused specifically on transgender people, while two studies reported included LGBT couples and spouses [52,62], but only one [52] reported gender identity. The predominant sexual orientation reported was gay , although one qualitative study focused only on lesbian women [56], and some studies focused on people who identified as bisexual [48,54] or bisexual individuals constituted most of the sexual minority sample [50]. Only two studies reported queer identities in their sample [54,60]. --- Definition of resilience Eleven of the papers offered definitions of resilience as either a trait, a process or an outcome [43,46,47,52,53,55,57,59,62,69,71]. However, there is no uniformity in the definitions offered. In the remainder of the papers, resilience was not clearly defined although it was alluded to and perceived as a strength or a positive attribute generally [10, 44, 45, 48-51, 54, 56, 58, 60, 61, 65, 66, 68, 75]. Where resilience was defined as a process, it was mainly seen as a process of adaptation where older LGBT+ individuals responded to adversities over time. � Denotes the language used in the studies to specify demographics . 1 Studies with a stated or inferred objective were classified as entry point studies. Studies in which resilience inductively emerged from the data were classified as outcome studies. 2 Studies drawing from the same data set . 3 Studies drawing from the same data set . https://doi.org/10.1371/journal.pone.0277384.t001 Adversities were perceived as the context within which LGBT+ resilience was operationalised and in some of the papers being a member of the LGBT+ community was associated with adversity in itself. Handlovsky et al. [53], writing about gay men's exposure to systemic discrimination, describe resilience as a response to risk exposure where environmental harms are mediated through internal and external protective processes. This allows older LGBT+ individuals to cope with adversity using both internal and external processes. King and Richardson [71] also refer to resilience as a coping mechanism, but refer to its health-promoting capacity as well its capacity to buffer physical and psychological distress. Nelson-Becker and Thomas [62] viewed resilience as 'the capacity to manage significant difficulty and stress and is both a process and an outcome' , while also describing it as a character trait. As a trait, it is seen as the capacity to manage stress which is facilitated by personality characteristics and environmental supports which mirror Handlovsky et al's. [53] definition. However, some definitions include outcome and process elements, such as Nelson-Becker and Thomas [62], who define resilience 'as an ability to integrate life learning and expand coping repertoires, reaching a new understanding that encompasses what came before but also moves beyond it' and as 'the ability to access one's inner wisdom and strength enhanced by time and experience' . Both Dziengel [52] and Nelson-Becker and Thomas [62] also bring in the notion of adversity shifting older LGBT+ individuals off-balance with resilience providing the capacity to help them to restore equilibrium. Foster et al. [57] also define resilience in the context of it being a trait which is the capacity to respond and successfully negotiate stressful life events in addition to it being a set of protective factors. Foster et al. [57] incorporate the ability of resilience to build capacity to also respond to future challenges which stresses the developmental nature of resilience. Higgins et al. [59] refer to Foster et al's [57] definition of resilience as well as using Ungar's [14] trait and process orientated definition which states that resilience is the capacity to navigate resources that build and sustain wellbeing. Fleishman et al. [46] use the American Psychological Associations definition of resilience as a process of adaptation to adversity, while Fredriksen-Goldsen et al. [47,69] discuss resilience in the context of behavioural patterns, functional competence and cultural competence that individuals and communities use in stressful situations. Emlet et al. [43] refer to resilience as a resource as well as a process, outcome and trait. Six qualitative papers did not set out to research resilience, therefore a conceptualisation or definition of resilience was not required at the outset of these studies. In one example, the researcher referred to resilience emerging only briefly in participants' narrations in relation to how historical experiences led the participants developing resilience in later life [66], while other authors included resilience in the discussion section of their papers [54,56,58,60,66]. In these papers, the reference to resilience is most often related to the experiences of discrimination, oppression and surviving the HIV/AIDS epidemic. These lived experiences led LGBT+ older adults to develop coping skills, which leads to resilience in later life. For example, Jen and Jones [54] relate resilience to the broader cultural and historical events and intersectional identities evident within the participants' narratives while also recognising resilient communities' efforts around HIV/AIDS. Green and Wheeler [61] discuss how the participants in their study demonstrated resilience as they aged which acted as a buffer to negative health care experiences related to their minority sexual orientation and HIV status. Others only briefly mention resilience as part of participants' strengths [56,60] and as a potential topic for training that would shift practices from pathology to seeking indicators of resiliency [58]. Only Rowan and Butler [56] set out to define resiliency. They define it by using Blundo's [78] understanding of resiliency as 'finding a stronger and more meaningful way to deal with life difficulties and stressors' . --- How resilience was studied Adopting a clear definition appeared to be influenced by how the researchers approached the study of resilience. Twenty-one studies were classified as entry point studies as they had resilience as a stated or inferred objective and six studies were outcome studies as resilience inductively emerged from the data. In all the quantitative studies , [10, 43-51, 68, 69, 71, 75] resilience was an entry point, while the qualitative studies, were divided between entry , [52,53,55,57,59,62,65] and outcome studies , [54,56,58,60,61,66]. Entry point studies were further classified based on the focus researchers gave to studying resilience. Studies that utilised resilience theory and aimed to explicitly research resilience were classified as primary-focused. These studies specifically explored resilience in terms of how it was developed and the factors that influenced it. Where resilience was a secondary focus or a secondary research objective, resilience per se was researched in the context of other phenomena, such as multiple stigmas in relation to LGBTQ aging and cultural generativity [65], integration of sexual orientation and faith [57], physical and mental health [45,50], as one of the several mental health and well-being indicators [49], and others. Two primary focused studies specifically researched spiritual resilience [57,62], where the focus was on the struggles LGBT people had in reconciling their sexuality with their sense of spirituality or how the persons spiritual beliefs or practices supported them in addressing life's challenges. Others examined more general protective processes or factors that foster positive health outcomes and decreased psychological distress associated with sexual identities [53]. Resilience was also studied as an adaptive factor [44] that helps older LGBT+ people to mediate the impact of stigma and discrimination and adapt to prejudice and loss [59]. As a primary focus, resilience was also studied alongside other predictors of mental distress in gay men such as stigma, discrimination and internalised homophobia [71] and as a predictor of sexual satisfaction [46]. Resilience, along with the concept of mastery, were studied as two separate psychological resources that relate to each other but may be influenced by different factors in the structural-environmental and individual contexts [43]. In addition, Haldane et al. [55], set out to study how prominent 'sexual orientation minority elders' perceived the LGBTQ+ community as developing hope and resiliency in relation to major events that led to the development of rights for LGBTQ+ people. In papers where resilience was a secondary focus, it was often seen as something that was exercised or demonstrated in the face of adversities. In Bower et al's [65] study for example, they set out to research multiple stigmas and cultural generativity in relation to LGBTQ and ageing. They also set out a secondary objective to discuss how the experience of stigma and discrimination support a generational legacy of resilience. Similarly, Foster et al. [57] explored the integration of sexual orientation and faith and discussed how spirituality may become a method for meaning-making and resilience. Resilience as a secondary focus was more prominent in quantitative papers and often researched in relation to the presence of different physical and mental health characteristics, which were used as proxy measures of resilience. Cortes et al. [45], for example, set out to find and confirm evidence of resilience in older LGBT adults by measuring levels of anxiety, depression and alcohol use and then comparing them with younger LGBT adults. Similarly, Monin et al. [50] examined whether sexual minority status confers vulnerability or resiliency in older adulthood by using a variety of scales to measure mental health, social support and exposure to trauma. Included papers from the 'Caring and Aging with Pride' and 'Aging with Pride' studies, in most cases did not report measuring resilience specifically, but used a variety of tools to measure key health indicators, risk and protective factors and health outcomes. In these studies, higher scores in the variables measured suggest the exercise of resilience. Emlet et al. [43,75] also drew their findings from 'Aging with Pride' but differed because they reported directly measuring resilience. --- Tools and conceptual frameworks used In the studies by Batista and Pereira [44] and Pereira and Silva [51], resilience was a primary focus and was measured using the 10-item Connor-Davidson Resilience Scale that utilises a four-point Likert scale . This self-reporting scale takes the perspective that resilience is a personal quality that reflects the ability to cope with stress [80]. Fleishman et al. [46] and King and Richardson [71] used Wagnild and Young's Resilience Scale [72], although the former used a 25-item scale, the latter used a 14-item scale without a clear rationale for omitting some of the items on the original scale. Wagnild and Young's [72] self-reporting scale views resilience as a positive personal characteristic that enhances individual adaptation. Emlet et al. [43,75] used a three-item scale adapted by Fredriksen-Goldsen and Kim [73] from the Brief Resilience Scale by Smith et. al [74]. The BRS was designed to assess the person's ability to bounce back or recover from stress [74] and was used in its original form in the study conducted by Lyons et al. [49]. A number of studies used a conceptual framework to guide the research design or the analytical process. Fredriksen-Goldsen et al. [47] used a Resilience Conceptual Framework that incorporated risk and protective factors. The framework included five components: ' background characteristic ; key health indicators ; risk factors ; protective factors ; and health outcomes ' [47, p. 665]. While this framework was used to inform a number of the 'Caring and Aging with Pride' studies [47,68,69], there was variation in the title given to the framework and in how the dimensions of the framework were reported within the different papers published from this study. Fredriksen-Goldsen et al. [10,48] and Emlet et al. [75] used the Health Equity Promotion Model [70] that situates health more broadly within the life course, asserting that optimal ageing is linked to the availability of opportunities over time to promote health. 'This model highlights how social positions and structural and environmental contexts intersect across the life course with health-promoting and risk processes to culminate in the health and well-being of LGBT adults as they age' [10, p. S73]. The authors argue that the key variables in the model influence resilience, therefore higher scores in health behaviours suggest the exercise of resilience. Finally, Foster et al. [57] adapted the Model of Global and Situational Meaning [81] to develop a theoretical model for Lesbian and Gay Christian Spiritual Resilience [57]. --- Discussion Of the 27 papers included in this review, the majority did not provide a formal definition or operationalisation of resilience at the beginning of the study. In these papers, resilience is either briefly mentioned and/or related to other concepts, which leaves the reader to deduce how researchers conceptualised resilience from the methodology used. Papers that utilised conceptual frameworks often did not provide a definition of resilience but identified it as part of the larger framework. Furthermore, considering both the papers in which resilience was an outcome of the research and papers in which resilience was the secondary focus, only ten of the 27 papers definitively set out to research resilience as a primary focus. This points to resilience being rarely researched in the older LGBT+ population as a concept in itself. This review also shows that resilience is indeed difficult to categorise as either a trait, process or an outcome, because of the fluid nature and the interrelationship between all three aspects of resilience. Therefore, researchers often write that they are researching resilience as a trait but discuss the findings as an outcome of a lifetime of responding to adversities through a process of adaptation. Though a clear delineation might be unnecessary, there is a lack of clarity on what view of resilience is taken by some researchers and how their understanding of resilience informs their study and its methodology. Lack of clarity is further hampered by authors using other words when talking about resilience, sometimes interchangeably, words such as resiliency [e.g. 50, 52, 55], coping strategies, adaptive factor, and strengths. Although some of the quantitative studies used scales to measure resilience, some operationalised the concept of resilience by measuring temporal changes in mental distress or mental health outcomes such as depression, anxiety and substance misuse [45] where resilience is equated with lower scores of distress or mental illness. Equating resilience with the absence of, or, reduced 'symptomatology' has been critiqued from the perspective of perpetuating negative perceptions of LGBT+ people as 'at risk' and 'vulnerable' as well as failing to acknowledge the wider socio/structural and system factors that impact LGBT+ lives [82,83]. Although the study of resilience is seen as one way to shift the agenda from a risk-deficit focused approach to a more strengths-based approach [83,84], most of the literature on resilience in the papers reviewed, discuss the specific circumstances and histories of LGBT+ older adults and research or describe resilience as a trait LGBT+ older adults possess to help them bounce back during adversity. More importantly, the findings highlight how resilience theory is insufficiently applied throughout all stages of the research process. In keeping with previous commentary on the use of theory in research [33,34], within the studies included in this review there was less of a tendency to use resilience theory to inform older LGBT+ research. Based on Bradbury-Jones et al. [33] typology of theory use, most of the papers met the criteria of either implied theory , partially applied or retrospectively applied theory . Retrospective application can be observed in papers that were classified as outcome studies, where resilience theory was 'considered at the end of the study as a means of making sense of research findings' and/ or 'introduced as an afterthought' [33, p. 137]. According to Bradbury-Jones et al. [33] the highest level of theory use is the consistent application of the theory throughout the research process. This level was rarely observed, except for the studies classified as primary/ entry focus which set out to research resilience , provided a review of literature and applied resilience theory throughout all stages of the research process. On the other hand, secondary focus studies rarely define resilience, and in one case even set out to measure resilience, but give no attention to presenting the theory of resilience throughout any of the phases of the study [49]. The findings also highlight, that many of the papers position their research specifically in the context of the history and effects of the HIV/AIDS pandemic on gay and bisexual men. While important, this overshadows the wider social and political landscapes that shaped gay and bisexual men's lives, shifting the focus away from lesbian's, bisexual women's and transgender people's experiences, as well as those of older LGBT+ people from Black, Asian and other ethnic minority backgrounds. Furthermore, queer identities rarely constituted a part of the sample [54,60]. This may be due in part to the age of the study participants, as they may have viewed the terms queer as a derogatory and insulting, as opposed to a term being reclaimed by the younger activist community. --- Strengths and limitations of the review A member of the team with specific expertise in literature searching and retrieval conducted the database searches without any time limitations being applied . This helped to ensure that the review was as comprehensive as possible. In addition, the review process, data extraction and quality assessment were conducted by more than one author, making the process less susceptible to bias. However, many of the included studies had methodological weaknesses which may influence the quality of the findings in this review. Only peer-reviewed papers in the English language were included. Furthermore, papers that included resilience might have been excluded where resilience was not mentioned in the title/abstract/keywords section of the paper or in cases where researchers used terminology other than resilience although our search strategy sought to minimise this. While some might consider the inclusion of multiple papers from the same study a limitation, with the potential to skew the findings, their inclusion shows how even within the same study resilience is differently applied and conceptualised across the papers. Deciding on whether resilience was a primary or secondary focus had an element of subjectivity attached. For example, we classified Haldane et al. [55] study as a primary focused paper, as the study of resiliency was stated as an aim; however, as resiliency was not included in the subsequent research questions or reflected in the interview schedule others may have classified this paper differently. --- Future research In terms of future research this review points to the need for researcher to explore in greater depth from an emic perspective how older LGBT+ people conceptualise resilience, including the socio structural factors that promote or hinder its development. It is also important that future studies are designed in a way that takes account of peoples' multiple identities might intersect and influence their understandings and experiences of resilience. The review demonstrates that white, gay and middle-class men were more prevalent in the samples of the included studies, therefore the question of whom is being studied needs to be considered in any future research on resilience in LGBT+ later life. In addition, future researcher should be more explicit about how they are using resilience theory, including the discourse or perspective they are taking when employing the concept. As highlighted in the review more studies are also required to make explicit how concepts such as resilience, resiliency, coping and adaptation differ. If the study of resilience is to move towards a strengths-based approach, there is a need for researchers to consider their choice of measurement scales and move beyond scales that measure symptomatology. There is also a need for future resilience studies to move beyond the 'individual' adaptation discourse to one that addresses or politicises resilience in the context of socially embedded discrimination experienced by the older LGBT+ community. Finally, while not specifically focused on resilience, to avoid confusion there is a need for researchers to be cognisant of the language they use to report demographic characteristics, especially when reporting on sex, gender and sexual orientation. --- Conclusions There has been an increasing growth of research into the experiences of LGBT+ ageing internationally, which has attempted to connect with more recent advances in developing theoretical models of resilience both in the LGBT+ population [26,27] and ageing [21]. The body of research reviewed here has considered the interactional and contextual features of both ageing and LGBT+ identities sometimes without a clear definition of resilience, and the variety of studies tend to show a lack of coherence in the way in which the concept of resilience is utilised and theorised in LGBT+ ageing. This review is the first in our knowledge to have examined how resilience is actually defined and conceptualised in LGBT+ ageing research and highlights some of the challenges in providing an adequate comprehension of resilience and how this can be better integrated. Further work to develop a typology of resilience theory taking a more ecological perspective may help to apply resilience to LGBT+ ageing research in a more comprehensive way. Understanding the mechanisms involved in resilience, including the minority stress and structural issues that impact resilience has the potential to advance our knowledge and create new and innovative theoretical approaches [85] to inform affirmative policies and practices. Furthermore, greater clarity around the concept of resilience would support the education and practice of professionals and providers in relation to resilience and unique experiences of older LGBT+ adults [32,86,87]. --- All relevant data are within the paper and its Supporting Information files. ---

Within the literature, resilience is described as either a trait, an outcome or a process and no universal definition exists. A growing body of research shows that older LGBT+ adults show signs of resilience despite facing multiple inequalities that negatively impact their health and social wellbeing. The aim of this review was to examine how resilience is defined in LGBT+ ageing research and how it is studied. A mixed-study systematic search of peer-reviewed research papers published before June 2022 was conducted using the electronic databases CINAHL, Embase, Medline, PsycInfo, Social Science Database and Web of Science. This resulted in the screening of 7101 papers 27 of which matched the inclusion criteria. A quality appraisal was conducted using the Mixed Methods Appraisal Tool. Findings show that papers often lack a clear definition of resilience and application of resilience theory within the studies, although many of the papers conceptualised resilience as either a trait, process or an outcome. However, resilience was rarely the primary focus of the studies and was researched using a variety of measurement instruments and conceptual frameworks. Given the socioeconomic disparities, diverse social relations, histories of discrimination and stigma, and acts of resistance that have shaped the lives of older LGBT+ populations, resilience is a topic of growing interest for researchers and practitioners. Clear definitions of resilience and application of resilience theory could help improve methods used to study the concept and lead to more robust findings and the development of effective interventions. Greater clarity on the concept of resilience could also broaden the focus of research that informs policies and practice, and support practitioner training in resilience and the particular experiences of older LGBT+ adults.

Background In rapidly growing cities in low-and middle-income countries , changing working and living conditions leave families struggling to care for their children. The World Bank estimates that 350 million children lack quality childcare globally . Evidence from population surveys in 61 LMICs showed that up to 17% of under-5 children are left alone at least 1 day a week in East and Southern Africa and up to 29% in South Asia . It is estimated that 200 million under-5 in LMICs are at risk of not reaching their developmental potential, largely due to infections, nutritional deficiencies, and a lack of responsive caregiving . This is despite global recognition that healthy development in the early years is a pre-requisite for future health and productivity of individuals and societies . Families in cities in LMICs face an unprecedented childcare challenge as gendered work patterns change and both parents work outside the home for long hours . With reduced support from the extended family, low-income parents, particularly mothers, are left with limited options for childcare. Many must either take their children to work or leave them at home unattended . This undermines child health, already compounded due to detrimental living conditions in deprived urban neighborhoods with children exposed to injuries , poor nutrition , and poor hygiene . Threats to child health are further compounded by the challenges that urban health systems face in meeting the needs of their expanding populations resulting in significant intra-urban inequity in health outcomes . While distance may not be a barrier, the limited provision of free, public primary care clinics, open at times when parents are not working, is a significant barrier to access . This leads to limited uptake of child health programs among urban poor households as exemplified by the finding that only 58% of children in Kenya's slums are immunized . Given changing work patterns and social structures, there is high demand for center-based childcare within cities. Few studies have quantified the demand for center-based care . Those that have, have found high demand, for example, 84% of residents in a deprived neighborhood in Dhaka wished to use center-based childcare, with 3.8 times higher odds of demand among slum than non-slum households . The burgeoning of poor-quality, unregulated centers is clear to see in poor urban neighborhoods across LMICs . In one of Nairobi's informal settlements of around 200,000 residents, a rapid survey identified over 50 informal home-based centers . Centers frequently consist of one room with poor sanitation and facilities where a woman watches over children for a set fee. While NGO providers are often able to provide better quality childcare than informal providers, they rarely meet parents' needs for long hours of care and our research in Dhaka has found that the sustainability of such externally funded services is a major limitation . Good quality center-based care offers the potential to support healthy child development whilst also allowing parents, particularly mothers to work for an income and older children to remain in school rather than care for younger siblings. There is limited evidence, particularly from Africa and Asia, of the impacts of center-based care on child health, early childhood development , and wider benefits for mothers, families, and long-term social and economic benefits. Available evidence points to positive impacts on short-and long-term ECD outcomes . More recent primary studies have identified improvements in cognitive and socio-emotional skills and long-term impacts on vocabulary . Evidence of health impacts is particularly limited, although Behbehani's et al. review of center-based childcare for under-3s found some low-quality evidence of improved immunity and normal growth, especially for those attending for longer . In Bangladesh, community-based childcare centers for under-5s in rural areas reduced the risk of all-cause mortality by 44% , drowning by 82% , and injuries by 88% . Evidence of the social and economic benefits of center-based childcare is considerable, particularly the impact on women's participation in the labor force in high-income countries and in LMICs . Evidence of the impact of childcare centers on family health and wellbeing is limited. However, some limited impact on women's mental health has been identified in India . Importantly, evidence from both HICs and LMICs indicates that low-cost but poor-quality center-based care may worsen ECD outcomes and lead to discontinued breastfeeding and increased infections . Figure 1 illustrates the potential benefits of quality center-based childcare to children and their families. Improving the quality of childcare centers in complex, poor urban neighborhoods is vital if these centers are to benefit, rather than undermine children's healthy development. Given the challenges working parents face in taking up child health programs, childcare centers also offer an ideal setting for outreach to increase access to child health programs such as immunization, deworming, and vitamin A provision among "hard-to-reach" lowincome families. In this study, we present the process of developing an intervention to support center-based childcare in informal settlements in Kenya. Despite the challenges facing public health systems in cities in LMICs, any intervention to support childcare centers must be embedded within this system if it is to be sustainable and with potential for scale-up. The local governmentmanaged community health teams which consist of community health assistants and community health volunteers are a considerable asset within urban areas in Kenya. Their valuable contribution has been recently recognized through the passing of the Nairobi City County Community Health Services Act which provides a monthly stipend and health insurance to CHVs in Nairobi. In recognition of the existing knowledge, experience, and emic perspective of the CHVs and childcare providers, we draw on the Communities of Practice model, which, based on the situated learning theory , allows peers to share their experiences and practice-based knowledge to learn and improve their practice, rather than through formal instruction or training. The model has been successfully used to harmonize and improve ECD services among various stakeholders in South Africa . The intervention development process is part of a larger study to test the feasibility of the resulting intervention . Here, we address the following objectives: 1) To map and assess the childcare environment and provider skills in informal settlements. 2) To co-design with childcare providers, parents, government, and ECD experts a supportive assessment and skill-building CoP approach which can be delivered at scale within informal settlements in Kenya. --- Methods --- Setting The study was conducted in Korogocho and Viwandani, two informal settlements in Nairobi, Kenya. These two settlements were selected due to their differing characteristics, with Viwandani close to the industrial area being relatively less poor than Korogocho. Viwandani has seven villages, most structures have roofs made from iron sheets and walls from tin and iron sheets. The Ngong River flows to the south of the settlement and is heavily polluted by the industries which are mainly to the north of the settlement. There are also seven villages in Korogocho, and most homes are built of mud and timber and recycled tin cans or sheets as roofing materials. Korogocho is one of the most congested slums in Nairobi with over 250 dwelling units per hectare . Both informal settlements have a high proportion of under-5 children and women working outside the home, often in informal employment. Located about 7 km from each other, Korogocho and Viwandani are densely populated with 63,318 and 52,583 inhabitants per square km, respectively. These settlements are characterized by poor housing, poor sanitation, lack of basic infrastructure, insecurity, high crime rate, and poor access to maternal and child health services and healthcare in general . Variations in these two communities will support the transferability of study findings to wider urban-poor settings in the East African region. --- Study design To co-design the intervention, we used a sequential mixedmethods design in combination with the Experience Based Co-Design process . EBCD uses a series of interviews, discussions, and observations to share perspectives and issues between different stakeholder groups to develop health service innovations and improvements. EBCD often relies on the use of videos to share the perspectives of different stakeholders to inform intervention design. We mapped childcare centers in two informal settlements to identify their locations for the implementation and feasibility testing phase of the project and to understand their key characteristics to inform the design of the intervention. This was followed by qualitative interviews and group discussions with different stakeholder groups , using video to capture key issues that could then be shared during workshops to agree on the design of the intervention. The data collection period coincided with the COVID-19 pandemic, with subsequent lockdowns and social distancing regulations in Nairobi. During this period, many childcare centers closed in line with COVID-19 guidance for schools. Several changes to our protocol were required. Mapping and assessment of childcare centers, planned as the first activity, had to be delayed until most centers reopened in February 2021. Social distancing requirements meant that we could not hold focus groups using participatory methods with parents, center providers, CHVs, and CHAs and videos to summarize key issues from different participant groups as recommended within the EBCD process . Despite waiting for childcare centers to reopen, many remained closed even after restrictions were lifted in both informal settlements selected for the study. Figure 2 illustrates the revised co-design process. The initial discussions with sub-county and county officials and nongovernmental organizations identified CHVs as an appropriate cadre to deliver the CoP sessions and supportive --- FIGURE The potential impacts of quality center-based childcare on child and family outcomes. --- FIGURE The co-design process. supervision in informal settlements. Subsequently, qualitative interviews were held by phone with parents and guardians using childcare, and with childcare providers. Five workshops were held with parents, community health teams, i.e., CHVs and CHAs, and their sub-County managers and Nairobi City County policy makers. Through a review of published literature, Kenyan ECD and child health policies , and discussions held with ECD, child health, and nutrition experts, we identified appropriate content for the intervention training and materials. The NGO partners in the study, Kidogo, who work to support and improve quality among informal childcare centers in Nairobi's informal settlements, shared key materials and childcare center assessment tools, which were adapted and included . Following preliminary analysis of qualitative data and reflections on the co-design workshops 1-5, two final workshops brought together parents, center care providers, government officials, and county and sub-county ECD experts to finalize the CoP model with supportive assessment tools. In the protocol, we planned to map and record brief details of all childcare centers in the two areas and conduct questionnaires --- Home-based Home-based are centers that are within the dwelling units of the childcare center providers. In some cases, the providers hired separate rooms within the plot where they lived and designated the rooms for use as childcare centers. In most cases, however, the same room where the childcare center provider lived also served as the childcare center. Most of the center providers in this category started the childcare center business out of necessity and initially started by looking after neighbors' children and then turned this into business. As a result, the majority do not have any training on ECD or any other childcare-related training. School-based School-based are centers that are based within the school system, usually attached to a primary school. In these cases, the schools have a childcare center unit, which also serves as a pre-primary school unit. Most of the care providers in this category are trained ECD and primary school teachers; some are also pursuing degrees in education. Center to assess the knowledge, attitudes, and practices of childcare providers and CHVs early in the co-design process. However, given the childcare center closures and workload of CHVs due to COVID, the mapping exercise and questionnaires were conducted immediately before the training of CHVs and the start of the intervention. With further training sessions being held over the 4 months of implementation , we were able to use the questionnaire findings to inform training content and intervention design for the later modules of training. --- Qualitative methods --- Interviews with parents and guardians With the aid of CHVs in the two informal settlements, we were able to identify childcare providers from different types of centers using purposive sampling . We asked providers from a range of childcare centers to help us identify parents and guardians of the under-5 children using the centers. We then purposively selected mothers, fathers, and guardians of a range of ages, number of children in the household, and different occupations. This allowed us to explore a range of different perspectives and childcare needs during the telephone interviews. Interviews explored participants' current childcare center use, experiences, and preferences, including expected standards and willingness to pay for different services . Details of qualitative interview participants are shown in Table 2 below. --- Interviews with childcare providers Once all childcare providers were identified by the CHVs, we purposively sampled providers from childcare centers with different characteristics . The interviews with providers explored their experiences of running a childcare center, details of any training and support they had received, how they felt they could improve their centers and the quality of care provided, their interactions with CHVs, and suggestions for the intervention. Interview guides for both the providers and parents/guardians were translated into Kiswahili and the interviews were conducted by a team of 6 qualitative researchers all fluent in Kiswahili or other local languages to enable good communication with participants. Interviewers received a 5-day training on the study, qualitative interviewing, and ethical issues. The telephone interviews were arranged at a time convenient to the participant and all costs of phone data were covered. Interviews were recorded and transcribed into English soon after the interview. All transcripts were checked by the interviewing team for accuracy. We continued interviewing until similar issues were emerging repeatedly from each participant group, i.e., parents/guardians and childcare providers, and we felt data saturation had been reached . As described earlier, data generated from the interviews were discussed in the co-design workshops to ensure that the development of supportive supervision and CoPs was grounded in the realities of childcare center provision in informal settlements. Throughout the process, the study team checked back with government officials to clarify any issues raised and gain their feedback on emerging issues. --- Mapping and rapid assessment of childcare centers Field interviewers , many of whom lived in the two areas, were trained to use OpenStreetMap © to map all the childcare centers in the two informal settlements. Each FI was paired with a CHV and together they visited childcare centers within the CHV's catchment area. The CHVs helped to locate the childcare center and subsequently introduced the FI to the childcare center provider. The FI then proceeded with the consent process before administering the mapping tool. Basic information about the centers was collected by asking the provider questions, observing the facilities and practices in the childcare centers, and checking any records available to ascertain: opening hours; staffing levels; number and ages of children; rooms; hygiene facilities; fees; any organizational/NGO support; and the name of local CHV. --- Knowledge, attitude, and practices questionnaires --- Childcare providers Center providers who agreed to be involved in the intervention feasibility testing component were also asked to complete a questionnaire on their childcare knowledge, attitudes, and --- Community Health Volunteers Before the initial intervention training session, the CHVs allocated by the sub-county to deliver the intervention were asked to complete a brief questionnaire to assess their knowledge in relation to child protection, safety, discipline, and abuse; stimulating environment; responsive caregiving; learning through play; health; nutrition; and water, sanitation, and hygiene. In addition, they were asked about their current practices of visiting childcare centers in their catchment areas and their motivation to facilitate group sessions and provide support to the center providers . --- Analysis An initial rapid analysis of parent and childcare provider interview transcripts was conducted using codebook thematic analysis to identify key areas to include in intervention design. The 5 areas of the Nurturing Care Framework were used as predetermined codes and structured codebooks to analyze from a deductive perspective and then further themes were added as they emerged inductively from the data to structure the rapid analysis: Opportunities for early learning; good health; adequate nutrition; responsive caregiving; and safety and security. The process highlighted additional themes on the community context, center infrastructure/environment, and considerations for intervention delivery. The rapid qualitative analysis supported a clear specification of the types of childcare providers , and this informed the data to be collected during the quantitative mapping of childcare centers. The rapidly identified issues and areas were presented and discussed in the final co-design workshops 6 and 7 held in January 2021. A detailed coding framework grouped around these themes was then applied to all transcripts using NVIvo 2020 software. Detailed notes from the co-design workshops were added to the qualitative analysis and coded using the same themes. Attributes were set for all the respondents based on gender and location and for the parent/guardian interviews: the relationship to the child; marital status; and occupation; and for the provider interviews: the type and years of operation of the center. The attributes aided the analysis team to explore different patterns in knowledge and practice between different types of centers and for parents and guardians with different characteristics. Data from the mapping exercise and the KAP questionnaires with CHVs and childcare providers were analyzed using descriptive statistics in Stata Version 17. Consent forms and data, including the location of childcare centers, were kept confidential and locked away to ensure that childcare center providers' privacy was protected. --- Results The mixed methods results of the brief survey of center characteristics, knowledge, attitudes, and practices surveys and qualitative interviews with parents, guardians, and center providers are presented showing the strengths and challenges of the childcare centers in addressing the nurturing care domains within the context of informal settlements. The meta-inference from these findings and the potential solutions identified during the co-design workshops and how these culminated in the final intervention are presented along with the details of the co-designed intervention. In total, 129 childcare centers were identified, 55 in Korogocho and 77 in Viwandani. All consented to provide essential information about their centers which is presented in Table 4. Of the 129 centers mapped, almost two-thirds were in Viwandani where small home-based centers dominated. These home-based centers received less external support and were less likely to have had any training in ECD than other center types. This confirmed the information shared during the co-design workshops with community health teams that the home-based centers had poor standards and the women running them were untrained in child health and development and unsupported . In light of this, a key decision in the intervention development was to exclude the school-based centers and focus instead on the home-based, slightly larger centers and faith-based centers. Of the 68 centers that fell into these categories, 66 agreed to be part of the implementation and feasibility testing of the intervention and completed the more detailed questionnaire . --- Early learning Both the qualitative interviews and questionnaire results provided insights into early learning practices and perceptions. Most providers talked about how they sang and played with the children using whatever toys and space they had. The questionnaire results reinforce this with singing, dancing, and movement games which are the most common activities reported, particularly in the home-based centers . Some providers in smaller home-based centers had found ways to create space for play despite the constraints they faced: "My house has a space for play because I remove the table. I have two large seats, and a cupboard. I create room in the middle for them to play." . Parents identified multiple gaps in the skills of providers, and even when some providers were skilled in learning through play, they lacked experience in responding to health, nutrition, and hygiene needs of the children. "All that happens at that daycare centre. . . including playing with children. . . the children know nursery rhymes such as 'gari ya moshi'. She is very okay with that because she has enough space. The problem with her is not giving the child food on time and not changing the child's diapers." . Some providers, particularly those in larger centers, mentioned reading and writing with the older children and described how children had toys and space to allow play: "Mostly we group them, there those who like to play and moving around. They play with playing materials like blocks, dolls and toy cars and at certain times they read orally. You write with bold letters you read to them, and they recite so they can get the concept of reading." . The lack of play materials was apparent from the questionnaire results with only 22% of home-based and 29% of faith-based providers reporting that children had a toy or other appropriate object to play with. This was slightly higher among the larger centers . Parents frequently commented on the lack of play and learning materials. Those who were able to, took their own toys for their child to play with at the childcare centers. Many parents saw the childcare center as an important step in the transition to primary school. They valued the skills of the providers in playing and teaching their children. "So she can interact with other children, to learn how to walk, talk and he goes to school it will be easy" . --- Home-based Centerbased Faith-based Total N = n N = n N = n N = n Business license displayed 0 2 1 3 The lack of training of providers, particularly younger women without children of their own, was identified by many parents as a barrier to responsive caregiving and the ability to support children to learn and develop. "More teachers who are trained in childcare should be brought on board, so that the children who are about to join Baby Class can be taught how to be with other children. They can also be introduced to what they will be taught in the next level so that they can know. . . the day-care service provider is mainly playing the role of a parent right now, but it would be good if she would get more training." . Sub-county and county-level workshop participants focused on the need to create a healthy environment and protect and promote child health, with less of an emphasis on ECD. However, during workshop 5 with sub-county officials, there was a recommendation that the intervention could do much to encourage center providers to make use of any local materials they could find to improvise playing material for child stimulation. The importance of learning through play was identified by participants in workshop 6 who were keen to ensure that providers had these skills to allow their children to develop their skills and talents. Some of the attitudes toward interactions with children were illustrated in the responses to the CHV questionnaire with 27% feeling there was no need to show an interest in children's conversation and 32% believing very little talk was needed with under-5 children . --- Responsive care Many parents and grandparents emphasized that their main motivation for choosing a childcare center was the caring nature of the provider and most were confident in the abilities of their chosen care provider to provide responsive and caring attention to their children. However, high child-to-caregiver ratios were frequently mentioned, and parents were concerned about how this impacted the responsiveness of the provider to their children's needs. The mapping exercise found the highest child-to-provider ratios of 1:30 among the school-based centers with some centers having over 400 children. Homebased centers had a mean ratio of 1:8. However, there were some outliers with over 20 children and only one provider . High child-to-provider ratios were raised as key concerns during interviews with parents and the workshops with community health teams . Parents identified the negative impacts of these high child-to-provider ratios and expressed concern about how responsive single providers could be, particularly with many children of different ages: "That is the biggest challenge like I told you. They need to be like two or three teachers. Because what happens is that the older ones tend to be kept on the side as the young kids are taken care of. She might have two or three children crying and she is alone. She may carry one of them while she the other crying ones continue. The other challenge is on feeding them because they are all in her custody and she is all alone." . Being responsive during feeding was an area that was explored in the questionnaire and the majority of providers reported supporting children to eat. However, 11% of providers in home-based and 21% of providers in larger centers reported leaving children to eat on their own. A relatively high proportion of home-based and center-based providers felt children should be treated harshly, with only 56 and 57%, respectively, saying this should never happen, compared to 86% of faith-based centers reporting this never happened in their centers. --- Nutrition While the questionnaire responses indicated that a high proportion of all providers were able to provide uji and lunch, within the qualitative data, both parents and providers emphasized their concerns regarding the ability to ensure this level of provision of food at the childcare centers. Within the centers, there were different approaches to the provision of meals depending on the provider's ability to store and heat food. Some parents gave the providers money in addition to the regular fees to buy and prepare cooked food, but the majority in both informal settlements provided pre-packed food to support the implementation of new skills/practice and to monitor quality using the simple assessment tool. • Supervisors meet volunteers regularly and accompany them on childcare visits or during CoP group sessions as needed. --- Who Community volunteers • Community workers or volunteers with an interest and minimal training in child health and development would be appropriate to deliver the intervention. They need to have time available to visit childcare centers and run monthly group CoP sessions within their neighborhood. CHVs receive a stipend to cover the expenses of their work. --- Supervisors of community volunteers • Within our intervention, the CHAs supervised the CHVs through their routine meetings and supervision structures and provided training for CHVs. The CHAs hold at least a diploma in any health discipline and they are the primary supervisors of the CHVs. They will provide supportive supervision to the CHVs during the CoP sessions and assessment visits. --- Local experts on child protection ECD, health, and nutrition • To supplement the knowledge/skills of the main supervisors during the training. Experts will require a good understanding of the context to be able to ensure topic knowledge is applied and relevant. --- How • Community volunteers facilitate monthly face-to-face CoP sessions to facilitate trained center providers to share knowledge and experience on the four module areas. Sessions last several hours depending on provider availability. • Visits from the community volunteers to the childcare centers are in-person, informal, and supportive. --- Where • Training can be held in a suitable community venue. • CoP sessions can be held in childcare centers when no children are using the center. Hosting CoP sessions can rotate between group members. This helps to ensure discussions are grounded in the realities of the center environment and facilities. --- When and how much • The total intervention can be delivered in 4 months, however, 6 months may be more appropriate, with the 6 days training of community volunteers spread over the period. This ensures the CoP sessions on the same topic are held soon after the training. • Monthly CoP sessions last 1 or 2 h, or longer depending on the interest of the members. • Community volunteers should visit all center providers from the group during the month between CoP sessions . --- Tailoring and implementation • The changes made to the intervention during delivery will be described in a subsequent study. for their children. Providers, particularly those from home-based centers, found the pre-packed food difficult to manage with some children having more and better food while others had very little or nothing. "I have a big challenge here because there are those who bring food and a parent might come and tell me they don't have food. I make sure they all eat, but some don't come with food and I can't let them starve the whole day." . "The centre is a small room, and the children are around 30 or 40. I am not so sure how many children they are but they are very many. So, they really squeeze in there. They don't have toys there, and sometimes there is no food. They used to be given porridge there but now they are not given." . Our qualitative analysis highlighted the limited nutritional variety in the foods provided by both parents and providers. Rice, ugali , and uji were mentioned most frequently. Occasionally, providers and parents were able to add cabbages, beans, and spinach depending on available funds. The questionnaire pointed to gaps in provider knowledge on a healthy diet, particularly in advising parents on what they should bring for their children or feed them at home. Few providers were able to plan meals with diverse food for the week ahead . Given the timing of our study, participants frequently discussed the impacts of COVID lockdowns. Without the ability to earn a daily wage, many families shared that they were unable to feed their children. For some families, lack of food at home meant that any provision of food at a childcare center was reason enough to send them there, despite other factors: "I saw other children being given porridge and being taken care of well but some of the teachers were pinching the children, and beating them. But for me, of importance was food. I thought they were feeding them better." . Financial challenges of obtaining sufficient cooking fuel, as well as care in cooking food, were identified as constraints to cooking food appropriately and safely. These concerns resulted in parents packing their own food for their children. "Most of the childcare centre [providers] simply cook rice and beans for the children. What I don't know is if the beans are usually well-cooked or not. With rice there is no problem but as for beans you never know if they are given cooked beans. So we preferred to pack food for them." . --- Health Given the environment of the informal settlements, many parents did voice concerns about the ability of the centers to maintain hygiene and a clean environment for children to eat, sleep, and play. In particular, they were concerned about changing children's diapers, providing clean and sufficient mattresses, floor coverings and bedding, keeping children clean, ensuring hand washing, and provision of clean drinking water. "They should also get a water tank where they can store water for using, like for children to wash their hands." . The co-design workshops with CHVs and CHAs reinforced parents' concerns about the centers. Participants commented on how dirty and unhygienic many of the centers were, particularly the home-based centers where the available space was frequently a single room where the center provider lived and also operated the business. Participants shared that: "You find children sleeping in one corner, there are used potties in another corner and food in another corner" . Parents were particularly concerned about the hygiene standards in storing and preparing food in the centers. They frequently expressed concerns that the childcare provider was not storing the food in a clean area and protecting it from contamination. This was mentioned most frequently in relation to the home-based centers. "When I left my child with neighbours, every week my child would fall ill with common colds and other diseases such as coughs. During cold season, he would be left to play with water and then the next week, he is unwell. When I started to take him to a daycare centre I stayed for about two months, for the first time, without having taken him to the hospital." . Underlying this theme was the overriding influence of the slum context and the economic hardship facing families. Several parents explained how they were using childcare centers that they felt were a risk to their child's health simply because they were cheap and what they could afford. Frequently, the challenges facing the centers were due to the poor slum environment around the centers. "They border a sewer line. Before they fenced the school, children used to fall into the sewer. Then the children have to breathe in the fumes from that sewer line. The fumes make them sick; my children have blocked noses from that" . Regarding minor ailments, injuries, and emergency care, parents and providers described the challenges of getting appropriate and timely care. For parents working long hours far away from home and the childcare center, there was great concern that their children would remain untreated until they returned from work when it would be too late to find a healthcare center open. Several parents recommended that childcare centers should keep first aid kits and basic medications. "I told you about a pharmacy in the daycare for when a child is not feeling well. When they call you when your child is not feeling well you leave everything to come and take your child to the hospital but if they would have a pharmacy there to give first aid you can come in the evening and see what you will do" . During the interviews with providers, most focused on the challenges of running their centers rather than the health of the children. However, the questionnaire results indicated that 97% of providers felt they knew what to do if a child was sick and 83% said they checked the children's health each day and knew the immunization status of the children in their care. Several of the more experienced providers from both home-based and larger center-based care mentioned referring parents or even taking the children themselves to local clinics for immunizations or check-ups: "You find the child is underweight and the child doesn't have a good diet, there is a hospital called 'Provide, we advise the parent to take the child within Korogocho and if she is lucky there is these protein sachets that they are given to eat so they regain the weight'." . The need for providers to be trained to manage child health, including checking immunization and knowing what to do if a child was sick, was identified as an important part of the support needed by childcare providers by 95% of the 22 CHVs completing the KAP questionnaire. --- Safety and infrastructure A priority for parents during Workshop 6 and a frequent theme within the qualitative analysis was the safety of their children. The high level of crime and low social capital within the slum environment meant parents were scared to leave their children with neighbors, and few had relatives nearby to provide trusted care. "When I go to hustle for a job, leaving her with another person is another source of stress. You are better off leaving her in a place where you know you will not find him in danger. You know neighbours [xxx] have many issues, so it is better to leave the child where you are assured of his safety" . "I trust her because when I go to work, I can leave her with my baby and I know my baby will be safe. I have to leave her in a place that I deem safe, she will not stay hungry. Those are what I look out for. A place where she will not be defiled, and she will not stay hungry." . "When I am at work, like I said she should be given food, when she falls asleep, she is put to bed, and when I am at work, I don't fear news such as, 'your child has been defiled'." . "I saw a certain lady who used to take her child there and I wanted a place where my child could spend the whole day, so that I would be secure in knowing that my child was safe in a certain place, where I can go and pick him from when I am back. I didn't even bother about the environment. All I cared about was a place where I could leave my child." . These concerns for the safety of the child were expressed by providers, particularly in relation to managing the collection of children. With parents working long hours, agreeing on who should pick children when the parent was delayed was a real concern. Providers who had been running centers for many years explained their policies for pickups. However, managing these issues was particularly challenging for smaller home-based and inexperienced center providers. "The parents bring them and when they leave, they come back so late. We agree with her in the evening they can't tell anyone to come and pick the child in the evening, so she has to come with the one who will be picking her child and tell me when they are late this person should take my child" . The crowding of the childcare centers, particularly the homebased ones, was frequently highlighted by parents as a situation that undermined the safety and healthy development of children. --- Management challenges The harsh economic realities facing families in informal settlements meant that many providers struggled to run their childcare centers as viable businesses. There were many examples of providers working long hours, accepting under-payment and late payment, and providing food for children when their parents could not do so. The provider questionnaires showed how home-based providers worked particularly long hours with some starting as early as 5 am and closing at 9 pm. Furthermore, 84% of home-based providers opened on Saturdays compared to 50% among larger centers and 14% among faith-based centers, and only home-based centers operated on a Sunday. Even providers who had been in business for over 5 years explained how they frequently took responsibility for the children when parents struggled to pay, feed, or pick them up on time. The need for providers to play this role was clearly heightened by the economic conditions during the COVID pandemic. However, the unreliability of informal work and lack of economic opportunities were factors identified as influencing the financial viability and management of the centers independent of the pandemic. "So, let's say it's a job, the parents go and work overnight and the next day they come at around 11 a.m. Sometimes even if they left their phone number, it is not useful because they change the number so that you can't reach them. So, I have passed through this with like five parents. In fact, I am forced to stay with them [the children] like mine and when they return, they don't give me thanks." . Managing centers in these conditions was clearly challenging. For instance, home-based centers had fewer good management practices such as weekly budgeting, tracking income and expenditure, and keeping an attendance register . --- Community health systems and support for childcare centers The qualitative interviews and mapping exercise illustrated that many childcare providers were aware of the CHVs, and several mentioned that they already visit the centers to check the children's health. Similarly, all but one of the 22 CHVs completing the questionnaire reported supporting childcare providers and 62% reported visiting weekly . "When the CHV's are making their rounds as part of their check on children's health, often you find them going to daycares as they know they will get children in the daycare centers. So, they have to visit me and check them and if they find out the child has poor health, they tell me what to do." . The idea of using the CHVs to deliver the intervention was discussed with county and sub-county officials and considered by all stakeholders during the co-design workshops. Ultimately, it was agreed that CHVs would be the most appropriate members of the community health team to be trained to deliver the intervention. This was reiterated in the workshop where participants felt that as CHVs were already visiting households, and many were already including childcare centers in their rounds, this would not be an extra burden. "It will not be too much work because the day care centres are in the community so the CHVs can work in them as they do their other duties. Some of the childcares are part of homes and the CHVs do household visits so for childcares within the household environment, it will be easy to visit. The CHVs live in the community, and they know where the childcare centres are. If the CHVs are facilitated, they can do the work well. However, they need to be trained on what they should look out for." . The CHVs and their CHA supervisor agreed that supporting child health and development in the childcare centers was within the CHV's remit. However, they identified resources, training, and systems adaptations that would be required to support implementation. The questionnaire results concurred with the 22 CHVs and showed that they felt motivated to support the centers, with 16 feeling strongly motivated. The most common barriers identified by the CHVs were their lack of tools and resources , lack of training , and perceptions of low interest among center providers . "CHVs will be able to support the project but for them to do so, they will need training on the project. They will also need stipends, referral books, field supplies such as gum boots, raincoats, bags and umbrellas; they will also need hand washing equipment to place in the day care centres; they will also need certificates and badges from the project/organization." . "The public health reporting tool does not have a component to capture childcare centres so this can be added. Currently there are no data on childcare centres." . Sub-county and community health teams were quick to identify the potential of the intervention to improve the health and early childhood development outcomes of children, particularly through the early identification of disabilities such as hearing impairment and autism, increased awareness of child rights and action to protect children from abuse , and the potential to increase uptake of immunization, vitamin A, and deworming programs and improve providers skills to assess sick children and know what to do in case of a health emergency . They also highlighted challenges in implementing the intervention, particularly due to the instability within informal settlements with both families and childcare centers frequently moving locations due to evictions and the continual struggles for daily survival and economic opportunities. Some CHVs felt that training on income-generating activities for childcare providers could help to overcome these difficulties, and all agreed that some component of training to support providers to run their centers as viable businesses would be an important component of the intervention. During workshop 3 with representatives from Nairobi Metropolitan Services , the challenges of operationalizing the current Nairobi City County Childcare Facilities Act 2017 were highlighted, in particular the registration of childcare centers within informal settlements. The Act specifies a minimum standard for facilities, staffing, and training for childcare centers, and the center must meet these standards and pay a registration fee. "Some of these centres are not registered and therefore no one has gone to assess the quality of service they provide because no one even knows where they are because they have never been mapped." . --- Intervention design The co-design workshops and rapid analysis of the qualitative data informed the development of a list of topics to be covered in the training. The process of close engagement with childcare providers, CHVs, and parents built the teams' understanding of the operational context and how the intervention should be delivered. It was evident from the interviews that there was diversity in the levels of experience and knowledge of the childcare providers. Their need to work long hours providing care, particularly for the home-based center providers, was also evident. Deliberation on these factors within the workshops led to many design decisions. First, a CoP approach which allows providers to share experiences and approaches and to apply good practice to the context of informal settlements would be more appropriate than a didactic training course. Second, the challenging and unstable economic conditions of the informal settlement and limited business experience of the providers, particularly the home-based providers, would need to be a component of any support to improve the quality and functioning of the centers. Third, CHVs would be well placed to facilitate these CoP meetings due to their existing role in the community, but they required further training particularly in areas of early childhood development, child safety, and business management, with top-up training on child health, hygiene, and nutrition. Fourth, given the hours worked by childcare providers, particularly the home-based centers, CoP meetings should be held at the weekend with compensation paid to providers to enable them to pay a back-up career during their absence. Fifth, support to individual centers was required between the CoP session, and CHVs were well-placed to include these supportive supervision sessions within their regular household rounds. The details of the intervention are provided in Table 7 below, using the headings provided by the Template for Intervention Description and Replication to describe each aspect. The implementation flow of the intervention over the 4-month feasibility study is provided in Figure 4 and the theory of change is shown in Figure 5. --- Discussion Our results show the presence of a high number of childcare centers in Korogocho and Viwandani which is an indication of the growing demand for childcare services in these informal settlements. The qualitative and quantitative findings highlight the challenges facing these centers within every domain of the nurturing care framework. Our findings reflect those of the limited number of other studies that have explored the quality of childcare in informal settlements . Other studies within the African context have assessed quality and the impacts of types of preschool environment on childhood development , finding the importance of play materials, infrastructure, and training of providers on ECD. Previous work in informal settlements in Kenya found an average child-to-adult ratio of 1:22, limited nutrition, poor sanitation and facilities, and untrained caregivers . The informal homebased centers were often identified by parents and community health workers and volunteers as having the most limited facilities, environment, and poor practice in ECD and health. Similar quality issues have been reported in other studies that have examined the provision of childcare within similar urban settlements in Africa . The ability of these centers to provide "nurturing care" is severely limited, leaving parents anxious about the quality of childcare and undermining child health and development. The Nurturing Care Framework provided a valuable framework for identifying appropriate content for the intervention. We identified significant gaps in center resources, practices, and knowledge of providers in all areas of the framework. This was the case across all types of providers, particularly the small informal home-based and center-based providers. However, the inputs for all participant groups through the co-design workshops were key in determining the focus and depth of the topics within each domain of the NCF. The decision to follow a CoP approach was reinforced by the findings from parents and childcare providers on the diversity of approaches and the considerable strengths and levels of experience among the providers, with clear benefit to be achieved from sharing these experiences through group CoP sessions. One element emerged that is not emphasized within the NCF and not originally considered in depth by the team was the importance of supporting center management. The challenges of childcare providers attempting to balance limited incomes with high expenditures, particularly following the economic impacts of the COVID lockdowns were very evident in the qualitative interviews and co-design workshops in our study. Previous studies have noted an association between center management capacity and center quality , yet management and leadership are rarely considered in the assessment of the quality of childcare centers . There is limited evidence on how to support childcare centers, particularly smaller home-based informal center to improve healthy childhood development, particularly in LMICs . While Kenya has a strong track record in ECD policy, starting with the ECD Policy Framework of 2006 and encouraging recent attempts to bring multiple sectors together to strengthen the response to ECD , there is clearly a wide gap between policy and practice. This is particularly the case in Nairobi where the County government leads policy provision in the country with the Nairobi City County Childcare Facilities Act 2017 which clearly specifies standards for childcare services. We found that most informal childcare centers were unregistered and only received ad hoc visits from community health teams. Our study highlights the need to ensure that policies and strategic plans reflect the realities facing the growing number of informal home-based and small private centers and can realistically be implemented within low-income neighborhoods. Implementing and scaling up the registration process linked to achievable standards, combined with support from CHVs and the community health teams, has the potential to bridge this policy and practice gap. Our intervention development process has much in common with the characteristics of co-design described by Vargas et al. where multiple stakeholders actively collaborate to design solutions to prespecified problems. This contrasts with a process of "co-creation" where diverse stakeholders are deeply involved in the process of understanding problems and evaluating solutions , a process with similarities to participatory action research . Within the context of our intervention development process, we took as our starting point the consensus and evidence base underpinning the Nurturing Care Framework and the need to ensure the intervention could be delivered sustainably with the existing system due to the lack of any additional resources. Our codesign process focused on developing an intervention within these parameters was preferable. We were cognizant of the need to ensure that the intervention was well-integrated into existing systems and accepted by parents and childcare providers. We were particularly conscious of the need to ensure CHVs were not overburdened and received sufficient training and supportive supervision as these factors have been identified as undermining their wellbeing and willingness to continue their vital work . Our qualitative findings and the discussions in the co-design workshops highlighted that some CHVs were already visiting childcare providers but in an ad hoc manner and with limited knowledge of how to support the centers. The lack of organized support, training, and regular contact of childcare centers with the community health services was an important finding and influenced co-design decisions to focus the intervention on the informal, home-based childcare centers. Agreement across community health teams, and local and national government of the potential contribution of CHVs in providing structured support to childcare centers, was a major step in the co-design process. Recent government plans to provide a stipend of KES 2,000/month for CHVs and Nairobi county's commitment to their community health bill bodes well for improved functioning and sustainable functioning of the CHV program . We developed a highly participatory process to co-design ensuring there were multiple ways to hear the voices of the different tiers of government, CHVs, their supervisors, as well as a range of childcare providers and parents. By using qualitative interviews as well as the co-design workshop and knowledge, attitudes, and practices questionnaires, we felt able to shape the intervention to fit well into the health system and community context. In addition, we were very aware of the important role played by NGOs in filling many of the service gaps in informal settlements. Our partners in this study, Kidogo provided detailed guidance and resources. However, scaling up and sustaining initiatives to improve the quality of childcare is vital in the context of rapid urbanization . Our study will also publish findings on the feasibility of implementation. However, further research is needed to assess the sustainability and effectiveness of such interventions to improve child health and ECD and contribute to wider outcomes such as women's participation in the workforce and household income and wellbeing. --- Strengths and limitations This study had some limitations as the COVID-19 pandemic hit the country in March 2020 when study implementation was just taking-off. This meant that co-design workshops and baseline qualitative interviews were conducted virtually and elements of the co-design process, which were planned as highly participatory group events with the use of video, had to be simplified to be conducted virtually or in accordance with social distancing requirements. The pandemic period also resulted in the closure of some childcare centers. This means that our estimates of the number of centers and the number of children using the centers may not reflect non-pandemic conditions. Few studies have explored childcare conditions in informal settlements. Our long-term engagement using both quantitative and qualitative methods as well as the co-design process enabled deeper insights into the assets and challenges facing different types of childcare providers in informal settlements. An additional strength of our study is the co-design approach which enabled the complete engagement of national and county governments, parents, childcare providers, and community health teams. The engagement of these multiple stakeholders was a prerequisite to the design of an intervention embedded within government systems to enable potential scale-up and sustainable delivery. --- Conclusion Implementing a co-design process embedded within existing community health systems and drawing on the lived experiences of childcare providers and parents in informal settlements facilitated the development of an intervention with the potential to be scalable and sustainable. Such interventions are urgently required as the number of home-based and small center-based informal childcare centers are growing rapidly to meet demand; yet, they receive little support to improve quality and are largely unregulated. Given the context of informal settlements, support to strengthen management within the centers is vital in addition to the core domains of WHO's Nurturing Care Framework. Further research on the effectiveness and sustainability of support to private and informal childcare centers in the context of low-income urban neighborhoods is needed. --- Data availability statement The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author. --- Ethics This study was approved by Amref Health Africa's Ethics and --- Ethics statement The studies involving human participants were reviewed and approved by Amref Ethics and Scientific Review Committee. The patients/participants provided their written informed consent to participate in this study. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. --- Supplementary material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2023. 1195460/full#supplementary-material

Background: Globally, million under-s do not have adequate childcare. This may damage their health and development and undermine societal and economic development. Rapid urbanization is changing patterns of work, social structures, and gender norms. Parents, mainly mothers, work long hours for insecure daily wages. To respond to increasing demand, childcare centers have sprung up in informal settlements. However, there is currently little or no support to ensure they provide safe, nurturing care accessible to low-income families. Here, we present the process of co-designing an intervention, delivered by local government community health teams to improve the quality of childcare centers and ultimately the health and development of under-children in informal settlements in Kenya. Methods: This mixed methods study started with a rapid mapping of the location and basic characteristics of all childcare centers in two informal settlements in Nairobi. Qualitative interviews were conducted with parents and grandparents (n = ), childcare providers, and community health teams (n = ). A series of co-design workshops with representatives from government and non-governmental organizations (NGOs), community health teams, and childcare providers were held to design the intervention. Questionnaires to assess the knowledge, attitudes, and practices of community health volunteers (n = ) and childcare center providers (n = ) were conducted. Results: In total, childcare centers were identifiedin Korogocho and in Viwandani. School-based providers dominated in Korogocho ( %) while home-based centers were prevalent in Viwandani ( %). All centers reported minimal support from any organization ( % supported) and this was particularly low among home-based ( %) and center-based ( %) providers. Home-based center providers were the least likely to be trained in early childhood development ( %), hence the co-designed intervention focused on supporting these centers. All co-design stakeholders agreed that with further training, community health volunteers were well placed to support these informal centers. Findings showed that given the context of informal settlements, support for strengthening management within the centers in addition to the core domains of WHO's Nurturing Care Framework was required as a key component of the intervention.

BACKGROUND More than three million refugees have been resettled in the United States since 1975 with more than 58,000 resettled in 2012 . Since 2011, many of the resettled refugees have come from African countries, primarily Somalia, Liberia, Sudan, Ethiopia, and Burundi as determined by the processing priorities of the United States Refugee Admissions Program . The challenges that refugees face include trauma exposure, family loss and separation, poverty, learning English, low-wage work, troubled schools, inadequate housing, social and cultural transition, discrimination , and gender inequalities . Prior research has documented the consequences of these challenges for the mental health of refugee adults and children, including post-traumatic stress disorder , depression, anxiety, somatic problems, alcohol and substance abuse, domestic violence, learning difficulties, and other behavioral and family problems . Adolescent refugees in the U.S. are a particularly vulnerable group because their exposure to trauma and displacement coincides with key developmental transitions. Adolescents are at a time in their lives where they are experiencing shifts in responsibilities, identity explorations, and perhaps role confusion . Adolescent refugees in the U.S. may experience a combination of both the child to adult transition familiar from the culture of their home country, and the adolescent transition seen in U.S. populations. Studies of adolescent refugees in resettlement show high rates of psychopathology . Several longitudinal studies of psychiatric symptoms have shown high rates of PTSD and depressive symptoms that generally diminish over time . Adolescent refugees with strong parental support and community engagement have diminished negative behavioral outcomes . Additionally, the adaptability that is part of adolescents' development, may assist adolescent refugees in better adjusting to their new environments . A prior study characterized the patterns of psychosocial adjustment among adolescent African refugees in U.S. resettlement . A purposive sample of 73 recently resettled refugee adolescents from Burundi and Liberia were followed for two years and qualitative and quantitative data was analyzed using a mixed methods exploratory design. The researchers first inductively identified 19 thriving, 29 managing, and 25 struggling youths based on review of cases. Multiple regressions indicated that better psychosocial adjustment was associated with Liberians and living with both parents. Logistic regressions showed that thriving was associated with Liberians and higher parental education, managing with more parental education, and struggling with Burundians and living parents. Qualitative analysis identified how these factors were proxy indicators for protective resources in families and communities. The present study drew from the ethnographic data of that same study and aimed to build knowledge on the factors and processes that might account for differing patterns of adolescent refugee's adjustment, specifically their psychosocial well-being, which was defined as a state of mental and social well-being which is more than the absence of psychiatric symptoms or disorder. --- Protective Agents, Resources, and Mechanisms Conceptual Framework This research was grounded in family ecodevelopmental theory . Family ecodevelopmental theory envisions adolescents in the context of family systems and community networks interacting with educational, health, mental health, social service, and law enforcement systems. Ecodevelopmental theory explains how risk and protective factors may interact across adolescent's multiple social environments . Additionally, the study also drew upon trauma theories that explain the impact of and meaning of trauma exposure , and migration theories that explain why people move and how mobility impacts them .We also drew upon resilience theory, which explains a person's or community's capacity to withstand or bounce back from adversity . Specifically, family therapists have used resilience to refer to domains of family life that may contribute to family adaptation after adversity, including shared family belief systems, family connectedness, and communication processes . Based upon these theories, we first defined three new constructs so as to identify and deliberately distinguish between those which protect adolescents , their capacities for protecting adolescents , and how they do it . For the purposes of this study, these constructs were defined as follows. Protective Agents are those individuals, groups, organizations, and systems that can contribute either directly or indirectly to promoting adolescent refugees' psychosocial wellbeing. Protective agents can include parents, schools, and churches. Protective Resources are the family and community capacities of protective agents that can promote adolescent refugees' psychosocial well-being. Again, this means not only that they can stop, delay, or diminish negative mental health or behavioral outcomes among adolescent refugees, but also that they contribute to positive growth and development in a broader sense. Protective Mechanisms are the processes fostering adolescent refugees' competencies and behaviors that can promote their psychosocial well-being. Protective mechanisms are really about the unfolding of positive changes over time for adolescent refugees. A conceptual framework was devised that maintained the distinction between protective agents, protective resources, and protective mechanisms as it endeavored to characterize their interactions. It posited that: 1) protective agents and resources exist in refugees' family and social environments to protect against negative outcomes from their exposure to risks through war, migration, and resettlement; 2) through protective mechanisms, these protective agents and resources mitigate the family and ecological risks for negative individual behavioral and mental health consequences and promote psychosocial well-being. This framework, grounded in the previously mentioned theories, led us to focus this investigation on characterizing the protective mechanisms by which the protective agents and their protective resources yielded positive changes among refugee adolescents. It also informed our data collection and analysis in several ways. One, it helped in identifying domains of interest for the interviews and observations, including trauma experiences, family connectedness, and support networks in the community. Two, it helped in formulating questions used in minimally structured interviews and choosing sites and persons for observations. Three, it helped in identifying a priori domains of interest that focused the qualitative data analysis . --- Prior Studies of Refugee Youth To further inform this study of protective agents, resources, and mechanisms, we also reviewed prior studies of refugee youth that identified protective resources. Several studies have identified protective resources that contributed to the psychosocial well-being of refugee youth. These included higher: acculturation, social support, collective self-esteem, strong ideological commitment, and family connectedness . Regarding school performance, other studies have identified: sense of belonging to family, school, and community and positive perceived school performance and commitment to education . Protective resources associated with refugee youth's school adaptation and outcomes include higher: acculturation, social support, social integration, goal commitment, and appropriate grade placement, as well as having healthy parents . Protective factors against psychiatric disorders include: having connectedness to family, peers, and the larger community, social support, strong religious beliefs, mothers with greater duration of education, higher socioeconomic status, and active community involvement in residential school system . These prior studies informed the conceptual framework, described in the previous section, which shaped data collection and analysis in several ways. One, data collection and analysis focused on protective agents in both family and community domains. Two, throughout the analysis, a distinction was maintained between those who protect adolescents , their capacities for protecting adolescents , and how they do it . --- Burundian and Liberian Refugees Burundians-In 1972, Burundians from the Hutu ethnic group fled a violent campaign from the Tutsi controlled government that led to the death of approximately 200,000 Burundians and the flight of 150,000 refugees to Tanzania, Rwanda, and the Democratic Republic of Congo . The genocide led by Michel Micombero, an army captain turned dictator, was designed to kill all educated Hutus older than fifteen years of age in order to prevent Hutus from coming to power . Micombero was overthrown in 1976 by Lieutenant Colonel Jean-Baptiste Bagaza who continued to suppress and deny education to the Hutus . Living in exile in Tanzania for three decades, the so-called 1972 Burundian refugees experienced ongoing political and criminal violence, sexual assault, poverty, unemployment, dependency, no freedom of movement, family break-up, and poor education for children. While Burundians had access to food within the refugee camps, the lack of money caused many parents to send their oldest son to the capital, Dar es Salaam, to find a job . Although illegal to move between refugee camps, some refugees were able to falsely identify themselves as Tanzanian . Ten thousand Burundian refugees from isolated Tanzanian refugee camps were resettled in the U.S. beginning in 2007. Many Burundians were unable to repatriate because of fear and the Tanzanian government's unwillingness to permanently resettle all Burundian refugees due to a large influx of refugees in the 1990s . In relocating to the U.S., Burundian adolescents faced a unique resettlement paradigm because they were born and raised in refugee camps and lacked direct connections to their parents' home country of Burundi. Burundian refugees in the United States have been resettled in small numbers and it is likely they do not have close relatives living in the United States . Liberians-From 1989 until 2003, Liberia suffered a series of ethnic conflicts among armed groups. Rebel forces, supported by Libya and Burkina Faso, led by Charles Taylor and Prince Johnson, tortured and killed Liberia's president Samuel Doe . It is estimated that up to forty percent of the combatants were child soldiers, under the age of eighteen with the youngest soldier documented to be six years old . An estimated 200,000 people were killed during the civil war and hundreds of thousands were forced to flee to other countries such as Guinea and Ivory Coast in 1990. The World Health Organization stated that at least two-thirds of high school students in the capital city, Monrovia, had witnessed torture, rape, and murder, and were psychologically scarred . From 1992 to 2004, the U.S. resettled approximately 23,500 Liberian refugees who fled the war. Similar to Burundians, recently resettled Liberians did not have many family connections in the United States, with less than one third of those resettled in 2004 being reunited with relatives . In addition to family separation, they experienced economic, social, and cultural pressures in resettlement . Unlike the adolescent Burundians who were raised in Tanzanian refugee camps, many Liberian adolescent refugees experienced or participated in the war adding to the complexity of resettlement and integration into the American society. --- Study Aims This longitudinal ethnographic study addressed the following two research questions: 1. What protective agents and protective resources appear to contribute to psychosocial well-being in adolescent refugees? --- 2. What are the protective mechanisms through which these protective resources appear to contribute to psychosocial well-being in adolescent refugees? Based upon the research findings, in the discussion the authors consider: How might policies, programs, and research better promote the protective agents, resources, and mechanisms that can contribute to refugee adolescents' psychosocial well-being? --- METHOD Data Collection This was a two-year, multi-site, longitudinal ethnographic study of Burundian and Liberian refugees resettled in Chicago, Illinois and Boston, Massachusetts. Study subjects were 73 purposively sampled at-risk refugee adolescents, their families, and service providers, interviewed within the first three years following resettlement. The sample consisted of 37 Liberian and 36 Burundian adolescent refugees with comparable numbers of Burundians and Liberians in Chicago and Boston. Based on a previous publication, there was no statistically significant difference between the psychosocial well-being of the adolescents resettled in Boston verses Chicago . The entire adolescent refugee population could be considered at-risk for psychosocial difficulties due to their exposure to trauma and displacement. However, for the purpose of this study, at-risk was defined as refugee youth with one or more of several specific factors that have been empirically associated with mental illness or behavioral problems in published studies of migrant youth . These were: 1) a one parent family; 2) poverty (monthly family income below U.S. Census poverty threshold ; 3) living in a linguistically isolated household ; 4) a mother or father with less than a high school education; 5) a parent who has sought or received mental health treatment . They were recruited from the refugee communities through refugee service organizations in metropolitan Chicago and Boston by fieldworkers from those communities. The purposive sampling strategy was intended to maximize diversity on three axes: 1) gender; 2) age; 3) Burundian or Liberian. All participants gave written informed consent as approved by the institutional review boards of the University of Illinois at Chicago College of Medicine and Harvard Medical School. Data collection consisted of minimally-structured interviews and shadowing observations of individual study participants, and focused field observations carried out with each family in homes, communities, and service organizations. Minimally-structured interviews consisted of discussions with the participant that began with a small number of introductory questions . The interviews included addressing the following domains: 1) Family and ecological protective factors; 2) Protective mechanisms linked to outcomes; 3) Experiences of outcomes by at-risk refugee adolescents; 4) Sociocultural contexts of protective factors, protective mechanisms, and outcomes. The conversation proceeded in whatever direction allowed the participant to speak most meaningfully to the research questions from his/her personal experience. These interviews also gathered basic demographic characteristics which enabled some quantitative exploration as described below and displayed in Table 1. Shadowing field observations involved the ethnographer accompanying the family or its members on his/her normal daily routine in a variety of sites . Shadowing observations allowed the ethnographers to directly witness the interactions between protective resources, risks, culture, and service sectors over time. Over the two-year study period, for each adolescent subject, we conducted six individual interviews, four hours of shadowing observation per quarter, plus four interviews with their family members and two interviews with their providers. The research team included several Liberians and several Burundians who were familiar with the community being studied, one psychologist, one nurse, one social worker, two public health academics, and one human development academic. The interviewers spoke the same language as the participants and were trained in ethnographic methods and supervised by the principal investigator and co-investigator, a psychiatrist and a medical anthropologist. All interviews and field notes were transcribed in English. Data was collected and analyzed based upon well-established approaches to ethnography and qualitative analysis . Because this study was trying to capture changes over time for each of the families, the investigators applied a case study research design whereby the data for each youth were read sequentially so as to identify factors that changed over time . --- Data Analysis The initial study questions were refined through an iterative process of data collection and analysis that followed standardized qualitative methods utilizing a grounded theory approach to data analysis . We coded the same sections of text and compared their results, discussed, and resolved disagreements until a level of .80 agreement was reached. After establishing coder reliability, all transcripts were coded by multiple coders using an initial coding scheme agreed upon by the entire research team. The codebook consisted of codes that corresponded to demographic, contextual, and experiential matters, including gender, ethnicity, family, school, and trauma. Then, through pattern coding and memoing, the analysis formed typologies used to produce a grounded theory model. The different sources of data including adolescents, parents, and providers were first analyzed separately, and then they were integrated to enhance the value of multiple perspectives. Not surprisingly, the data collected provided evidence of the challenges and difficulties that refugees face . However, these were not the focus of analysis, which was instead on building knowledge on possible protective agents, protective resources, and protective mechanisms. This model is described in the text and represented in Table 2 and Figure 1. All the findings were reviewed by the entire team to check for contrary evidence until consensus was achieved. The team determined the trustworthiness of the data based on the following techniques: 1) prolonged engagement by investing sufficient time to ensure accurate understanding of the phenomena being studied; 2) triangulation by using multiple sources, methods, theories, and investigators during the research process . --- RESULTS --- Protective Agents Nine protective agents were grouped into three categories: 1) Youth; 2) Family; 3) Service Providers. These individuals, groups, organizations, and systems contributed either directly or indirectly to potentially promoting adolescent refugees' psychosocial well-being. --- Youth Friends and peers-As for many teenagers, friends and peers played a central role in the daily lives of refugee adolescents. Because many of them lived, prayed, and went to school in neighborhoods heavily populated by other refugees and immigrants, their friends often came from their own ethnic community or another facing similar challenges. A 15-year-old Burundian boy talked about what he does with his friend, "J. is a close friend. We talk about everything that happens, do homework and play soccer and basketball together." Most of these adolescents did not have friends who were part of the mainstream society. --- Family Parents-The adolescent refugees resettled in the U.S. with one or both parents or with other family caregivers. Parents provided their teens with emotional, material, and educational support, monitoring and supervision, cultural connections, and access to faith communities. A Liberian parent described the importance of family time for their children, "Once in a while my parents are here so we go and visit their grandparents. And they want to go to the pool, and I take them swimming. And they want to go to the park, and we go and play games from back home. And they like to go out, so once a month we go out to dinner at the restaurants around here. Just change the environment once in a while if I can afford it, I just want to make them feel happy once in a while." However, most parents reported that they were highly burdened by concerns about their own well-being, family finances, and learning English, and these detracted from helping their children. Older siblings-Some adolescent refugees had older siblings living with them who looked out for them in neighborhoods and schools, and served as cultural and language brokers. A 15-year-old Burundian boy explained how he helped his younger brother with schoolwork, "My younger brother knows division but he does not know multiplication...Yes, we help each other like that. I don't show him the final result. I show him how to do it and he does it himself. I do that wanting to see if he has understood what I showed him." Of course, older siblings shared that they also faced their own barriers to learning and adjusting. Extended family members-Some adolescent refugees had extended family members close by or elsewhere in the U.S. These extended family provided family connectedness as well as material support. In some cases, when teens faced trouble at school or with the law, their parents sent them to live temporarily with their extended family. A Burundian mother explained how the children's grandparents lived in a different state but still provided material support to the family, "The children's grandparents in New Jersey send presentsclothes, shoes, and house things like plates." However, extended family expressed that they often had their own worries and immediate family to provide for. --- Service Providers School teachers, staff, and coaches-Refugee adolescents worked with bilingual education or mainstream teachers. For most this took place in public schools, but for some it occurred in parochial schools or special schools for immigrants and refugees. The teachers' work with adolescent refugees was not just instruction, but also counseling, family support, and conflict resolution. One teacher described her work with adolescent refugees as, "Success for any of my students would be for them to be able to understand their reality, and be able to manipulate and control it, to really enjoy engaging in the pursuit of knowledge. I really want these kids to have the foundation, so they have a chance in high school. These kids need the basic skills to help them, I mean, they have amazing insights, they have so much more lived experience than I do." However, teachers and staff complained that they lacked specialized knowledge and skills on how to work with refugee children in resettlement, even though some had years of experience. Church staff and congregants-Church members from multiple denominations provided spiritual guidance, material support such as rent money, food, and vehicles, afterschool and social activities, and case-workers for adolescents experiencing trouble in school or at home. A 14-year-old Burundian boy explained, "This priest with two Ugandan fellow Christians have supported the family. The priest once in a while he gives food for the family, the two Ugandans have helped the father to get to the hospital when he was sick couple weeks ago." However, church members stated that they lacked knowledge and training on how to work with refugee children in resettlement. Resettlement agency caseworkers and activity leaders-Agency staff members were responsible for managing the logistics of resettlement which included travel, housing, initial material needs such as food and clothing, employment, job training, and language instruction. Agency caseworkers also connected families to financial resources, medical aid, after-school programs for youth, and tutoring, and helped youth get settled at their new schools. A Burundian parent explained how their caseworker has helped, "C is the family contact for health, financial, and school issues, he's the one who registers all the kids in a public school." Resettlement agency caseworkers reported that they had limited time and resources for helping refugee families during resettlement and were often seen by the families as withholding money and not on their side. Volunteers-Sometimes resettlement agencies or churches offered adolescent refugees and their families help through volunteers and sponsors. They augmented the support provided by resettlement agency staff by offering primarily companionship as well as help with paperwork, childcare, and financial assistance. A Liberian mother explained how a resettlement agency paired her with a volunteer, "She was my volunteer from the agency and for 3 months they pay your rent but you have to do everything on your own and they help you find a job and everything. And they put my son in school." As volunteers, these persons reported that they often lacked time to provide ongoing support to refugee families after their initial resettlement. Health and mental health providers-Teens were seen for routine care, or referred for special needs by teachers, school counselors, church staff, volunteers, and resettlement agency staff. An assistant principal explained how helpful the school social worker was, "I know that Z's dad has spoken to the social worker, but the problem is that she is the only social worker for 1200 kids. And she sees many kids individually with severe problems." The health and mental health providers working with the refugee teens often reported being understaffed, under-supported, and under-trained to provide adequate care and support. --- Protective Resources The following eight family and community capacities appeared to potentially promote psychosocial well-being in the adolescent refugees. Finances for necessities-Refugee families who had money for basic needs such as food, clothes, and school supplies, either via employment, entitlements, or community donations, appeared more able to provide stability and support for their teenage children. Parents who could afford these necessities felt more secure and were more confident providers. A Burundian father explained, "I decided to buy a computer with Internet access. She doesn't have to go to the library to do her homework. I want to avoid any kind of excuses. I want her to have what other American children have...if they can do it, she should be able to do it." English proficiency-Arriving in the U.S. already being able to speak English or learning English quickly upon resettlement helped refugee parents to find jobs, navigate social services for their families, and support their children's schooling. A Liberian father was better able to assist with his child's schooling by being able to communicate with the teachers, "The father has been to the school two times and he talks on the phone with the teachers frequently. He said that talking to the teachers was helpful because they tell him what is going on with his son and when he's good." Families who arrived without English were able to attend English language classes at resettlement agencies or at community organizations. Social support networks-Refugee families who were connected to and supported by members in their community, including friends, sponsors, church staff and congregants, and social service agencies, were better able to access and share information and resources for their adolescent children. For example, they were able to access tutoring and after-school activities for their children, or jobs that helped with family finances. A Burundian father stated, "The Burundians I found here introduced me to the agency. They helped me find a house and the agency paid the house rent for three months. They helped me find a job and had been giving me a ride to and from job until they contributed money for me to get me own car." Engaged parenting-Refugee parents who adapted their parenting styles and expectations through increased involvement and parent-child communication with their teenage children appeared better able to help their teens address issues of school and community adjustment, and making friends. A 15-year-old Liberian boy expressed the importance of his father's involvement in his life; "He always comes to things that I need him for. If I'm in school and I need a ride, I always call him, and he can come. Or a parent conference or something like that. Like yesterday, they had an honors breakfast, and I was invited, so he actually went there with me." Family cohesion-Families that shared positive family relationships with open communication, spending time together, and flexible attitudes, had more potential to add structure, security, and normalcy to their teens' lives during resettlement. A Liberian mother stated, "We have family meetings and we all get together and talk about the problems that they are having with one another." Cultural adherence and guidance-Some families encouraged their adolescents to hold on to their traditional culture to prevent their teens from developing negative behaviors that are more associated with American culture. This was especially so for daughters. A Burundian father explained, "It's not good to lose the culture and I've been telling this to my daughters." Although adhering to aspects that are deeply rooted in one's own culture, including religious beliefs and family values, may be protective for the adolescent, this does not suggest that adapting to some American traditions and behaviors, including learning the English and appropriate social behavior, are not protective as well. Educational support-Parents, teachers, tutors, and the school encouraged the educational success of adolescent students by valuing education, talking about school days, offering homework assistance and support, problem solving, and mentoring. A 17-year-old Liberian girl described support from her teacher, she explained, "Teachers in the U.S. are good, they care about you a lot. They keep an eye out for me and my sister and try to make sure that we are getting the extra help we need. They push us to learn. They make sure we understand." Faith and religious involvement-Families with strong religious beliefs could use their faith to support adolescents in maintaining hope, making wise decisions, and developing effective coping skills. Attending church services and worshiping allowed families to feel spiritually connected, guided, and better able to overcome daily stressors and challenging transitions. A Liberian mother explained, "I believe God brought me here for a purpose, he brought my kids here for a purpose. He wants a better future for them, that's why he brought us here. I just decided to be strong enough and see a better future for my children." --- Protective Mechanisms Nine protective mechanisms were identified. They are defined below and are grouped into three categories: 1) Relational; 2) Informational, and; 3) Developmental. The following protective mechanisms were processes fostering adolescent refugees' competencies and behaviors that can promote their psychosocial well-being. --- Relational Supporting-Providing adolescent refugees with material, social, emotional, and advocacy needs. A Burundian father explained the support provided by the resettlement agency staff, "The case manager helped to get the social security cards, IDs, employment authorization documents. We told him about a problem with food stamps and he told us to talk to the food stamps director. He helps with social needs and he helped to register the kids in school." A Burundian mother explained that their neighbor, a Somali woman, "helped us to get clothes and food and taught us how to take the bus." Connecting-Helping adolescent refugees become associated or joined with individuals, groups and organizations either in the resettlement or mainstream community. A 17-year-old Liberian student explained that he, "has stayed involved with the Student Leadership Program, which he attended the summer before he went to college. Sometimes a supervisor asks for recommendations for other students that he or past participants think would benefit from the program. One time a teen encouraged a friend to apply..." Belonging-Providing close and loving relationships to adolescent refugees through acceptance, closeness, and attachment. A 14-year-old Liberian teen described quality time with his family, "We usually play games together, like board games and checkers. Sometimes on special occasions, like birthdays, or Father's or Mother's Day, we go out to eat and go to church. We spend a lot of time with our grandparents and other Liberians." A Burundian adolescent described his closeness with three boys from his home country, he explained that they, "help one another, we correct each other when we are doing bad things. We keep each other's secrets and hang out all of the time. Because of our friendship, our parents are best friends." --- Informational Informing-Imparting information to adolescent refugees regarding expectations, challenges, problems, opportunities, and help. A middle school teacher explained how she informs her students of expectations in the U.S., she stated, " well, you know what, in this society, in this social structure, you need, you may want to consider, or you know, directly instruct that this, you need to do these things." A pastor explained that he informs adolescents about laws and expectations because, "When people come and see such a freedom, they think, 'Wow, I can just do anything.' They don't realize that this country has laws and laws upon laws. For us who are pastors, when we come over is that we try to help them get over the cultural shocks. They need to know they are in a different society." --- Preparing Getting adolescent refugees ready for future events such as for school transition, moving, employment, college applications, and family caregiving. A project coordinator from a school explained that she teaches the refugee adolescents social norms that they will be expected to know when they seek employment and also helps them to plan college visits. A Reverend explained that a host family requested a mentor to help a female adolescent succeed, "a Liberian young lady who is a psychology major at an outof-state university and a good Christian was asked by us to mentor her over the summer." --- Developmental Defending Keeping adolescent refugees from harm or danger in resettlement, especially concerning alcohol and drug use, gangs and criminality, and HIV/AIDS. A Liberian mother explained how she sent her teen to relatives to remove them from neighborhood trouble, "I gave my two children to them to help me straighten the children up, and to take care of them also." A teacher explained the gangs and how he tries to help adolescents avoid negative influences: "new arrivals are not at risk for getting involved in gangs. These kids do not leave their homes at night but as time goes on and they begin to adjust to the culture they become more vulnerable." He walk students home and talks with them about staying away from gangs. Promoting-Advancing adolescent refugees maturation and development through being given new positions, responsibilities, or leadership. A 15-year-old Liberian youth explained his leadership position in school, "One of the positions they wanted me to take after they saw me in student council was sergeant of arms. Someone who will keep time, and keep the talking to a minimum when other people are talking. I signed up for that. They wanted me to give another speech again, so I had to write another one, and my worst teacher was the sponsor, and she told me that, "you can go out and do it because I am in there, and I will help you out if you need any help, so just talk to me," so I said, "ok," and I joined the African Club." Adapting-Facilitating adolescent refugees' flexibility, responsiveness, and change regarding learning, working, interpersonal relations, identity, and future plans. A 17-yearold Liberian boy described how valuable his teacher was: "He was, like, really dedicated to teaching and ready to inspire any student who was ready to learn, and he was also helping every student to be the best that they can be. Like, he will stay in there and help you." A teacher explained her approach to helping adolescent refugees in the classroom adjust to new circumstances: "I wondered how much of that type of behavior might be survival strategies learned previously, behaviors and mentalities that might have been necessary to survive in conditions of scarcity and uncertainty. Part of addressing their behavior is to create an environment that is safe and predictable, and where every student is assured of having everything they need in the classroom. Not needing to fight to get what they want." --- Integrating Protective Agents, Protective Resources, and Protective Mechanisms Figure 1 summarizes the overall grounded theory model derived from investigation of the study sample characterizes how adolescent refugees' psychosocial well-being is shaped by identifiable family and community protective agents and protective resources that are linked to protective mechanisms. Protective Agents are the individuals, groups, organizations, and systems that can contribute either directly or indirectly to promoting adolescent refugees' psychosocial well-being. The nine protective agents identified were grouped into three categories: 1) Youth ; 2) Family ; 3) Service Providers . Protective Resources are the family and community capacities that can promote psychosocial well-being in adolescent refugees. The eight protective resources identified were: 1) finances for necessities; 2) English proficiency; 3) social support networks; 4) engaged parenting; 5) family cohesion; 6) cultural adherence and guidance; 7) educational support; and, 8) faith and religious involvement. Protective Mechanisms are the processes fostering adolescent refugees' competencies and behaviors that can promote their psychosocial well-being. Nine protective mechanisms identified were identified and grouped into three categories: 1) Relational ; 2) Informational , and; 3) Developmental . Table 2 summarizes the consensus patterns observed among protective agents and mechanisms in the study sample. Protective agents within the community domain were divided into sub-domains by location: school, church, agency, and health care. Overall, relational protective mechanisms were cited as being the most prevalent among the participants, followed by informational and developmental. The protective agents most actively involved with protective mechanisms were, from most to least active: school and agency ; followed by family and church ; youth ; health and mental health . Explanations for these different patterns of involvement are described below. --- Discussion This study looked at Burundian and Liberian refugee adolescents in U.S. resettlement. Despite abundant challenges and hardships, we found evidence of resilience in their families and communities. Using longitudinal ethnographic data and grounded theory analysis, and based on family-ecodevelopmental theory, we built a model that describes resilience in terms of the multilevel protective agents, resources, and mechanisms that appeared to promote adolescent refugees' psychosocial well-being. The study found that family members, teachers and peers, church members and volunteers were protective agents towards adolescents' psychosocial well-being. However, regarding these protective agents, we found that as a whole these individuals, groups, and organizations were often overwhelmed, under-supported, lacked adequate knowledge and skills, and were disconnected from other protective agents. Despite some highly significant contributions, these deficiencies limited their capacities to promote the psychosocial wellbeing of refugee adolescents. These deficiencies should not be understood as predominantly individual deficiencies, but more as reflective of systems problems. In particular, we found that health and mental health providers were under-performing relative to their potential contribution. This reflects both challenges accessing appropriate and affordable health and mental health services focused on adolescent refugees, but also the reluctance of adolescent refugees and their families to seek health or mental health services as a way of addressing their needs. Refugee adolescents and family members prefer turning to family and prayer for support rather than mental health services, which are often highly stigmatized. We also found that in the study sample, because the friends and peers were all refugees or immigrants who were generally no better integrated into mainstream society than the adolescent refugee themselves, they were not able to contribute much to their psychosocial well-being as protective agents. Through examining the constructs of protective resources, we were able to better specify the capacities which protective agents need in order to be able to better promote refugee adolescents' psychosocial well-being. Most importantly, their needs are multi-faceted in that they included both material, linguistic, social, familial, cultural, educational, and religious capacities. Some, such as faith, were more under the control of the refugees to maintain, whereas others, such as finances for necessities, were highly difficult for the refugees to negotiate without support from organizations and government. Regarding protective mechanisms, we found that the nine identified were not distributed evenly, with developmental and informational mechanisms under-emphasized in comparison with relational mechanisms. Adolescent refugees were not being provided with adequate and accurate information and they were not being helped to anticipate, prepare for, and adjust to the many new challenges that they face. This does not mean that support is not critically important, but that informational and developmental assistance were harder to come by and call for more deliberate strategies to bring about. --- Recommendations Given the broad spectrum of family and community processes involved in fostering resilience, what can policymakers and service providers do to enhance the psychosocial well-being of refugee adolescents in U.S. resettlement? The results suggest that enhancing adolescent refugees' psychosocial well-being could be approached through strengthening the protective agents, resources, and especially mechanisms that were identified in this study sample. To start, it is important to recognize that a mental health or health clinician in practice seeing individual patients may not be able to do much to alter these family and community processes. However, clinicians who are engaged in service provision or consultation with schools, resettlement agencies, and faith-based organizations are better positioned. Furthermore, service providers do not only include clinicians, but refers also to a broad spectrum of workers who are positioned to help adolescent refugees in various contexts, including: school teachers, staff, and coaches; church staff and congregants; resettlement agency caseworkers and activity leaders; and volunteers. This broad spectrum of service providers can all function as protective agents for adolescent refugees. Because the current limitations of protective agents reflect systems problems, it follows that protective agents could be strengthened through a systems approach that involves: 1. Building or strengthening the relationships between groups and organizations that can help refugee adolescents ; 2. Convening dialogues about disruptions and solutions for refugee adolescents ; 3. Linking informal networks with institutions involved with refugee adolescents ; --- 4. Creating feedback loops regarding information pertinent to refugee adolescents ; --- 5. Enhancing the abilities of groups and organizations to be flexible in addressing the needs of refugee adolescents . It is especially important to prioritize strengthening health providers and families, who in this study sample, appeared to be under-performing relative to their potential. Targeted trainings of health providers, and family support and education interventions for refugee families could be helpful. Policymakers and service providers could focus new initiatives on sustaining those protective resources that were strong , strengthening those protective resources that were weak and initiating those protective resources that were the weakest . It is especially important for policymakers and service providers to focus on the protective mechanisms being promoted because these are the actual processes by which positive changes occur. The findings indicated that there were deficiencies in informational and developmental protective mechanisms that call for targeted interventions. Moreover, the model indicates that because resilience works through protective mechanisms, greater attention should be paid to understanding how to enhance them through new programs and practices. Of course government cannot solve these problems alone. However, government can provide resources, support, and guidance to increase the capacities of service providers in places where adolescent refugees can be helped. These in turn can work directly with youth and parents. In particular, the three highest priorities that emerged from this study were: 1) parenting educating and support interventions; 2) promoting greater parental involvement in education; 3) adequate jobs for refugee families. Programs and providers could use the results of this study to reflect upon their current work with adolescent refugees and to develop new strategies. They might begin by asking: 1) Do current strategies and actions help to strengthen the protective agents, resources, and mechanisms that were identified in this study? If so how? 2) What other protective agents, resources, and mechanisms are instead being targeted? 3) What are the obstacles to enhancing protective resources and mechanisms and how can these be ameliorated or overcome? --- Study Limitations and Further Research This study had several limitations. One, given differences in language and culture, misunderstandings were possible. This challenge was addressed through a multi-lingual research team and ongoing review of translation and cultural issues. Two, the sample was not representative of all Liberian and Burundian refugees, let alone other refugee groups in the U.S., or U.S. adolescents. Thus, the study findings might not be generalizable to these other groups. Three, two years was a reasonably generous period over which to study change, however, not long enough to document processes, let alone outcomes. There is a need for further research that builds knowledge on protective agents, resources, and mechanisms. One priority is to conduct intervention research in communities to examine: which acts of enhancing protective agents, resources, and mechanisms work with whom, under what circumstances, and why? A second priority is bring the insights from this study into a study using survey and quantitative methods that would test models of relationships between protective agents, resources, mechanisms, and a range of outcomes. Both of these research activities would also have to involve constructing assessment measures for protective resources and mechanisms, as well as other key constructs. A third priority is to look at a more granular level at how protective mechanisms work in the life cycle and daily lives of adolescent refugees such as through experience sampling method . Integrating Protective Agents, Protective Resources, and Protective Mechanisms

Background-Adolescent refugees face many challenges but also have the potential to become resilient. The purpose of this study was to identify and characterize the protective agents, resources, and mechanisms that promote their psychosocial well-being. Methods-Participants included a purposively sampled group of 73 Burundian and Liberian refugee adolescents and their families who had recently resettled in Boston and Chicago. The adolescents, families, and their service providers participated in a two-year longitudinal study using ethnographic methods and grounded theory analysis with Atlas/ti software. A grounded theory model was developed which describes those persons or entities who act to protect adolescents (Protective Agents), their capacities for doing so (Protective Resources), and how they do it (Protective Mechanisms). Protective agents are the individuals, groups, organizations, and systems that can contribute either directly or indirectly to promoting adolescent refugees' psychosocial well-being. Protective resources are the family and community capacities that can promote psychosocial well-being in adolescent refugees. Protective mechanisms are the processes fostering adolescent refugees' competencies and behaviors that can promote their psychosocial well-being.

The Gelberg-Andersen Behavioral Model for Vulnerable Populations has been used in research with homeless people to explain and predict their health services utilization and other health related behaviors . The vulnerable-populations model is a version of the Andersen Behavioral Model which originally described relationships among predisposing, enabling, and need factors that are associated with HSU and access to medical care in the general population . Predisposing characteristics exist prior to the perception of illness , enabling resources facilitate or impede HSU , and need variables pertain to physical illness. The expanded vulnerable-populations model incorporates traditional Andersen model factors and also includes specific vulnerabilities common among homeless people such as substance abuse, mental illness, and competing needs that are additional barriers to obtaining health care. U.S. rates of hepatitis B virus and hepatitis C virus infections are high among homeless people. In national household surveys, estimated rates for persons ever infected with HBV are about 4.7% and with HCV are about 1.6% . In contrast, prevalence rates among homeless people have been reported as high as 17% to 31% for HBV , and 17% to 53% for HCV . HBV and HCV can be transmitted through exchange of infected blood during injection drug use , which is prevalent among homeless people . HBV and HCV are also common among individuals with incarceration histories due to IDU and unprotected sexual activity within and outside of prison . The expanded Gelberg-Andersen behavioral model is the framework we use to predict forms of HSU among homeless adults as well as to predict HBV or HCV seropositivity and lack of knowledge of one's own positive hepatitis infection status . The behavioral model has not been used previously to assess associations between the predisposing and enabling variables and hepatitis. This study has three main objectives: to report the prevalence of hepatitis B and C in a representative sample of homeless adults and to assess whether those who are positive know their status, i.e., we assess a perceived need and an evaluated need and the association between them]; to provide an explanatory model of hepatitis infection as a function of predisposing and enabling variables; and to provide an explanatory model of health services use by homeless adults as a function of the Gelberg-Andersen model domains of predisposing, enabling and need variables. This analysis will contribute to an understanding of health care disparities within a vulnerable population. See Figure 1 for a schematic representation of the expanded Vulnerable-Populations Model for this study, which includes specific hepatitisrelated variables. --- Predisposing variables As depicted in Figure 1, predisposing factors include traditional sociodemographics and specialized variables that reflect vulnerability which are associated with HSU in both homeless and general populations . Demographics are also associated with hepatitis risk. For instance, Stein and Nyamathi found that older homeless men were more likely to be HCV positive. The vulnerable predisposing variables from the expanded Gelberg-Andersen model are relevant to homeless people. Severity of homeless history has been shown to impact numerous health problems including HCV and maladaptive HSU . We include childhood abuse. Homeless people often report high rates of childhood abuse, which in turn has negative effects on later health . Homeless people are also more likely to have prison or jail histories than those in the general population . Due to the association between incarceration and hepatitis, we expect an incarceration history to predict hepatitis and also impact other variables in the model . We further expect alcohol and drug use, especially IDU, to have a direct impact on HSU patterns as well as hepatitis prevalence. Risky sexual behaviors are associated with HBV, but the possibility of sexual transmission of HCV is controversial . We expect risky sexual behaviors to impact hepatitis infection. Homeless adults with mental disorders are more likely to report barriers to needed medical care than other homeless adults . Thus, we include a history of mental illness among the vulnerable predisposing variables. --- Enabling variables Traditional enabling variables include health insurance, a regular source of health care, and income, which we expect to predict more use of ambulatory services, the most desirable form of care in terms of patient outcomes and economic considerations. Homeless-specific variables include perceived barriers to obtaining health care, competing needs, and having case management, and they are also expected to impact HSU. Barriers to care and competing needs, such as struggles for shelter and food, lead to delays in seeking care and may precipitate reliance on acute care . Lewis, Gelberg, and Andersen found that unmet needs for medical care were associated with barriers such as long office waiting times and not knowing where to go for care. Lewis et al. found that the odds of unmet needs were lower among individuals with a regular source of care, although health insurance was not related to unmet or competing needs. We further hypothesize that using case management will facilitate better HSU. Sadowski, Kee, VanderWeele and Buchanan reported that homeless individuals offered housing and case management services had fewer hospitalizations, hospital days and emergency department visits compared with a usual-care group that was not offered housing and case management. Horvitz-Lennon et al. reported ethnic disparities in case management services. Our model also will identify any ethnicity-related disparities in the use of case management services. Finally, we expect that hepatitis-related variables will be more likely among adults with access to a regular source of care and health insurance. --- Need variables Traditional need variables include perceived poor health and specific health conditions in the current study. We also include HBV or HCV infection separately as a need variable. The vulnerable need variable is represented by not knowing one's seropositivity status, which endangers health by delaying vital medical treatment and monitoring and by increasing transmission risk. The Vulnerable-Populations Model assumes that the direction of relationships between HSU and the predisposing and enabling variables will be the same for evaluated and perceived need . For instance, more education and having a regular source of care may lead not only to better utilization but also to less hepatitis and a greater likelihood of knowing one's infected status , although we would expect perceived status would be more closely related to evaluated status for persons with higher education and a regular source of care. --- Outcome variables Some forms of HSU are more desirable than others, such as obtaining preventive care in an outpatient setting so that health problems can be dealt with earlier and more effectively . Less desirable HSU includes reliance on ERs or hospitalizations for problems that could have been addressed more easily and earlier by timely outpatient office visits. The model assesses the relative impact of the traditional and vulnerability-specific variables on these three types of HSU: more desirable outpatient office visits and less desirable ER and hospitalizations. Based on prior findings, we expect the need variables of medical conditions and perceived poor health to drive HSU because homeless people have less opportunities for preventive health visits in the absence of illness. In addition, we expect that the traditional-and vulnerable-model enabling variables will be relevant for HSU as well. --- Method Participants Participants were recruited for a community-based probability survey of homeless adults living in the downtown "Skid Row" area of Los Angeles in [2003][2004]. The final sampling frame consisted of 41 sites: 19 shelter programs at 10 locations and 22 meal programs at 9 locations. Homeless participants were identified by having spent the previous night either in a public or private shelter, or on the "streets" . Additional eligibility criteria included a minimum level of cognitive competence and English fluency. Only an estimated 5% of the sample were monolingual Spanish speakers. As needed, the 6-item Short Blessed Test was used to assess cognitive competence . The study employed a conventional two-stage representative sampling design to select homeless adults from shelters and meal programs. First, the sampling frame was stratified by site using site-days as sampling units. Second, selection of subjects employed either simple random sampling or systematic random sampling strategies tailored to each site. One site refused to participate, giving a site refusal rate of 1/41 or 2%. Of the 903 persons sampled from remaining sites, 588 were identified as eligible for interview. Of these, 41 refused to be enrolled in the study, 10 were later identified as repeaters, 1 could not be located, 1 had a language barrier, and 1 interview was discarded due to concern about client credibility. The combined screening and interview completion rate was 83%; the final study sample was 534 adults. The racial/ethnic makeup of the sample was 10% Hispanic/Latino, 16% White, 72% African-American, and 2% other ethnicities. Nearly 80% percent of the sample was male; the average age was 46 years. Interviews were conducted by the RAND Survey Research Group. Eligible adults were interviewed using a face-to-face computer-assisted questionnaire that included mental health and substance abuse modules from a shortened version of the Diagnostic Interview Schedule-IV . Blood serum was tested for HCV, HBV, and HIV. Data collection averaged 90 minutes per person for which participants were compensated $30. Participants were given appointments to return one week later for test results and referrals. Participants who returned were paid $10. If necessary, they could call for results but would not receive the incentive. After three months of fieldwork, payment was raised to $25 to increase the return rate. The overall rate of notification of test results was 92%. The study was approved by the Institutional Review Boards of UCLA, the Public Health Institute, and the RAND Corporation. --- Measures Single-item variables, composites, and multiple-indicator latent variables were indicators of the variables representing the domains of the expanded Gelberg-Andersen model. --- Predisposing variables Traditional predisposing variables-One dummy variable represented ethnicity . Hispanic ethnicity was not significantly associated with any outcome variable and was not included. A general minority ethnicity variable also was also tested but was not associated with the other model variables. Continuous variables of age and education were included. Gender was coded 1 = female, 2 = male. Vulnerable predisposing variables-Percent of adult life homeless was based on the percentage of their lives spent homeless after age 18. Childhood Abuse History was a latent variable indicated by 2 items: 1) whether they had been hit, beaten, burned, or hurt by someone they lived with before they turned 18; and 2) whether they had been made to have sex against their will before they turned 18. Incarceration history was dichotomous . Alcohol use was a latent variable based on 1) How many days they drank 5 or more drinks in the previous 30 days , 2) frequency of binge drinking in the previous 12 months , and 3) classification as alcohol dependent using DSM-IV criteria . Drug use was indicated by the mean of 3 parcels of 9 drug use items scaled from 0 , 1 , 2 , to 3 . The drugs included marijuana, heroin, street methadone, opiates, sedatives, crack cocaine, cocaine, hallucinogens, and speed. Other measured variables represented injection of illegal drugs , and classification as drug dependent based on DSM-IV diagnostic criteria . Mental illness history was represented by a dichotomous variable that indicated one or more overnight hospitalizations for mental health problems. Risky sexual behavior was a latent variable indicated by a history of trading: 1) sex for cash, and 2) sex for drugs. Other items such as regular condom use were not significant predictors and were not included. --- Enabling variables Traditional enabling variables-Regular source of care was assessed by: "Is there one place in particular, like a doctor's office or clinic, where you go when you are sick or need advice about your health?". Health insurance was assessed with a single item: "In the past 12 months, were you covered at any time by health insurance or medical coverage?" . Income was based on respondent report of total monies received over the past 30 days . This variable was not retained in the model because it had no significant impact on any other model variable. Vulnerable enabling variables-Barriers to care was a latent variable with 3 indicators that were each the sum of "yes" answers to 9 types of barriers to health care that the participant may have encountered in 3 situations : 1) when attempting to receive medical care ; 2) when attempting to receive care for mental problems ; and 3) when attempting to receive treatment for alcohol or drug use . Items include: "You didn't know where to go," and "It takes too long to travel to the treatment site." Competing needs was indicated by 4 items scaled from "never" to "always" that assessed how often in the past 30 days it was a problem 1) Getting a place for the night; 2) Getting food to eat; 3) Finding a place to wash up, and 4) Finding a place to go to the bathroom" . Case management was indicated by 2 yes/no items: 1) Whether any service programs were working with them to help them get on their feet , and 2) Whether they had a case manager, counselor or advocate who handled things for them. --- Need variables Traditional need variables-Perceived bad health was a latent variable indicated by 4 items from the SF-12 : 1) The degree to which pain interfered with a participant's normal activities in the previous 4 weeks [1 to 5 ]; 2) a subjective rating of personal general health [excellent to poor ]; 3) a mean of responses to 2 items that assessed whether during the past 4 weeks they had accomplished less than they would like as a result of their physical health and whether their physical health limited the kind of work or other regular activities they do; 4) the mean of 2 problems with physical function including a) moderate activities such as moving a table and b) climbing several flights of stairs; ["a lot" , "a little" , to "not at all" ]. Conditions was the sum of 10 possible serious health conditions that they reported ever having such as a heart attack, cancer, or tuberculosis. HBV or H CV seropositivity was a yes/no variable that was based on objective blood screening for history of presence of either virus . Blood samples were tested for the hepatitis B core antibody and hepatitis B surface antigen, with confirmation using the neutralization assay. HBV seropositivity was determined if the participant tested positive for either or both tests. Blood samples were tested for HCV antibodies using a second-generation enzyme-linked immunosorbent assay . Per Centers for Disease Control recommendations, only ELISA tests with a signal-to-cut-off ratio < 3.8 required confirmation. These tests were confirmed using a supplemental recombinant immunoblot assay . Vulnerable need variable-Does not know hepatitis-positive was based on whether participants reported being told that they had been infected with HBV or HCV. Participants who tested positive for HBV or HCV, but who reported no knowledge of their positive status, were given a 1 . Otherwise, they received a zero. Those who did not test positive for either HBV or HCV were also given a zero. --- Outcome utilization measures HSU items included the use of ERs, Hospitals, and Ambulatory office visits. ER use was based on 1 item: "In the past 12 months, how many times, if any, did you go to an emergency room for any health reason?" Hospital visits were assessed by: "In the past 12 months, how many times, if any, were you hospitalized overnight for a physical problem." This item excluded hospitalizations for mental problems or substance use problems. Ambulatory office visits were based on 1 item: "In the past 12 months, how many visits did you make to a doctor or nurse practitioner for a physical health problem. Do not count times when you only picked up medication or when you saw a doctor when you were hospitalized." --- Analysis The analyses were performed using the EQS structural equations program . Goodness-of-fit was assessed with the maximum-likelihood χ 2 statistic , the Comparative Fit Index , and the root mean squared error of approximation . The CFI ranges from 0 to 1 and reflects the improvement in fit of a hypothesized model over a model of independence among the measured variables. Values at .95 or greater are desirable, indicating that the hypothesized model reproduces 95% or more of the covariation in the data. The RMSEA is a measure of fit per degrees of freedom; values less than .06 are desirable. A confirmatory factor analysis tested the adequacy of the hypothesized measurement model and provided bivariate associations among the model variables. We allowed a correlation between IDU and hepatitis B or C positivity a priori because we expected that IDU, a measured indicator of the latent variable of Drug Use, would be associated with hepatitis. After minor modification to this model based on plausible suggestions from the LaGrange Multiplier test , we tested a saturated predictive path model that positioned all predisposing variables as predicting the enabling variables, all of the predisposing and enabling variables predicting the need variables, and all the predisposing, enabling, and need variables predicting the HSU variables. Relations that were significant among the predisposing variables were also retained in this model as within-domain correlations, and significant correlations among the residuals of the intermediate and outcome variables were also allowed. Nonsignificant paths and correlations were trimmed gradually. Indirect effects were also examined as evidence of possible mediation between the predictor and outcome variables. --- Results --- Confirmatory Factor Analysis After minimal model modification, the final CFA model was acceptable: ML χ 2 = 766.23, 495 df; CFI = .95, RMSEA = .03. All factor loadings were significant indicating that the model was plausible. Table 1 presents summary statistics including the factor loadings, means, and standard deviations of measured variables. Four correlated error residuals were added between some of the drug items and some of the alcohol items . As expected, IDU was correlated with hepatitis. The table of correlations among the variables in the model is large and is not presented in the interest of saving space. Focusing on bivariate relationships with the hepatitis-related variables, testing positive for HBV or HCV was significantly associated with: predisposing variables of African American ethnicity, male gender, greater age, less education, greater percent of life homeless, a childhood abuse history, an incarceration history, drug and alcohol use, IDU, risky sexual behavior, the enabling variable of more competing needs, and the need variables of perceived poor health and medical conditions. Testing positive was significantly associated with fewer office visits. Not knowing one's positive status was significantly associated with predisposing variables of greater age, less education, risky sexual behavior, and drug use, and the enabling variables of a regular source of care , more competing needs, and less case management. Regarding HSU, ER use in the past year was significantly associated with not knowing one's infected status , female gender, childhood abuse, alcohol use, a mental illness history, a regular source of care, health insurance, perceived bad health, and more medical conditions. Hospitalizations were associated with non-African-American ethnicity, more education, female gender, percent of life homeless, and more medical conditions. Office visits in the previous year were significantly associated with a regular source of care, having a case manager, perceived bad health, and more medical conditions, and negatively with not knowing one's infection status. --- Path analysis The final trimmed path model had excellent fit statistics: ML χ 2 = 921.09, 658 df, CFI = .95, RMSEA = .03. Table 2 presents the significant paths in the predictive model. A correlation suggested by the LM test was added between male gender and less sexual abuse . Not all significant correlations in the CFA remained significant predictors in the path model. Regarding study outcomes, seropositivity was significantly predicted by the predisposing variables of greater age, alcohol use, IDU, and risky sexual behavior. Not knowing one's hepatitis-infected status was predicted by predisposing variables of greater age and risky sexual behavior and enabling variables of not having a regular source of care, and less case management. ER visits in the prior 12 months were significantly predicted by items in all domains including younger age, alcohol use, health insurance, perceived bad health, and more medical conditions. Less use of ERs was predicted by seropositivity. Hospitalizations in the prior 12 months were predicted by female gender, more medical conditions, and a greater percentage of life homeless; hospitalizations were less likely among African Americans. Office visits in the prior 12 months were predicted by a regular source of care, case management, more education, and perceived bad health. However, visits were less likely among those testing positive for Hepatitis B or C, who averaged less than 1 visit in the prior 12 months compared to nearly 2 visits by the other participants. Other significant predictive relations among the elements of the Vulnerable-Populations Model can be seen in Table 2, including predisposing predictors of enabling variables, and predisposing and enabling predictors of the need variables. Of note, age and gender were important predictors of enabling and need variables. --- Indirect effects Indirect effects of predisposing variables on need variables include the following. Bad health was indirectly impacted by a greater percentage of life homeless, more childhood abuse, drug use and alcohol use. There were indirect effects of age and female gender on more medical conditions. Greater age and drug use impacted seropositivity. Not knowing one's seropositivity was significantly impacted indirectly by male gender. These effects were mediated by the enabling variables that directly predicted the need variables as reported in Table 2. As far as HSU is concerned, some indirect effects of the predisposing and enabling variables are not meaningful or are spurious because the mediators are poor health or more medical conditions which in turn predicted greater use of services. For instance, poor health items were predicted by more problematic variables such as more childhood abuse, competing needs, and barriers to care. If the direct relationship is significant and opposite from the indirect effect, the indirect effect is not reported. Several significant indirect effects on HSU were observed. ER use was indirectly impacted by predisposing variables of greater age, less education, female gender, a greater percentage of life homeless, childhood abuse, alcohol use, and mental illness. It was also impacted by having a regular source of care, health insurance, and more barriers to care. Office visits were indirectly predicted by female gender. There were no significant indirect effects on Hospitalizations. --- Discussion We outlined three objectives in using the expanded Gelberg-Andersen behavioral model. First, we wanted to assess the prevalence of hepatitis B and C in a representative sample of homeless adults and assess whether those who tested positive knew they were infected. Forty-three per cent of the sample tested positive for HBV or HCV or both: 31% were HBV positive, 28% were HCV positive, and 16% were positive for both. Importantly 71% of those testing positive did not know they were positive. Our second goal was to test a model of HBV and HCV infection and awareness of infection, conceptualized as need variables as a function of predisposing and enabling variables. Infection was related most strongly to predisposing variables associated with hepatitis risk in all populations: alcohol use, IDU, age, and risky sexual behaviors. In addition to the predisposing variables of risky sexual behavior and age, not knowing one's infection status was negatively predicted by enabling variables of case management and having a regular source of care. The third goal was to provide an explanatory model of health services use by homeless adults as a function of the Gelberg-Andersen model domains of predisposing, enabling and need variables augmented with the hepatitis-specific variables. As a unique contribution of this study, we found that seropositive people were less likely to report ambulatory office visits or even ER use, which is less desirable but at least exposes an individual to the health care system at some level. People infected with HBV or HBC are in need of screening, evaluation, monitoring, continuing care, and hepatitis-related education. Further, more attention to seropositive people might reduce transmission risk since patients might change their behavior if they learn they are infected. In this population, even an ER visit may be "better than nothing" and present a screening opportunity for seropositivity. In addition, significant associations were noted in the path analysis between not knowing one's hepatitis infection status and not having a regular source of care or case management. Establishing supportive links between homeless persons and the health care system would help with appropriate management of those that are seropositive and may not know it. Traditional predisposing variables were generally as important in predicting HSU with homeless adults as they are in general population. However, among these homeless adults, minority ethnicity did not predict negative HSU outcomes. African-Americans were more likely to have a regular source of care and had fewer hospitalizations in the past 12 months than those of other ethnicities in this study. Similar results have been found among homeless women by Stein et al. , Austin et al. and Gelberg et al. . Stein et al. hypothesized that the homelessness of the African-American women may have been based more on loss of affordable housing, lack of family supports, and extreme poverty rather than associated with mental illness and substance use that are found more often among white homeless women. This situation may extrapolate to both homeless men and women. Income was not important in this model, perhaps because most homeless are poor; also, their income can vary widely from month to month, making its predictive utility in any one month minimal. Men, as compared to women, generally reported more substance use and higher rates of hepatitis but fewer enabling resources to get care including less case management and a lower likelihood of having a regular source of care or insurance. Women, as compared to men, reported poorer health, more health conditions, and more hospitalizations. However, despite their greater reported health needs, women did not report more office visits than the men. As in the general population, age was important for health-related outcomes. Although older homeless adults were more likely to have a regular source of care and insurance, they were less likely to have case management and more likely to report bad health and medical conditions. They were more likely to test positive for hepatitis and less likely to know that they were HBV or HCV positive. Despite greater medical needs and better access to care, older age did not predict more ambulatory office visits, and older adults were less likely to use an ER for their exigent needs. More outreach to older homeless adults appears warranted, including targeting them for hepatitis screening. More education predicted more office visits and better health. Among vulnerability-specific predisposing variables, there was a significant direct influence of childhood abuse on barriers to care and more medical conditions. Childhood abuse was also significantly associated with other predisposing variables of percent of life homeless, having a jail/prison history, drug and alcohol use, risky sexual behavior, and mental illness. These results reinforce and expand the findings of Wechsberg et al. who reported that childhood abuse is associated with later stress-filled lives and drug use among homeless people. Abuse also has an impact on health-related outcomes including illnesses and difficulties negotiating the health care system. A more severe homeless history predicted more competing needs, less likelihood of receiving case management, and more hospitalizations. Thus, people who have suffered proportionately more time homeless may be the least effective at connecting with case management and health-related services as they often resorted to hospitalizations, the least discretionary and most expensive form of care. Surprisingly, homelessness history severity did not directly predict hepatitis seropositivity or lack of knowledge of infected status. Stein and Nyamathi found a significant and positive relationship between severity of homelessness and HCV positivity. Although an incarceration history was not a significant predictor in the path model, it had significant bivariate associations with male gender, more alcohol and drug use, more mental illness, more risky sexual behavior, and more competing needs. Its association with the outcomes was probably attenuated by its associations with other more potent predictors of the outcomes. Historically, incarceration has been a risk factor for hepatitis . A regular source of care predicted poorer health, more medical conditions, more recent office visits, and a greater likelihood of knowing their HBV/HCV-infected status. It is likely that among homeless adults, medical conditions and poorer health may predict having a regular source of care as much as the reverse influence as specified in the model. Clearly, a regular source can provide the continuous and comprehensive care needed to identify and treat health problems. Having health insurance predicted more ER use and more medical conditions. Exigent care for homeless people may be provided more readily when they are insured, and they may be more likely to go to an ER when they have insurance. This finding is similar to one reported by Stein et al. among a younger sample of women. It was speculated that poor women with illnesses were more likely to qualify for public insurance which may also be the case in this broader sample of men and women with a wider range of ages. Also, participants may have perceived the ER as their regular source of care. Vulnerable enabling factors of barriers to care and competing needs played only an indirect part in this model because neither directly predicted HSU. These items predicted more Perceived Bad Health. Case management was facilitative. Those who had case management were more likely to know that they were positive for hepatitis and also to make ambulatory office visits. This variable reflects some connectedness to community social services. In this sample of homeless adults, need was driving HSU. Poor health and medical conditions were the strongest predictors of use of emergency services. Alcohol use also predicted more ER use. These chronic conditions would be treated better if homeless adults had greater use of non-emergency services and received preventive screenings, continuity of care, and alcohol treatment in more appropriate settings. Not surprisingly, hospitalizations were predicted by serious medical conditions, and office visits were predicted by perceived bad health. The more desirable ambulatory office visits were facilitated by enabling factors of a regular source of care and case management. --- Limitations This study tested the Vulnerable-Populations Model, which posits a particular directionality of influence in HSU . Alternative predictive associations are always possible when data are cross-sectional. Certainly, having a regular source of care does not "cause" more illness, although it can increase diagnoses of existing illnesses. Further, homeless adults with illnesses may be more likely to have medical care available to them. Further, prior utilization could have led to knowledge of seropositivity. However, although most of the sample with hepatitis probably became infected years ago, and only a minority knew their status. Another limitation is that our findings rely on selfreports , and self-reported data are subject to recall bias or other errors. However, prior research using chart reviews to compare with self-report indicated that homeless persons were surprisingly accurate in reporting their use of health services . We used percent of adult life homeless which may have underestimated total lifetime homelessness; we did not have access to data on length of homelessness during childhood. However, only 7% of the sample reported experiencing homelessness before the age of 18. Furthermore, although our sample from Skid Row in Los Angeles was carefully and probabilistically derived from many community sources, our findings may not generalize to other homeless adults in other geographical areas. Of Los Angeles County's estimated homeless population of nearly fifty thousand in 2009 , about 5000 are located in downtown Los Angeles , 2009). In the downtown population, about 46% are sheltered. This corresponds closely with the data from the current study assessed a few years earlier in which 44% were sheltered. In greater Los Angeles County about 33% of homeless people are sheltered. The relatively high sheltered proportion in Metro LA is largely due to the high number of homeless service providers operating emergency and transitional housing in Skid Row . --- Conclusion Based on our findings, outreach among homeless adults may need to include efforts that provide continuous and comprehensive medical care that addresses multiple and complex health problems. One obvious example of enhancing medical services is related to HBV and HCV detection, which would include screening homeless adults for hepatitis, especially those who are injection drug users. For those testing positive for HBV or HCV, follow-up visits and monitoring may help prevent transmission as well as help the infected adult avoid alcohol and drug use and other risk behaviors for progressive liver disease. For those testing negative, patient education may help prevent exposure by reducing high-risk behaviors, and eligible persons could be encouraged to be vaccinated against HBV . Timely treatment might capture various medical problems before they become so serious that hospitalization or over-reliance on burdened emergency rooms is required . Permanent supportive housing for homeless people combined with in-house medical programs would be helpful as well . Changing housing and health policies to prevent homelessness in the first place among populations at risk would be a pivotal change that would provide a benefit to the health of our society as a whole . Hypothesized Gelberg-Andersen Behavioral Model for Vulnerable Populations specialized for homeless populations including specific hepatitis-related variables.

Objective-Homeless people have disproportionately high rates of viral hepatitis. The Gelberg-Andersen Behavioral Model for Vulnerable Populations (predisposing, enabling, and need variables) was expanded to predict prevalence and awareness of hepatitis B (HBV) or hepatitis C (HCV) infection as well as health services utilization (HSU) among homeless adults using structural equation modeling. Design-A population-based sample of 534 homeless adults in Los Angeles' "Skid Row" was interviewed and tested for HBV and HCV. Main outcome measures-HBV/HCV seropositivity, awareness of seropositivity, and HSU in the previous 12 months. Results-Seropositivity (43%), usually unknown (72% of seropositives), was predicted by injection drug use, alcohol use, older age, and risky sexual behavior. No regular source of care, risky sexual behavior, less case management, and greater age predicted not knowing one's positive status. Health insurance, younger age, alcohol use, perceived bad health and more medical conditions predicted emergency room (ER) use; ER use was less likely among seropositives. Hospitalizations were predicted by more medical conditions and greater percentage of life homeless and were less frequent among African-Americans and males. Ambulatory visits were predicted by a regular source of care, case management, more education and perceived bad health; they were less likely among seropositives.The Gelberg-Andersen Behavioral Model provided a useful guide for predicting HBV/HCV positivity as well as HSU in homeless adults. Most hepatitis-positives did not know

Introduction It is common for people to dread some risks but not others: They tend to be very afraid of epidemic diseases, nuclear power plant failures, and plane accidents but are relatively unconcerned about some highly frequent and deadly events, such as traffic crashes, household accidents, and medical errors. One key distinction of dreadful risks seems to be their potential for catastrophic consequences [1], threatening to kill a large number of people within a short period of time [2]. In this paper, we propose that people may be particularly sensitive to risks that could wipe out the number of people that corresponds to the size of a typical social circle-family, friends, colleagues, and others that people interact with on a regular basis [3,4]. What is the size of a typical social circle? In hunter-gatherer societies-typical for most of human evolutionary history-people frequently lived in bands of 20 to 50 individuals [5,6]. Several adjacent bands often maintained regular contact for collaborative activities or ceremonial purposes, creating a local community of about 100 to 200 individuals [7]. In modern societies, the same groupings appear in personal and professional social networks [8]. For example, in a recent study of a Dutch population , participants maintained regular face-to-face contact with a median of 50 individuals . The number of individuals with whom people are in regular-but not necessarily face-to-facecontact is somewhat higher: Participants in a study in the United Kingdom sent Christmas cards to a median of 138 people [9]. Similarly, the average number of contacts that people have at a large social networking site is 130 [10]. Larger social circles are relatively rare, possibly because humans have a cognitive limit allowing them to maintain stable relationships with at most 150 individuals [8]. In sum, although there is a large variability in reported sizes of social circles, active contact seems to be rarely maintained with more than 100 to 200 people. Why would people be especially afraid of risks that could kill roughly the number of people that corresponds to the size of a typical social circle? One possibility is that the fear developed in our evolutionary history. Given the very low population densities throughout most of the Paleolithic-about 1 person per 10 square miles [11]-losing one's community might have implied being alone for days or longer before encountering another group. This would have been a serious threat to one's fitness, as being in a group reduces predation risk, helps with finding food and hunting, and increases survival chances when injured [12][13][14][15]. People who were more afraid of risks that could kill their whole group might have been more likely to engage in behaviors that helped the group to escape or reduce the impact of the dreaded risk , thus saving their group and with it increasing their own fitness. Therefore, although today population densities are many times higher than in the Paleolithic and the next human being is typically just minutes away, people may still have a tendency to be particularly sensitive to risks that could destroy a typical social circle. If it is true that people dread risks that threaten to kill roughly the number of people that corresponds to the size of a typical social circle, then risks that threaten a much larger number of people should not be dreaded more. To investigate this idea, we conducted a series of 9 experiments. In Experiments 1 to 4, we tested a very specific hypothesis: Risks that kill 100 people would be feared more than those that kill 10 people, but risks that kill 1000 people would not be feared more than the risks that kill 100 people. That is, we expected this pattern of results rather than a monotonic increase of fear with increase in group size. The rationale is that risks that kill 10 people are insufficient to wipe out everyone in a typical social circle and, therefore, will cause moderate fear, risks that kill 100 people suffice to eliminate most of a typical social circle and hence will produce more fear than risks that kills 10 people, and risks that kill 1000 people cannot cause much additional damage to one's social circle than those that kill 100 people and hence will not be feared more. In Experiments 5 and 6, we used a different methodology to directly test the hypothesis that dread peaks when a risk threatens to kill the number of people that corresponds to the size of a typical social circle. In Experiments 7 and 8, we ruled out an alternative explanation of the results. Finally, in Experiment 9, we show that this pattern of results is not the same as the ''psychophysical numbing'' phenomenon. --- Experiments 1-4 --- Method Ethics statement. All experiments in this paper were approved by the Ethics Committee at the University of Granada. Participants had to confirm that they have read and agreed to a written paper-based or electronic informed consent form before starting the study. Three participants in Experiment 2 and one in Experiment 8 reported to be 17 years old and they were excluded from the analysis. Participants. Experiment 1 was conducted in January 2009 with n = 83 undergraduates in the lab at the University of Granada in Spain. Experiment 2 was conducted in February 2009 with n = 44 participants , recruited via two Web sites that linked to psychological experiments . Two thirds [29] of the participants in Experiment 2 were from the United States while a third lived in other countries . Participants in Experiments 3 and 4 were recruited via the Web marketplace Amazon Mechanical Turk, which provides a convenient pool for conducting some types of behavioral research [16]; all participants reported being from the United States. Experiments 3 and 4 were conducted in February 2011. Materials and Procedure. In Experiments 1 to 4, participants read a brief description of a scenario involving risks that threaten to kill a specific number of people. In Experiment 1 and 2, the risk was described as ''an unknown, deadly disease affecting your town.'' In Experiment 3, the risk was described as ''an accident in a factory near your town, resulting in a release of poisonous fumes into the air.'' Finally, in Experiment 4, the risk was described as ''an earthquake that struck a part of your town.'' In each experiment, randomly chosen groups of participants were told that the risk was forecasted to kill 10, 100, or 1000 people in their town within a week. Participants were then asked to rate how much they would be afraid of that risk on an 11-point scale ranging from 0 to 10 . As a control, we further asked participants to estimate the size of the town they lived in at that time and to report their age and sex. --- Results In line with our hypothesis, in all four experiments there was a significant increase in fear of risks killing 10 vs. 100 people, but no significant difference in fear of risks killing 100 vs. 1000 people . For example, in Experiment 1 the mean difference in ratings of fear of a disease killing 10 and 100 people was 1.460.55, d = 0.64, whereas there was a difference of 20.0460.54, d = 0.03, in ratings of fear of diseases killing 100 and 1000 people. In Experiment 3, the mean difference in ratings of fear of factory accidents killing 10 and 100 people as a result of the release of poisonous fumes into the air was 1.661.11, d = 0.71, whereas there was a difference of 0.1361.02, d = 0.06, in ratings of fear of factory accidents killing 100 and 1000 people. Similar results were obtained in Experiments 2 and 4 . There were no significant interactions of the number of deaths and sex, age, or town size of the participants. There was a tendency for women to give higher ratings of fear than men, but it did not interact with the pattern of results reported above. --- Experiments 5 and 6 To test if the results in Experiments 1 through 4 are partially due to some methodological artifact , in Experiments 5 and 6 we employed a different method. In particular, we asked participants to estimate the number of deaths that would make them feel different levels of fear in the scenarios used in the previous experiments . In addition, in Experiment 6 we asked participants to estimate the number of individuals in their social circle to test if our assumption about the size of a typical social circle is correct. If dread is indeed related to the risk of a typical social circle being wiped out, then, on the group level, the average number of deaths needed to feel the maximum fear should correspond to the size of a typical social circle. In addition, individual differences in the size of social circles may affect the results. --- Method Participants. Participants in Experiments 5 and 6 were recruited via the Web marketplace Amazon Mechanical Turk. Experiment 5 was conducted in February 2011 and Experiment 6 in March 2011. All participants reported to be from the United States. Materials and Procedure. Participants estimated the number of deaths that would need to occur in order for them to feel five different levels of fear ranging from ''not at all afraid'' to ''very afraid.'' Random subgroups of participants were asked to imagine different risk scenarios, all already used in Experiments 1 to 4: disease, factory accident, and earthquake. In addition, participants in Experiment 6 were asked to estimate the size of their social circle using the summation method [18]. In this method, participants report the size of different groups within their social circle, defined as the number of ''people you encounter and interact with regularly, or are otherwise important to you even if you don't meet them very often''. The total number of individuals reported is taken as an estimate of the size of their social circle. The groups were: immediate family, other close family, distant relatives one is in contact with, other people one is in regular contact with, and other people who one feels are important to him/her. Participants wrote the number of individuals belonging to each group in an open text field. A random half of the participants were asked about their social circle before the questions about fear, and the other half after. Note that measuring the size of social circles, or more broadly, the size of personal social networks, is a difficult methodological problem. The way a social network is defined, the questions used to ask about its size, and the ability of participants to recall their social contacts might influence the results [19]. The summation method [18] is based on the assumption that people's memories about other people are organized according to the social structure of their community and that using some common elements of social structure as prompts will improve recall [20]. Although the estimates are likely to be noisy because of various sources of error , they concur with results of other methods and are reasonable proxies of the size of participants' social circles [18]. --- Results Figure 2 shows the range of participants' estimates for the number of deaths that would cause different levels of fear, as obtained in Experiment 6. The results of Experiment 5 were very similar. Median number of deaths required for fear levels 1 to 5 were in Experiment 5 : 1 , 5 , 15 , 50 , and 100 . Even though the range of estimates was large, for the majority of participants the number of deaths that made them maximally scared was not larger than 500. For comparison, Figure 2 shows the average social circle size of the participants. Median size of the social circle was 77 individuals. Most participants had social circles smaller than 200 individuals. These results correspond to the indicators we gathered from the previous studies 1 [8][9][10], suggesting that the number of active contacts rarely exceeds 100 to 200 individuals. In sum, our hypothesis that dread reaches peak for risks that can kill all individuals in a typical social circle has been confirmed on the group level. At the individual level, our results suggest that participants might have used the size of their own social circle to estimate the number of deaths needed to evoke the highest level of fear. The estimated number of deaths needed for the maximum fear was not significantly different from the number of individuals in participants' social circle . Participants whose social circles were smaller than the median required fewer deaths for the maximum level of fear than those whose social circles were equal to or above the median size: Mdn = 40 vs. 150 deaths, respectively . On the other hand, the correlation between the size of individual social circles and the number of deaths that elicit most fear was small and not statistically significant . Taken together, these results suggest that the size of own social circle affects one's estimate of the most dreaded death toll, but it is likely to be only one of several factors that affect the results at the individual level. We comment on this issue in the Discussion. Neither the estimates of number of deaths nor those of social circle sizes were affected by participants' sex or size of the town where they lived, risk scenario, or whether the question about fear was asked before or after the question about social circle . Although women and participants from smaller towns tended to require fewer deaths for maximum level of fear, this tendency was not reliably present. --- Experiments 7 and 8 An alternative explanation of the results obtained so far might have been that people do not perceive much difference between the numbers 100 and 1000. In Experiments 7 and 8, we tested this using both a context-free evaluation of numbers and a monetary loss scenario. --- Method Participants. In Experiment 7, participants were 138 students at the University of Granada in Spain . Experiment 8 was conducted with n = 254 participants recruited from Sozioland-a large online panel of German Internet users. Experiments 7 and 8 were conducted in April and May 2009, respectively. Materials and Procedure. As in Experiments 1 and 2, randomly selected groups of participants in Experiment 7 were asked to rate their fear of diseases killing 10, 100, or 1000 people. After approximately 15 min of answering unrelated questions [21] randomly selected groups of participants were asked to simply rate the size of the numbers 10, 100, or 1000 without any context, using a 21-point scale ranging from 0 to 21 . In this latter context-free number size task, participants were assigned to 10, 100, or 1000 condition independently of the number they received in the former, fear of disease task. In Experiment 8, one third of participants received the fear of disease task used in Experiment 7, another third received the context-free number size task and rated the numbers using an 11-point scale, and the last third rated how upset they would feel if they lost 10, 100, or 1000 euros on an 11point scale ranging from 0 to 10 . Within each of the thirds, randomly selected groups of participants were assigned to 10, 100, or 1000 condition. --- Results In accord with our hypothesis and with previous results, participants' fear of a disease killing 100 people was larger than the fear of the same disease killing 10 people but was similar to the fear of the disease killing 1000 people . These results emerged consistently both in Experiment 7 and in Experiment 8. The mean difference in ratings of fear of a disease killing 10 and 100 people was 1.260.32, d = 0.51, and 1.860.85, d = 0.73, in Experiments 7 and 8, respectively, whereas the mean difference in ratings of fear of a disease killing 100 and 1000 people was 20.360.32, d = 0.15, and 0.1560.72, d = 0.07, in Experiments 7 and 8, respectively. In contrast to the results about fear of a disease, both the evaluated size of context-free numbers and the distress with monetary losses increased from 100 to 1000. The mean difference in ratings for context-free numbers 100 and 1000 was 2.060.90, d = 0.48, in Experiment 7, and 1.660.64, d = 0.89, in Experiment 8. The mean difference in ratings for being upset after losing 100 or 1000 euros was 1.360.64, d = 0.70, in Experiment 8. Taken together, these results suggest that the patterns of fear ratings observed so far are not due to a general insensitivity to the difference between the numbers 100 and 1000, and the effect is specific for scenarios involving people's lives rather than monetary losses or plain numbers. --- Experiment 9 On the surface, the patterns of fear found in our experiments appear to be similar to the results of studies on ''psychophysical numbing'' [22][23][24][25][26]. These studies have found that people's willingness to contribute to charitable causes decreases with the increasing number of people suffering. This phenomenon is thought to be a consequence of diminishing sensitivity to changes in number of deaths as the overall number of deaths becomes larger-similar to psychophysical relations observed for other physical quantities such as loudness and brightness [22]. In the present paper, we propose that the mechanism behind feelings of dread is different from psychophysical numbing. In particular, while psychophysical numbing occurs in altruistic behavior toward unfortunate people from other communities , the patterns in feelings of dread that we investigate originate from a potential threat to one's own community. Because the extinction of one's own communitybut not that of others-may threaten one's own survival, the specific form of the relationship between fear and the size of a threatened group can be expected only for people's own communities and not for that of others. In other words, when members of people's own community are threatened, fear should increase for risks killing 10 people to those killing 100 people, and then stay approximately the same for those killing 1000 people. When other communities are threatened, fear should be on a relatively low level independently of the number of people at risk. In Experiment 9, we tested this hypothesis. --- Method Participants. Ninety members of the online panel maintained by the Max Planck Institute for Human Development in Berlin, Germany participated in the study in June 2010. All participants were undergraduate students of local universities; 42% were men, with an age range of 21-36 years and an average age of 27 years. Materials and Procedure. We used the disease scenario of Experiments 1, 2, 7, and 8. As in these experiments, the size of the affected group was manipulated between subjects. In addition, there was a within-subject manipulation of the community affected by the disease. Each participant received three scenarios involving one group size in three different communities: Germans in Germany , Egyptians in Egypt , and German tourists currently in Egypt. The order of communities was counterbalanced. After a brief introduction, the participants read the following text: ''Imagine that [Germany/Egypt/German tourists currently in Egypt] [is/are] affected by an unknown, deadly disease. Health authorities forecast that the disease will kill [10/100/1000] [people in Germany/people in Egypt/German tourists currently in Egypt] within the next week.'' For each community, participants rated how afraid they would be of the disease on a scale ranging from 0 to 10 . Finally, to check the success of the community manipulation, participants rated for each community the extent of empathy and moral responsibility they felt toward the people affected by the disease, as well as how similar they considered themselves to be to the people in the particular community, all on the same 11-point scale used for the fear assessments. --- Results Manipulation checks. The type of community significantly affected participants' ratings of perceived similarity: Our participants rated themselves as most similar to people living in Germany , followed by German tourists in Egypt , and finally by people living in Egypt . In a repeated measures analysis of variance, all three communities were significantly different from each other . The patterns for empathy and moral obligation were similar: Participants reported higher levels of empathy and moral obligation for people in Germany than for either German tourists or people living in Egypt . Because of a substantial level of variation of these ratings between participants, in the analyses that follow we include all three variables as controls. Fear ratings. Fear was significantly stronger when scenarios involved either people living in Germany or German tourists than when they involved people living in Egypt . As Figure 4 shows, when scenarios involved people in their own community , fear ratings followed the same trend we found in previous experiments: There was a significant increase in fear of risks killing 10 vs. 100 people , but no difference in fear of risks killing 100 vs. 1000 people . In contrast, for people in other communities there were no significant differences in fear for different numbers of affected individuals. These results support our argument that the dread pattern we observed is specific for people's own community and is different from psychophysical numbing, which occurs for more distant communities [22]. While death of 100 people in one's own community is dreaded more than death of 10 people, in distant communities death of 10, 100, and 1000 people provoked similar, and relatively low levels of fear. --- Discussion In nine experiments, we found support for the hypothesis that people dread risks that threaten to wipe out the number of people corresponding to the size of a typical social circle. This pattern of fear appears consistently for several types of risk, including deadly diseases, earthquakes, and factory accidents resulting in a release of poisonous fumes into the air-suggesting that the underlying mechanism is not a specific adaptation to any particular risk but a more general concern about the possibility of losing one's social circle. We hypothesized that this relationship of dread and size of typical social circle has an evolutionary origin: loosing one's group might have been deadly in ancient human history. As the extant literature [5][6][7][8][9][10][11] and our own data suggest that people tend to maintain active contact with no more than 100-150 people, threat to this number of people would be expected to evoke maximum dread. Our results are in accord with this hypothesis. It is important to acknowledge other mechanisms that could have contributed to this pattern of results. For instance, because people seem to be cognitively adapted to maintain social contact with not more than 100-150 people [8], this group size may come to mind most naturally when trying to imagine a large group. In a related vein, people may have difficulty grasping the meaning of groups of people larger than about 100 people, as they may rarely encounter so many people in their everyday life. However, modern life seems to offer plenty of opportunities to encounter large numbers of people-from busy streets and public transportation, to sport and public events, to news about number of people using different products or being affected by war or disease. Therefore, imagining groups larger than 100-150 people is probably not very difficult in the present times. Another possibility is that participants imagined their own social circle being wiped off, and therefore their expressed dread could have been a by-product of feelings of emotional attachment to its members [27,28], rather than a product of an evolutionary adaptation to survival threats related to being alone. If this was the case, we would expect a strong correlation between the size of own social circle and the number of deaths that elicit maximum fear. However, the results of Experiment 6 suggest that the size of own social circle is only weakly related to the estimates of most dreaded death tolls, and that a general knowledge of the size of a typical social circle might be more important. This is supported by the fact that in our experiments feelings of dread were not reliably related to the size of participants' town, even though participants in smaller towns technically have higher chance to lose their social circles than participants in larger towns affected by the same risk. The patterns of fear found in our experiments may appear to be similar to the phenomenon of psychophysical numbing [22][23][24][25][26]. However, while psychophysical numbing occurs in the context of other communities, the dread risk investigated in this study concerns threats that can plausibly affect people's own community. Indeed, in our last experiment, we found the characteristic pattern of fear ratings only for the participants' own community-in particular for compatriots currently living in the same location as the participant. As expected, fear did not change with different number of victims in an unrelated community. These results do not invalidate the phenomenon of psychophysical numbing, but suggest that it is different from the phenomenon we describe in this paper. Regardless of whether the origin of dread is phylogenetic or ontogenetic, our results highlight the importance of people's social environments for the way they interpret and react to risks. The role of social circles in the understanding of risks has already been emphasized by Hertwig, Pachur, and their colleagues [3,4] who showed that people use their social circles to make judgments about frequencies of health risks in the general population. In a similar vein, Olivola and Sagara [18] proposed that people's preferences for risky solutions in problems involving human fatalities are guided by the distributions of death tolls in their environment. Specifically, in environments where high death tolls are relatively rare, subjective distance between events causing intermediate number of deaths is expected to be smaller. Hence, people in these environments will be more likely to prefer risk-seeking to risk-averse solutions to potentially deathly threats. The authors' results confirm this hypothesis: U.S. and Japanese participants, who according to the statistics rarely experience catastrophes involving large death tolls, prefer risky solutions more than Indian and Indonesian participants who experience large death tolls more often. Our findings cannot be directly compared to those of Olivola and Sagara [18] because we focus on emotional reaction to risks while they examined the effect of environmental distribution of death tolls on the curvature of individual utility functions and consequently their risk-seeking preferences. Given that it has been documented that emotion of fear affects risky choice [30], it is possible that both, fear of loosing one's community and statistical distributions of death tolls, contributed to both the present results and those of Olivola and Sagara. Further experiments could try to investigate the relative contribution of different mechanisms to the patterns we identified. Our findings are in line with the studies conducted by Wang and his collaborators [31,32], who showed that framing effects in risky choices involving human groups occur only when problems are presented in the context of large but not small group sizes, ''suggesting a 'live or die together' small group rationality'' [33]. That smaller groups have a special status when it comes to estimating risks is also echoed by Garcia-Retamero and Galesic's [34,35] findings that medical risks are easier to understand and recall if they are presented on the basis of smaller, evolutionarily plausible groups of people. We can exclude several possible methodological confounds of our findings. First, the results are not a consequence of a particular sample structure: We have replicated the basic patterns in different countries, with different demographic groups, and using different methodologies . Second, we believe that the results cannot be attributed to some side effect of the between-subjects design. In fact, as Birnbaum [36] showed, a more typical finding in between-subjects designs is a lack of differentiation between different numbers, or even results that violate mathematical laws . Third, this pattern of results does not appear to be a consequence of a ceiling effect: In none of the experiments was the average fear rating of a risk that strikes 100 people larger than 8 on a scale of 0 to 10, leaving enough room for further increase. In fact, the annoyance caused by monetary losses, measured in Experiment 8, grew monotonically with the increase in monetary amounts even though it was close to the upper part of the scale from the start . In addition, the pattern of results appears to be specific to scenarios involving people's lives. Evaluations of context-free numbers and monetary losses show the expected monotonic increase with the size of number. The present results have implications for communicating risks to the general public. When the goal is to raise public awareness about a certain risk that could claim 100 or more lives, it would be beneficial to stress that the risk could kill this or larger number of people. More broadly, our results stress the importance of considering social environments when studying people's understanding of and reactions to risks. ---

What makes some risks dreadful? We propose that people are particularly sensitive to threats that could kill the number of people that is similar to the size of a typical human social circle. Although there is some variability in reported sizes of social circles, active contact rarely seems to be maintained with more than about 100 people. The loss of this immediate social group may have had survival consequences in the past and still causes great distress to people today. Therefore we hypothesize that risks that threaten a much larger number of people (e.g., 1000) will not be dreaded more than those that threaten to kill ''only'' the number of people typical for social circles. We found support for this hypothesis in 9 experiments using different risk scenarios, measurements of fear, and samples from different countries. Fear of risks killing 100 people was higher than fear of risks killing 10 people, but there was no difference in fear of risks killing 100 or 1000 people (Experiments 1-4, 7-9). Also in support of the hypothesis, the median number of deaths that would cause maximum level of fear was 100 (Experiments 5 and 6). These results are not a consequence of lack of differentiation between the numbers 100 and 1000 (Experiments 7 and 8), and are different from the phenomenon of ''psychophysical numbing'' that occurs in the context of altruistic behavior towards members of other communities rather than in the context of threat to one's own community (Experiment 9). We discuss several possible explanations of these findings. Our results stress the importance of considering social environments when studying people's understanding of and reactions to risks.

Background Indigenous Australians have a lower life expectancy compared to non-Indigenous Australians, mainly due to higher rates of preventable chronic diseases [1][2][3]. Chronic diseases, such as type 2 diabetes and cardiovascular disease, are the leading cause of death for Aboriginal and Torres Strait Islander people [1]. Inequalities in health are further exacerbated by the location in which people reside, as rural and remote communities do not have the same opportunities to access services, education and employment [4]. The location and accessibility to Indigenous communities can also prove to be challenging when undertaking research into the health inequalities with Indigenous Australians [5]. There is a limited amount of research conducted with socially disadvantaged groups [6,7], fundamentally attributed to the continuous struggle for researchers to access, engage with and retain participants [7]. Indigenous Australians are unique in this regard as they are considered to be socially disadvantaged, however, there are concerns of being over-researched without corresponding health improvements [8]. Reasons for poor participation rates in 'hard-to-reach' populations can include mistrust in research and fear of exploitation [7], lack of education [9] and generally low population numbers compared to non-Indigenous people [10]. Conducting physical activity research with Indigenous Australians is challenging, but not impossible and the potential benefits could be significant. A recent systematic review investigated the effects on health outcomes of physical activity programs for Indigenous people in Australia and New Zealand [11]. While there was some evidence to suggest these programs improved health outcomes, most of the 13 studies in the review reported issues with recruitment, retention and adherence to the programs. For instance, several studies had less than 60% of participants complete the program, or failing to have complete data sets [12][13][14][15][16]. A randomised controlled trial by Biddle et al. [17] had an attendance rate of more than 80% to a physical activity intervention which involved small sided games over a four week period. Biddle and colleagues collected written feedback from the participants about their reasons for attending the program in which they stated that they had fun and enjoyed doing the exercise program. In addition to a quantitative evaluation, one of the studies in the review, a 12-week fitness program for Aboriginal and Torres Strait Islander women [13], also had a qualitative component to identify perceived barriers and facilitators to the attendance of the program [18]. Identified barriers included personal health, access to transport and competing obligations. In regards to logistics and attendance, transportation has also been recognised by Davey et al. [19] as a potential barrier to participation in an Indigenous cardiopulmonary rehabilitation program. The 8-week rehabilitation program involved twice weekly training sessions which were delivered at a gym in a private physiotherapy practice. Free transport to the practice was organised for the participants who would otherwise have had difficulties attending. Additional factors that are known to affect attendance to programs in rural areas are the need to travel away out of the community for work, medical appointments and education, as potential participants would be away for a substantial duration of the week [14]. Understanding the barriers and enablers to participation in physical activity programs can assist researchers in interpreting contextual factors that influence the implementation of and attendance to a program [20]. Thus, the aim of this study was to explore the barriers and enablers to participate in a community-tailored physical activity intervention for Indigenous Australians in a rural and regional setting in Far North Queensland. --- Methods --- Settings and recruitment of participants The two intervention sites were located in a rural and regional setting in Far North Queensland, Australia. The communities are not identified in this paper to protect the participants' anonymity. To be included in the study participants had to be Aboriginal or Torres Strait Islanders aged 18-45 who have a chronic disease or a risk factor for chronic disease. Based on the definitions by the Australian Institute of Health and Welfare Guidelines [21], risk factors for chronic diseases include a BMI of or greater than 30, a waist circumference of or higher than 94 cm for men and 80 cm for women and less than 150 min of physical activity per week. Participants were excluded if they were pregnant or advised by a doctor not to engage in physical activity. The program commenced in January 2016 in the rural community and in July 2016 in the regional location. At the commencement of the study, 34 participants were enrolled. A flowchart of participant recruitment and attendance is provided in Fig. 1. --- Interview guide development Interviews were conducted at the post-program assessments. The interviews were loosely based on the Health Belief Model [22], as it is one of the most widely recognised conceptual frameworks for health behaviours [22][23][24]. The Health Belief Model was selected as it was theorised that individuals would be more likely to voluntarily engage in the physical activity program if their current behaviour is perceived as a threat to their health. An interview guide was created to assist in the delivery of semi-structured interviews and included open-ended questions, which were designed to prompt responses relating to experiences with physical activity, the program delivery and future directions for interventions. The interview guide included the following themes: 1. What was your experience with the program? 2. What could have been made better about the program? 3. What would you like to see happen in the future? 4. Is there anything else you would like to comment on? Due to the complexity of a non-Indigenous researcher working with an Indigenous population group, the interview guide had limited questions to allow a better exploration of previously unidentified themes that may arise. --- Procedure An 8-week physical activity program was implemented in two communities. The free program was delivered through one-hour face-to-face group sessions, four times per week for eight weeks. Each week there were two sessions with vigorous activities and two with moderate intensity activities, delivered by a qualified sport and exercise scientist . The vigorous intensity sessions included small-sided team sports , such as soccer and touch football and circuit training with adaptations made to account for individual fitness levels. The remaining two sessions were self-paced moderate intensity walking sessions. After the program, participants were invited to partake in one-on-one interviews, regardless of their attendance to exercise sessions. Interviews were mostly conducted at the participants' house by the principal investigator who already had established relationships with the participants. --- Data treatment and analysis All interviews were digitally recorded by the PI and transcribed professionally by an external transcription service. Transcription files were presented in Microsoft Word and later uploaded into NVivo 11 for analysis. Interview transcriptions were read by the PI while listening to the corresponding audio file to cross-validate the content and to ensure the context was captured accurately. The interviews were coded and examined independently by a second person for a thematic analysis. The thematic network analysis was aimed at identifying global, organisational and basic themes [25], which were also compared between the two communities. The analysis involved coding the raw material into specific topics or words, such as 'money' , 'time' and 'work' in order to reduce the data. The next step of the analysis involved clustering the codes to identify the organizing themes. The three themes were labelled as 'barriers' , enablers' , and 'suggestions for future programs'. Finally, the thematic networks were constructed and arranged for further exploration and interpretation. To compare the findings between the rural and regional communities, the codes within the themes were reviewed and highlighted in two distinct colours to examine the frequency of appearance. The global theme of the data was named as 'participation in physical activity' as it best described the overall context of the findings. --- Results --- --- Motivations to participate in the program The motivations to join the physical activity program were different between the rural and regional community. During the analysis, the regional community had a sub-theme develop in regards to motivations of healthrelated outcomes such as weight loss and fitness. For example, reasons to join included 'To lose weight, get fit and be more active' and 'Weight loss and to get more into physical activity' . For one participant, weight loss was a means of controlling the progression of her chronic disease: 'Well, with the current condition I'm under it is very important that I control my weight…I've got diabetes' . The rural community members more explicitly stated motivations to participate were due to chronic diseases. One participant stated 'I'm determined not to be a [chronic disease] statistic… I'm pretty determined to break the cycle' , and another explained 'she [another participant] wanted to do it … her chronic disease wasn't looking good so she had to start making changes' . --- Barriers to attendance Many of the men who enrolled in the program stated that their main reason for non-attendance to activities was work commitments. Two of the participants who worked in the rural community stated 'I do work away, all the time up the Cape' while another often did shift work: 'Yes, well, at that time I was doing all day shifts at work and the timing was sort of out' . Both men were highly motivated to participate and attended the follow-up data collection, despite not attending a single training session. The men acknowledged their disappointment that they did not get any benefit from the program. One of the men in the regional community, who was also a university student, found it hard to attend as his work required him to travel away regularly, 'I think the biggest barriers for me are work, study and travel' . There were logistical barriers identified in the interviews, which were unique to the communities. In the regional community, logistical problems were mainly the timing of the class , whereas the rural community members experienced problems with access to transport. The rural community also had issues with access to facilities. During the first four weeks of the program, no indoor facilities were available and classes were regularly cancelled due to rain. 'You've got to take into consideration that a lot of outdoor activities can't be run on a regular basis sometimes, due to -obviously we live in the tropics, the wet tropics. That's why it's called wet, as -it can set in for days, months sometimes' . Despite being able to access a secondary facility five kilometres away, there were barriers in regards to access to transport to attend the classes. 'But organising stuff, even when there is wet [weather], we don't have indoor facilities…You've got to go in town or you've got to access a facility -fitness facilities, and then they cost money. Every time you want to do something, it costs money' . There was a lack of public transport services and 'not everybody has access to a car' . There were other barriers to program implementation and attendance in the rural community, such as 'sorry business' , which is a term used by Indigenous Australians that refers to a time of mourning in the community, following the death of a person. The process of mourning can take place over extended periods, and the intensity of the mourning is reflective of the importance of the individual that has passed away, respectively the importance A lack of family and peer support were identified as a barrier to participating in physical activity. One of the participants who had a young family and a partner that worked away explained: 'I think the only thing, sometimes I feel depressed there is not enough support from my partner, which is something that I'm wanting more is his support' . A male participant explained his hardship of needing to take control of his health conditions and the effect it has on his family. 'I suppose the receptive of the rest of the family is probably your key stumbling block if there's one … see I've got a wife and four kids so it's like I've got to have a separate meal'. He also expressed his feelings toward the impact that health choices have on his children: 'because I think she's borderline [his daughter]at the age of 20, borderline diabetic, and we try to do that with her, butyeahsort of heartbreaking when they get emotionally upset because they're not allowed to go and enjoy what they want to enjoy' . The lack of peer support and family also transitioned into themes of shame and stigma around undertaking physical activity. There were previous negative experiences in regards to people wanting to exercise for their health. 'Now or prior to the last five years, exercise was deemed as people who wanted to look good. There was always this negative attitude. Because I think obesity has become a norm in mostin the cultures now. You either need to look fat, so everyone stays fat, and if you want to lose fat, you're deemed as being sick. If you want to look good for the sake of being healthy, people think that you're better than anybody else. Whereas I think the culture is shifting a little bit now, or the mentality is shifting' . --- Enablers to participation Several participants mentioned finances and money during the interviews as barriers to participate in physical activity and that having a free program acted as an enabler. Reasons for joining were because 'it is good that it doesn't cost money. Everything else costs money' . Unemployment was as a financial barrier to undertaking exercise as 'Money is a major barrier for myself. Only because you're still looking for work' and 'it was free … Yeah, I couldn't afford it, no money, no job' . An identified enabler to joining in the physical activity program was the inclusion of family members into the activities. Figure 2 provides a kinship map of the participants in the rural community. Before the program, only the Indigenous mentor was known to the principal investigator. Despite flyers being displayed in the community, recruitment Fig. 2 A kinship map of the participants from the rural community. The health worker was not local to the community, but worked with the Indigenous mentor tended to occur via snowballing methods through family members. In the regional community, five members from the same family participated in the program. One of the family members stated 'It just made it all the better that it was family involved… And yourself of course. You're family' . Aside from family participation, peer support and guidance also assisted with facilitating participation. One person admitted to feeling uncomfortable when exercising on her own due to embarrassment: 'It's just if I'm by myself in public and I'm going for a jog on the road, I feel very self-conscious … Look at this fat chick she's jogging, what the hell' . Another participant stated they needed support and guidance from a qualified person, 'I tend to do things alone, and I don't know if I'm doing it right or not. Whereas if you've got somebody there who can guide you, you sort of have a better understanding of the process … most of the time I go to the gym, and one of the fears I have is not knowing if I'm actually doing it right. You know that as you're probably aware, you can hurt yourself…so I tend to do a lot of jogging' . Themes of resilience also emerged in regards to the lack of peer support. It was encouraging to hear statements such as: 'You have, yeah, those kind of people that would think [exercising is a disconnecting activity] like that, but, you know, just got to ignore them and do what's best for you, I guess … Turn around and tell them where to go, say this is my life, I'm not going to end up with a chronic disease, are you going to look after me when I'm on dialysis or whatever?' and 'I think for me that's what I'd like to see, is that exercising becomes the norm for our people … my mum comes from the dormitory [stolen generation] [26], so exercising for her was never the norm. I hope I've broken the cycle now, my kids will start to see it as more normal' . An enabler to participation in the physical activity program was the participants' relationship with the researcher. Many primary concerns to program implementation in the rural community revolved around the principal investigator being an 'outsider' , female and non-Indigenous. When participants were asked about whether having a non-Indigenous female run the program could have been an issue, many positive responses were received. One of the male participants explained his thoughts on the principal investigator, focusing on the relationship and level of comfort: 'Of course, because you really need to have a good relationship and it eliminates the shame factor. Because sometimes you're a bit too embarrassed, even though if you can'tif you don't have the strength or the capacity to do what you want to do or achieve' . For one of the female participants, having a female principal investigator was seen as an enabler to participation: 'Not really, when I met you, I was like yeah this is going to be easy. If it was like someone in their 30s, 35 and not like a male, yeah I'm probably not going to rock up to your session if they weren't so welcoming like you are' . Although there was not a negative reaction to the principal investigator being a non-Indigenous female, the need for a male role model was evident, particularly in the rural community. The principal investigator felt comfortable and welcome in the community, however, there were suggestions for future research. For instance, one of the men responded by saying there was a need for male role models: 'To some extent yeah, a possibility. I mean it doesn't matter … especially in communities, having a male would benefit, because a lot of the times, when you use girls, the girls always turn up to a lot of your stuff. But the blokes won't because there's too many women there. But if there was a bloke there and he sort of did it together, maybe they might turn up' . In agreement, another male participant also offered his suggestion to increase attendance: 'Just that it could have been a lot more user friendly in regards to that. I mean, I'm not saying that you're an uncomfortable person to talk to. You're very comfortable. But that's just me. Other people might find it uncomfortable and it's like role modelling too. So, if you've got aneven not so much of a young male or an older male who's got some problems, health problems, they can actually role model or they've come from a background where there's bad health in the families and so forth' . The need for male role models was further emphasised by higher program attendance by women. One woman suggested 'I think maybe it needs to have a male and a female, because it seemed to have worked more so for the women, than for the men' whilst another believed 'Culture is still focused on the men's and women's stuff … we did have a men's group, but they do a lot of social outings. That's something we could've tapped into' . It would have been ideal to have a male involved in the program, as this was an issue discussed during the study design, it was extremely difficult to involve a male Indigenous role model. This was due to the lack of male Indigenous health workers and qualified people to run physical activities. --- Discussion This study sought to explore Aboriginal and Torres Strait Islander participants' perceived barriers and enablers to attending a community-tailored physical activity program implemented in a rural and a regional setting. There were unique themes identified within the communities in regards to the barriers to attendance, however, the motivations to participate were similar across both sites. Firstly, the rural community had a lack of provisional infrastructure and access to resources. This impacted the program delivery and attendance due to inadequate access to an undercover facility during wet weather events. --- Barriers to participation There was a contrast between the logistical barriers between the communities where participants in the regional community were unable to attend due to poor timing, as opposed to those in the rural setting for whom lack of access to transport was a major barrier. The rural participants also expressed the importance of having both female and male role models for community members. For many their main motivation to participate in the program was that they saw physical activity as a means for prevention and management of chronic disease. In the regional community motivations were more centred on physical activity-related outcomes of weight loss and increased fitness. Participants from both communities highlighted the importance of support from their family and peers with their journey to becoming more active. Evaluations of physical activity interventions with Indigenous populations have reported several barriers to participation, which include the transient nature of the population, logistic issues and shame [27]. The Indigenous Australian population is more mobile than the remainder of the Australian population, as they often need to leave their homes to access medical services or for cultural obligations such as to attend funerals [28]. Particularly in smaller communities Indigenous people often need to relocate to larger cities to find employment opportunities [14]. For the rural community in this study, the need to travel to a larger city for personal, or family medical reasons was a frequent reason for non-attendance. Logistical issues, mainly in regards to transportation, are also known barriers to attendance to research programs [18]. Davey et al. [19] recognised access to transportation as an issue and included free transportation to exercise sessions, which was used by the majority of the participants and consequently attributed to retaining participants. In the context of this study, access to free transport was offered at both sites, but car-pooling was more common in the rural community due to the lack of public transportation. In the regional community, the provision of free transport was difficult due to the geographical spread of participants and due to the principal investigator also needing to conduct the exercise sessions. There were some participants who did use public bus services to attend the program. The concept of shame refers to embarrassment in certain situations and is often due to attention or circumstance, rather than the action by oneself [29]. The most prominent example of shame occurring was during a walking group in the rural community, where the principal investigator was asked to walk on the opposite side of a participant to block her from the view of men who were sitting across the road. Thompson et al. [30] examined the cultural and social context of physical activity in an urban Aboriginal population in Melbourne. A finding was that the Aboriginal people tended to view physical activity in three different aspects of everyday activities, sports and exercise. Everyday activities are described as necessary and performed for the benefit of the family and community, such as mowing the lawn or playing with children. Sporting competitions which can include individual and team disciplines are highly valued as the player represents the community and is a source of community pride and esteem. However, exercise is viewed as something conducted individually and specifically for personal fitness and health. In this context, the desire to personally improve oneself through exercise can be looked upon as a shameful and disconnecting experience by the community as it is often undertaken separately from family and community. Similar to our findings, Thompson et al. [30] reported that there were some elements of shame around exercising to improve health. The involvement of family in the exercise sessions may have reduced the shame associated with participation and acted as an enabler to participation. --- Enablers to participation Enablers to participation in the physical activity program primarily revolved around relationships to the other participants and with the principal investigator. In the rural community, all participants had a kinship connection with at least one other member of the program. It became evident that recruitment to the project was largely through kinship connections, rather than through the wider advertisement throughout the community. In the regional community, kinship connections were also an obvious factor in recruitment. One family had five family members involved, while another family had three sisters participate. The flexibility in the program delivery was also well-suited to participants who were the primary carer for children or dependant adults. It was often observed that participants would bring their children to training sessions, and in one instance a participant brought her dependant adult mother to a class. During the walking groups, some women would bring prams or bikes for their children to join in with the activity. There were also several community members who were not part of the program, but who began to attend training with the encouragement of the existing participants. The fact that the principal investigator who ran the physical activity intervention in the two communities was a non-local, non-Indigenous female was seen as a potential barrier to conducting the research. However, positive feedback was received and participants often 'vouched' for the principal investigator during the recruitment of their family members. The only barrier with the principal investigator conducting classes was the lack of mentorship for the men who were involved in the program. This was highlighted during the interviews with the members of the rural community. There was an understanding that it is challenging to find Indigenous men to assist with the program, especially in a voluntary role as it would have been with this project. It is important to be aware of the cultural ideas of men's and women's business which refers to certain customs and practices in Indigenous communities that need to be undertaken by men and women separately [31]. Therefore, researchers should aim to have mentors from both genders. --- Strengths The 8-week physical activity intervention lead to significant improvements in health outcomes, however, attendance was low across both communities. The research project utilised a participatory action research approach [32] and was guided by local Indigenous mentors, in conjunction with a local Aboriginal health organisation. The support from Indigenous mentors was invaluable as they provided a more in-depth understanding of the broader constructs of the community and the cultural ideologies to perform culturally appropriate research. Another strength of the research was the relationship between the principal investigator and the participants. Although the attendance rate was low, the responsiveness of participants was high in terms of availability for follow up assessments and interviews. Originally, there was no qualitative evaluation component planned for the physical activity program. However, after low recruitment and attendance rates were observed during the first few weeks of the program, the qualitative component was added to create a better understanding of the factors that lead to the low participation rates. This allowed for a better interpretation of the study results, as the interviews offered a rich source of data and new information in regards to improving the implementation of physical activity programs in Far North Queensland. --- Limitations The theoretical model utilised for the qualitative evaluation is a somewhat limiting model for a complex topic. The Health Belief Model was useful to identify barriers and enablers to program participation, but there are suggestions for future research. A limitation of this model is the lack of focus on external factors, such as access to transport, as opposed to individual factors, which can influence health behaviours [33]. A suggestion for future research would be to include the use of an ecological approach [34], as it could have been useful in this context. --- Conclusion The aim of this qualitative study was to explore perceived barriers and enablers for a rural and regional Indigenous community to participate in an eight-week physical activity program. Overall, there were positive attitudes and high levels of motivation towards the program. Enablers to participation were the inclusion of family members, no financial cost and a good relationship with the principal investigator. Barriers to program attendance were mostly beyond the control of the individuals, such as cultural obligations, the transient nature of communities and lack of community infrastructure. In conclusion, more consideration is needed before program implementation to understand community-specific barriers and enablers to physical activity programs. Furthermore, more work is needed to better understand how to improve participation rates. --- --- --- --- --- Competing interests The authors declare that they have no competing interests. ---

Background: Aboriginal and Torres Strait Islander people have higher rates of chronic disease and a lower life expectancy than non-Indigenous Australians. In non-urban areas these health disparities are even larger. The aim of this qualitative study was to explore perceived barriers and enablers to attending an eight-week physical activity program in a rural and regional setting which aimed to improve health outcomes, but had a low attendance rate. Methods: Thirty-four Indigenous Australians participated in the intervention from the rural (n = 12) and the regional (n = 22) community. Qualitative semi-structured individual interviews were conducted at the follow-up health assessments with 12 participants. A thematic network analysis was undertaken to examine the barriers and enablers to participation in the program. Results: Overall, there were positive attitudes to, and high levels of motivation towards, the physical activity program. Enablers to participation were the inclusion of family members, no financial cost and a good relationship with the principal investigator, which was strengthened by the community-based participatory approach to the program design. Barriers to program attendance were mostly beyond the control of the individuals, such as 'sorry business', needing to travel away from the community and lack of community infrastructure. Conclusions: More consideration is needed prior to implementation of programs to understand how communityspecific barriers and enablers will affect attendance to the program. Trial registration: ACTRN12616000497404. Registered 18 April 2016.

INTRODUCTION There is growing concern that the 'lockdown' measures to control the spread of the COVID-19 pandemic 1 2 have had unintended consequences including an increase in mental ill health. Several studies since the COVID-19 pandemic began have reported high levels of depression and anxiety [3][4][5][6][7] in the UK. However, some of these surveys are either cross-sectional or longitudinal within the lockdown period 6 7 with only three studies 3-5 comparing mental health in the pre-COVID-19 period to mental health during the COVID-19 lockdown. These studies consistently found worsening mental health in younger adults and those who were financially insecure. Other associations were reported in those who had a disability 3 ; pre-existing mental and physical health conditions 4 ; were living alone 4 ; in parents of young children 5 ; and in women. 4 5 In these studies, most participants were of White European origin and the larger of the studies were in a relatively affluent population. 4 Mental ill health is often reported as being more prevalent in people from ethnic minorities and the socially and economically disadvantaged, 8 9 but no longitudinal research to date has investigated the impact of COVID-19 lockdown on mental health in these populations. 10 We were able to explore these questions in depth by building on the Born in Bradford research programme which includes two longitudinal birth cohorts of ethnically diverse families living in the city of Bradford, many in deprived circumstances. These cohorts have recent in-depth information on the demographics, socioeconomic status and mental health of mothers before the COVID-19 pandemic, [11][12][13] as well as mental health during the March-June 2020 COVID-19 lockdown, 14 and so offer a unique opportunity to assess the impact of COVID-19 and lockdown longitudinally in a deprived and ethnically diverse population. We used the data to: ► Describe the changes in the prevalence of depression and anxiety in mothers living in Bradford from before the COVID-19 pandemic to during the March-June 2020 COVID-19 lockdown. ► Identify variables associated with a clinically important increase in mental ill health in this population, to identify vulnerable groups that may need additional support in the recovery from the pandemic. ► Explore whether there is a difference in the magnitude of the associations of key exposures with an increase in clinically important mental ill health by ethnicity . ► To explore mothers' lived experience during the March-June 2020 COVID-19 lockdown by assessing the frequency of worries and concerns relating to mental ill health obtained through free text responses to open questions. --- METHODS --- Study design A longitudinal study using data collected at two time points, before and during the March-June 2020 COVID-19 lockdown, from mothers who participated in one of two prospective birth cohort studies in Bradford: Born in Bradford's Growing Up study, with parents of children currently aged 9-13, 11 12 and Born in Bradford's Better Start , with parents of children currently aged 0-5. 13 Patient and public involvement Born in Bradford is a 'people powered' research study; the local community was consulted to identify key research priorities during the March-June 2020 COVID-19 lockdown. This included consultation with key community groups, seldom-heard communities and local policy and decision makers to ensure that the focus of the research was relevant to local needs. The COVID-19 survey and recruitment approach were tested through our established research advisory groups. The findings of the study were also shared with these groups to enhance interpretation and ensure useful dissemination back to the community. 14 --- Consent Participants had previously consented for their research data, and routinely collected health and education data, to be used for research. For the COVID-19 survey, verbal consent was taken for questionnaires completed over the phone and implied consent was assumed for all questionnaires completed via post or online. --- Data collection Full details of the data collection of the March-June 2020 COVID-19 survey can be found elsewhere. 14 15 In summary: Participants were recruited using a combination of emails, text and phone with a follow-up postal survey, and in their main language wherever possible. --- Pre-COVID-19 data Pre-COVID-19 data for BiBGU participants were derived from two sources: participant ethnicity and age were collected during pregnancy 11 ; recent follow-up data on mental health . 12 Pre-COVID-19 data for BiBBS participants were taken from data collected during pregnancy . 13 The median time since most recent pre-COVID-19 data collection was 15 months for BiBGU and 29 months for BiBBS. --- March-June 2020 COVID-19 lockdown data collection Mental health was measured using the eight item Patient Health Questionnaire 16 for depression and the Generalised Anxiety Disorder Assessment for anxiety. 17 These are widely used measures of the severity of symptoms of depression and anxiety that have been Open access validated in the general population, 16 17 and in many ethnic minorities including UK residents of Pakistani heritage. 18 Information was also collected on household circumstances; family relationships; social support and loneliness; financial, employment, housing and food insecurity; and physical health. This contextual information was captured in self-reported questions administered in the COVID-19 lockdown survey; details of variables used are in table 1 and our protocol paper. 14 Information on the participants' lived experience during lockdown was captured using free text questions that asked 'What are your three biggest worries at the moment?' and 'Has anything been easier or more enjoyable in lockdown?' --- Data preparation To describe the changes in prevalence of depression and anxiety from before the COVID-19 pandemic to during the COVID-19 lockdown, we derived a categorical variable for each PHQ-8 and GAD-7 score based on the standard clinical scoring classification: 0-4 no depression, 5-9 mild depression, 10-24 moderate-severe depression; 0-4 no anxiety, 5-9 mild anxiety, 10-21 moderate-severe anxiety, 16 17 we collapsed moderate, moderately severe and severe categories to indicate clinically important symptoms of depression and anxiety. To identify variables associated with an increase of mental ill health we used an increase in PHQ-8 and GAD-7 scores of 5 or more points as an indicator of clinically important change in symptoms. This cut-off was chosen following guidance from previous research and consultation with clinical colleagues. 19 An increase of 5 points or more would also always result in a change in the categorisation of symptoms while also capturing changes in severity within the moderate/severe categories. Ethnicity was coded using Census 2011 categories and also as 'White British', 'Pakistani heritage' and 'Other', which included all other groups due to small numbers of a wide range of ethnicities. Given the heterogeneity of the 'Other' ethnic group, the decision was made to not use this group in these analyses, but to focus on comparing those of Pakistani heritage and White British ethnicity. A number of categories within other explanatory variables were collapsed for analysis. These included: quality of relationship with partner: average to poor ; loneliness: not lonely and lonely ; social support: easy to get support and not easy to get support ; food insecurity: secure and insecure ; housing security: secure and insecure . --- Open access Missing data on measures were small for most variables and were not adjusted for in the analyses. --- Data analysis To describe the changes in the prevalence of depression and anxiety from before to during the pandemic, we explored the changes in PHQ-8 and GAD-7 categories using descriptive statistics and presented these visually to elucidate the patterns of both positive and negative changes. We used descriptive statistics to present the results of sample characteristics, including depression and anxiety scores at pre-COVID-19 and COVID-19 lockdown survey time points. We then modelled the ORs associated with an increase in PHQ-8 and GAD-7 scores by 5 or more points using separate unadjusted logistic regression models for each exposure variable of interest. Exposure variables included in the analyses were identified as indicative of an increase in mental ill health in the lockdown survey findings reported previously. 15 Pre-COVID-19 PHQ-8 and GAD-7 scores were controlled for in each model. In order to explore whether or not the magnitude of the association between exposure variables and a clinically important increase in symptoms of depression and anxiety differed between ethnic groups, we repeated the above analyses for Pakistani heritage and White British participants separately. This approach avoids the difficulties inherent in interpreting the ethnicity coefficient in regression models controlling for other variables. 20 All statistical analyses were carried out using Stata V.15. 21 To explore mothers' lived experience during the March-June 2020 COVID-19 lockdown, the free text responses to open questions were reviewed and themes that related to mental health of the mothers were pulled out. Responses were coded using thematic analysis. 22 The first 100 responses were analysed by one researcher , employing an inductive approach where coding and theme development were driven by the content of the responses. Two codebooks were developed, one for the questions on the three biggest worries and recent challenges during lockdown and another smaller codebook for the question on what had been made more enjoyable and easier during lockdown. Using Microsoft Excel, the remaining responses were then coded by three different researchers in order to test the strength and validity of the codebooks. Through frequent discussion between the researchers about this process, adjustments were made to the original codebooks so that they were reflective of the BiB, Born in Bradford. --- Table 1 Continued Open access total responses. The emergent themes relating to mental health were used to illuminate the findings from the quantitative analyses. --- RESULTS --- Study population A total of 2144 of those invited participated in the COVID-19 survey between 10 April and 30 June 2020. Full details of the study population are described elsewhere and demonstrate that the population are broadly representative of the BiBGU and BiBBS cohorts. 15 Of these 2144 mothers, 1860 had complete surveys and linked data from pre-COVID-19 surveys and were used for this study. Of these participants, 1316 were in the BiBGU cohort and 544 were in the BiBBS cohort and had a mean age of 37.5 years . Participants were ethnically diverse: 877 were of Pakistani heritage, 613 White British and 320 other ethnic groups; 1211 lived in the most deprived quintile of material deprivation in England. Table 1 describes the sample characteristics of the study population. Respondents were representative of the invited cohort on baseline levels of depression and anxiety. Ethnicity was skewed with more White British and less Pakistani heritage respondents than in the invited sample, although the study sample still offers a large ethnically diverse population . --- Study sample change in prevalence of depression and anxiety The prevalence of moderate/severe depression increased between the pre-COVID-19 and COVID-19 lockdown surveys from 11% to 20% . The proportion of mothers reporting mild symptoms of depression remained similar while the proportion of those with no depressive symptoms decreased from 65% to 56% . The proportion of mothers with moderate/severe anxiety increased from 10% to 16% . The prevalence of mild anxiety also increased from 16% to 23% , while the proportion of participants with no anxiety fell from 75% to 61% . Within-mothers change in depression and anxiety A total of 1760 participants had both pre-COVID-19 and COVID-19 lockdown depression data and 1634 had pre-COVID-19 and COVID-19 lockdown anxiety data. Table 3 shows the change in depression and anxiety categorisations for these mothers across the two time points . The majority of mothers stayed in the same depression and anxiety categories across the two time points, for example, 67% who had no symptoms of depression in the pre-COVID-19 data continued to have no symptoms during lockdown; and 54% of those who had moderate/severe symptoms of depression before the pandemic remained in this category in the lockdown survey. However, many mothers' mental health worsened, with 38% who had no/mild symptoms of depression pre-COVID-19 reporting moderate/ severe symptoms in the lockdown survey. A smaller number of mothers' mental health improved: of those mothers with moderate/severe symptoms of depression pre-COVID-19, 24% subsequently reported no symptoms and 23% reported mild depression in the COVID-19 survey. Similar patterns of change were seen with the anxiety categories. Factors associated with a clinically important increase in depression and anxiety Three hundred and sixty-seven mothers reported a clinically important increase in depressive symptoms, and 348 reported a clinically important increase in anxiety symptoms. --- Open access Table 4 presents the ORs for a clinically important increase in PHQ-8 and GAD-7 scores in relation to each exposure variable from the unadjusted logistic regression models. The estimates resulting from these models are imprecise, with wide CIs. Financial, food and housing insecurity during lockdown was associated with a higher likelihood of a clinically important increase in both depression and anxiety symptoms: the odds were more than six times greater for women who were financially insecure and over three times greater in mothers who were food insecure or housing insecure . Social circumstances were also associated with an increase in depression and anxiety: the odds of increased depression or anxiety were more than eight times greater in mothers reporting loneliness , and a lack of social support doubled the likelihood of an increase in depression and anxiety . An average/poor relationship increased the odds of experiencing symptoms of depression by 3.6 and of anxiety by 5.1 . A larger household size reduced the OR for depressive symptoms , but not for anxiety . A lack of physical activity during lockdown was associated with an increased OR for both depression and anxiety . Mothers who did the same amount of physical activity during lockdown as they had done pre-COVID-19 reduced the odds of an increase in depressive symptoms by 34% . There was no clear association between ethnicity and an increase in depression or anxiety . When the unadjusted regression analysis was repeated separately for White British and Pakistani heritage mothers, interesting differences in the magnitude of the association between exposure variables and a clinically important increase in symptoms of depression and anxiety were found, although CIs are wide . The odds of an increase in depression and anxiety were greater for Pakistani heritage women who reported loneliness compared with White British mothers . There was also a greater magnitude of association for increased depression, but not anxiety, for Pakistani heritage women who reported an average/poor relationship with their partner , compared with White British mothers . In contrast, the magnitude of the odds of an increase in depression and anxiety was greater for White British women who reported financial insecurity (OR=12. Open access --- Open access For Pakistani heritage mothers living in large households, the odds of an increase in depression was reduced compared with White British mothers for whom there was no association . There were no clear associations between ethnicity and household size for anxiety . --- Mothers Lived Experience Free text responses to the question 'What are your three biggest worries at the moment?' were available for 1799 mothers. Only a small proportion of women identified their mental health issues as one of their biggest worries, n=105 , slightly greater in White British mothers, n=51 than in mothers of Pakistani heritage, n=32 . More often, mothers reported how wider issues and concerns impacted on their mental health and well-being. --- Health anxieties about COVID-19 The most commonly reported worry was fear of bringing the virus home , and themselves or members of their family becoming ill or dying, as well as the fear of what would happen to their children if this did happen to them. I worry about contracting coronavirus particularly whilst at work and either becoming critically myself unwell or bringing it home to my family. Feeling anxious about the virus and constantly worrying about my kids which 2 of them have health issues and are quite vulnerable. I feel particularly anxious to even step out of the house even for food shopping or taking a walk/ exercise. I worry how this will affect my children. I'm terrified they will be separated as I have 2 children with my ex-husband and one with my current. So I haven't been outside in 10 weeks. --- Mental load Mothers often reported the mental load of managing work, home schooling, childcare and domestic tasks, without the break provided by children attending school, nursery or other childcare. Being or feeling stuck inside and unable to move around freely contributed to a sense of suffocation and feeling overwhelmed, and many mothers acknowledged that this was having a detrimental effect on their mental health and self-esteem: I'm worried [about] having a nervous breakdown or a panic attack…can't get a break from all the responsibilities and go somewhere for fresh air even. --- Loss of social support The loss of social support caused by lockdown, especially for those who did not live with their partner or were single parents, was highlighted as causing loneliness and isolation for some. My support network was my partner, who I cannot see due to lockdown. Not able to see my boyfriend-feel isolated and alone once children are in bed with no adult face to face interaction. The isolation and loneliness has been a challenge and having to take care of my child on my own full-time. Being a single parent of a disabled child I rely on my social life/friends for my mental wellbeing and I worry about when it'll be safe to see/hug my immediate family who live locally. --- Financial and employment insecurity Household finances and the stress of unemployment or job insecurity related to lockdown measures were also a major worry for women. Many families were in debt, reliant on credit cards and benefits, and in insecure employment before the lockdown measures began and were only just about getting by prior to the pandemic. I have struggled financially during this time. I would like to not worry about money and bills and shopping and outgoings. I have enough worries being a full time carer. Worried about the financial impact of covid 19. I am currently furloughed from work but I worry that the virus will have an impact on the business. My husband is self-employed and is not eligible to any funds. Being unable to switch off Participants described being frightened of the news reports but unable to switch off, and were wondering when, if ever, things would become normal again: Worry about my mental health, I know I'm struggling. Worry about the future and how this will affect the country financially. Worry about my children's education & what they've missed out on. --- Open access All the bad news on the TV, and the death rate on the News. All the information on the news makes me panic more. When will this end? Will it end? A loss of coping strategies For those who had existing mental health issues before lockdown, the lockdown measures had often taken away their sources of support, their normal routines and methods of coping. In addition, some respondents reported being unable to access mental health services due to COVID-19 and lockdown measures: Mental health-Exercise at the gym was a coping mechanism, now closed. Can do bits at home but nothing like at the gym. Cannot get any gym equipment-out of stock. Mental Health-struggling to motivate myself and to keep to a routine. Self-destructive behaviour-drink drugs and binge eating. I am missing my work and colleagues. Not getting my mental health support since the lockdown. My CPN not returning my calls. It has made me a lot worse. I try to talk to my husband so I am not keeping everything inside. Mental health, I have had previous issues in the past and am struggling and don't feel like I can approach my GP at the minute as it isn't an emergency. --- Positive aspects of lockdown Many participants reported positive aspects to changes enforced by the lockdown, commenting that they were getting to spend more quality time with their children, were enjoying a slower pace of life, a more relaxed routine and spending less time driving and commuting. Life has become a lot more relaxed over the last 3 weeks, no manic mornings trying to get everybody out of the house, time with kids, doing stuff with kids I would normally say 'not now' to. Get to know kids more. More time outside, [doing] jobs in the house that need doing. Spending more time with my husband and feeling appreciative of each other and having a relaxed day with the children instead of running to school to mosque and then all the extra clubs it's a more relaxed day. Ramadan is the easiest it has ever been, we are free to make up our sleep and not push ourselves too much, had time to do nice things during Ramadan including having a more peaceful time not having to Open access do school runs, be stressed out, my husband had a chance to take a slower pace to life and not get too stressed. --- DISCUSSION We compared depression and anxiety during the March-June 2020 UK lockdown to pre-COVID-19 depression and anxiety data collected in our longitudinal birth cohort studies. We found that clinically important symptoms of depression and anxiety increased from 11% to 20% and from 11% to 16%, respectively. These findings reflect those of other longitudinal studies which have reported similar changes in mental health from before to during the COVID-19 pandemic. [3][4][5] We hypothesised that our vulnerable population with diverse ethnicity and high levels of deprivation would be susceptible to increases in depression and anxiety during the lockdown. Financial, food and housing insecurity all increased the odds of an increase in depression and anxiety, as did loneliness, a lack of social support, an average/poor partner relationship and a lack of physical activity. There was no clear association between White British and Pakistani heritage mothers and a clinically important increase in depression or anxiety. However, when we separated out the regression analyses by ethnicity, we found interesting differences in the magnitude of the associations with an increase in depression and anxiety: mothers from a Pakistani heritage had greater odds of an increase in depression and anxiety if they were lonely or had an average/poor relationship than White British mothers. Pakistani heritage mothers had a much reduced odds of an increase in depression if they lived in a large household compared with White British mothers . In contrast, mothers of White British ethnicity had greater odds of an increase if they were financially insecure and/or physically inactive compared with Pakistani heritage mothers reporting the same exposures. The free text responses supported these findings, highlighting a complex interplay of challenges contributing to poor mental health in mothers including: acute health anxieties; the mental load of managing multiple roles and responsibilities; the loss of social support and other coping strategies; pressures of financial and employment insecurity; and being unable to switch off from the pandemic. The potential ethnic differences in the magnitude of the associations of different variables and increased mental ill health reported in this study warrant further investigation, including an understanding of potentially differing protective factors in different ethnic groups. We have previously reported, outside of the COVID-19 context, that White British mothers are more likely to have their mental ill health identified by health professionals than South Asian mothers, and that both White British and South Asian mothers are equally likely to disclose symptoms in self-report research questionnaires, as used here. 23 24 Ethnic differences in household structure and cultural practices might provide more support in times of adversity, as we have found in the case of food insecurity previously. 25 26 This is a longitudinal study comparing data collected before the COVID-19 pandemic and during the March-June 2020 lockdown which has allowed us to explore change over that time period. It also provides findings from a highly ethnically diverse population, the majority of whom live in the most deprived centiles in the UK. We are not aware of other studies that have explored longitudinal change in mental health from pre-COVID-19 to during the COVID-19 lockdown in a similar population. Respondents were mothers of children aged 0-5 and/ or 9-13 which may limit the wider generalisability, though our findings are broadly similar to those from a previous longitudinal study of two UK cohorts that included adult men and women , and found the increased risk of poor mental health in lockdown to be greater in women. 4 Our pre-COVID-19 measures were taken from data collected over the past 4 years, so we cannot, with confidence, attribute all changes to the pandemic and the lockdown. For example, it is possible that some of the difference reflects age-related change in the women and/or their children over time. It is also possible that we have underestimated some of the adverse impact of lockdown as a significant percentage of the BiBBS participants were pregnant at baseline, which itself is associated with raised levels of mental ill health. However, our analysis did not find increased levels of depression and anxiety for the BiBBS cohort compared with the BiBGU cohort. Similarly, our analyses did not find any association between time since baseline data capture and the odds of an increase in depression and anxiety, suggesting that the timing of our baseline collection is not influencing our findings. We undertook unadjusted logistic regression analyses to explore possible factors that might explain or influence changes in mental health from before to during the pandemic, and repeated these separating out the two main ethnic groups. This approach avoids the difficulties inherent in interpreting the ethnicity coefficient in regression models controlling for other variables 20 ; however, it also limits the interpretation of the data. In addition, the total sample size for this mental health analysis was ~1700 mothers and many estimates in the analyses were, as a consequence, imprecise with wide CIs. These results demonstrate the need for further studies with sufficient sample sizes, including boosted samples of ethnic minority groups to enable accurate understanding of the possible differing experiences and needs of the diverse UK population. While it is possible that our results are influenced by the survey response rate of 28%, these participants were representative of the BiBGU and BiBBS cohorts, 15 including on the rates of baseline depression and Open access anxiety scores, and have demonstrated a wide variability in most characteristics . We undertook stratified analyses to explore ethnic differences in how exposures related to changes in mental health from before to during the pandemic. However, while our analysis was able to look for differences between Pakistani heritage and White British mothers, we were unable to explore the heterogeneity of the 'other' ethnic group and we have no doubt missed important nuances in this population based on different social and cultural experiences. We will continue to follow the BiBGU and BiBBS families so that we can look at trajectories of change over time, including during any future national or regional lockdowns. At the time of writing , the UK was under a further nationwide lockdown, following a period of reduced restrictions during summer 2020 and tighter regional restrictions in the more deprived cities of Northern England, including Bradford, through the autumn. The impact of further and longer periods of restrictions and lockdowns will be a focus of our study moving forward. As the longitudinal follow-up lengthens and more data become available, it may be possible to model different transitions in mental health in relation to changing social and economic circumstances. It may also be of interest to explore positive/null transitions from clinically important symptoms to mild/no symptoms, or continued mild/no symptoms despite reported exposure to adversity. Our results highlight the potential public health impact of lockdown on mental health, particularly in those who are lonely and economically insecure. Mental health problems are in general less visible than physical symptoms and in particular the physical symptoms related to COVID-19, including the acutely ill patients in Intensive Care Units, but may have more significant longer term consequences. This study also suggests a complex interplay of factors between an individual's circumstances and the odds of worsening mental health during the pandemic. A 'one-size-fits-all' approach to supporting mental ill health will not be effective; instead, understanding and addressing the potential causes in different groups will be important. For example, in case of future need for social distancing measures, government and local councils should consider policies that permit 'social bubbles' that can be implemented to reduce loneliness for those at risk of mental health problems, and they as well as voluntary services should continue to focus support to those who are lonely/isolated. Policy and decision makers should also make provision for the continuing need to support and protect vulnerable families from financial, food and housing insecurity, all of which were associated with poor mental health in this study. These actions will be important throughout any further regional and national lockdowns and during the postpandemic period of recovery. Contributors JD was involved in the concept and study design, design of the data collection tools, overall supervision of the study and data collection; wrote the statistical analysis plan; and drafted and revised the paper. She is the guarantor of the study. BK was involved in the design of the data collection tools, wrote the statistical analysis plan, conducted all statistical analyses, and drafted and revised the paper. BL devised the coding framework for free text responses, analysed the data, and drafted and revised the paper. SB was involved in the concept and study design, design of the data collection tools, supervision of the study and data collection, and reviewing and revising the paper. CC was involved in the concept and study design, and reviewing and revising the paper. KW was involved in design of the data collection tools, data curation, analysis, and reviewing and revising the paper. KS was involved in the study design, design of the data collection tools, supervision of the study and data collection, and reviewing and revising the paper. KC was involved in the study design, design of the data collection tools, supervision of the study and data collection, data curation, and reviewing and revising the paper. MB was involved in the study design, and drafted and revised the paper. NS advised on the clinical significance of changes in mental ill health, revising the statistical analysis plan, and drafted and revised the paper. TS was involved in the concept and study design, supervision and support to junior members of the team, and drafted and revised the paper. DL was involved in the concept and study design, design of the data collection tools, and drafted and revised the paper. JW was involved in the concept and study design, design of the data collection tools, supervision and support to junior members of the team, and drafted and revised the paper. RM was involved in the concept and study design, design of the data collection tools, revised the statistical analysis plan, and drafted and revised the paper. KP was involved in the concept and study design, design of the data collection tools, supervision and support to junior members of the team, revised the statistical analysis plan, and drafted and revised the paper. --- Competing interests All authors had financial support from funding bodies listed for the submitted work. DAL declares previous support from Roche Diagnostics and Medtronic for research unrelated to that presented here. Patient consent for publication Not required. --- Open access Ethics approval This study involves human participants and was approved by the HRA and Bradford/Leeds Research Ethics Committee . Participants gave informed consent to participate in the study before taking part. Provenance and peer review Not commissioned; externally peer reviewed. Data availability statement Data are available upon reasonable request. Born in Bradford offers open access to their data resources. Available data and procedures to access this can be found at: https://borninbradford.nhs.uk/research/how-toaccess-data/_ Supplemental material This content has been supplied by the author. It has not been vetted by BMJ Publishing Group Limited and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations , and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Objectives To explore clinically important increases in depression/anxiety from before to during the first UK COVID-19 lockdown and factors related to this change, with a particular focus on ethnic differences. Design Pre-COVID-19 and lockdown surveys nested within two longitudinal Born in Bradford cohort studies. Participants 1860 mothers with a child aged 0-5 or 9-13, 48% Pakistani heritage. Main outcome measures ORs for a clinically important increase (5 points or more) in depression (eight item Patient Health Questionnaire (PHQ-8)) and anxiety (Generalised Anxiety Disorder Assessment (GAD-7)) in unadjusted regression analyses, repeated with exposures of interest separated by ethnicity to look for differences in magnitude of associations, and lived experience of mothers captured in open text questions.The number of women reporting clinically important depression/anxiety increased from 11% to 20% (95% CI 10%-13%; 18%-22%) and from 10% to 16% (95% CI 8%-11%; 15%-18%), respectively. Increases in depression/anxiety were associated with loneliness (OR=8.37, 95% CI 5.70 to 12.27; 8.50, 95% CI 5.71 to 12.65, respectively); financial (6.23, 95% CI 3.96 to 9.80; 6.03, 95% CI 3.82 to 9.51), food (3.33, 95% CI 2.09 to 5.28; 3.46, 95% CI 2.15 to 5.58) and housing insecurity (3.29, 95% CI 2.36 to 4.58; 3.0, 95% CI 2.11 to 4.25); a lack of physical activity (3.13, 95% CI 2.15 to 4.56; 2.55, 95% CI 1.72 to 3.78); and a poor partner relationship (3.6, 95% CI 2.44 to 5.43; 5.1, 95% CI 3.37 to 7.62). The magnitude of associations between key exposures and worsening mental health varied between ethnic groups. Responses to open text questions illustrated a complex interplay of challenges contributing to mental ill health including: acute health anxieties; the mental load of managing multiple responsibilities; loss of social support and coping strategies; pressures of financial and employment insecurity; and being unable to switch off from the pandemic. Conclusions Mental ill health has worsened for many during the COVID-19 lockdown, particularly in those who are lonely and economically insecure. The magnitude of associations between key exposures and worsening mental health varied between ethnic groups. Mental health problems may have longer term consequences for public health and interventions that address the potential causes are needed.► Three key longitudinal studies have highlighted that the COVID-19 pandemic and lockdowns have had a negative impact on mental health, particularly in younger adults, women and those from low socioeconomic circumstances, but with participants of predominantly White European ethnicity. ► The Born in Bradford research programme offers a unique opportunity to investigate the impact of COVID-19 lockdown on mental health in a deprived and ethnically diverse population in whom mental ill health is often reported to be more prevalent. ► This is a longitudinal study containing linked data collected before the COVID-19 pandemic and during the March-June 2020 lockdown which has allowed us to explore change over that time period in a highly ethnically diverse population, the majority of whom live in the most deprived centiles in the UK. ► Respondents in this study were mothers of children aged 0-5 and/or 9-13 which may limit the wider generalisability, though our findings are broadly similar (in prevalence and associations) to those from another longitudinal study that included adult men and women. ► We are not aware of other studies that have explored longitudinal change in mental health from before to during the COVID-19 lockdown in a similar ethnically diverse and deprived population.

Introduction The healthcare system has a focus on patient empowerment today with the goal to empower patients to become active participants in their care [1]. This requires not only an understanding of patients' experiences from disease and illness, but also expectations of healthcare professionals providing patient involvement and organizational support for these efforts [2]. Empowered patients are described through several concepts in the literature, often as active, engaged, empowered, and knowledgeable participants, even though there is no consensus of the definition in the literature [3]. In the early 2000s the term e-patient was initially coined by Ferguson and Frydman inspired by the digital development within society which was reflected in many patients' and informal caregivers' behaviors [4]. E-patients were originally described as patients or informal caregivers who used the Internet to find guidance regarding their own, or someone else's, health challenge. E-patients use their own health data to learn from and engage in innovations of digital tools to meet health-related needs. They find ways to actively make their current situation more sustainable, rather than passively waiting for potential future solutions or cures [5][6][7]. A concept in close relation to e-patients, is patient lead user defined by Riggare [8] as: "Patients or family members who take a larger responsibility for their own health and well-being. They meet their health-related challenges in a constructive and knowledge-based way, while taking their physical and mental abilities as well as capacity into account. Patient lead users make use of their own experiences to improve healthcare, on all levels of the system, for the sake of both themselves and other patients. Often you do not become a patient lead user by choice, it is something that you do to be able to manage and navigate the complex healthcare system" [8]. The concept of patient lead user originates from Von Hippel's lead users, describing lay persons having a great need within a specific context, without any available solutions at the market. Lead users will then innovate solutions for that need, and the general market will follow [9,10]. These concepts describe patients' and informal caregivers' active engagement, however their self-management tasks and healthcare system involvement are further described in a recent study of empowering behaviors of patients and informal caregivers. Here exploratory and influencing activities that patients and informal caregivers perform are characterized [11]. The exploratory empowering behaviors are labelled as the self-care expert, the knowledge seeker, the academic, the patient researcher, the tracker, the coping-expert, and the exposed and the influencing behaviors as the innovator, the entrepreneur, the communicator, the mentor, the healthcare coordinator, the healthcare partner, and the activist [11]. These empowering behaviors have evolved in parallel to the digital transformation in society [7] and require patients and informal caregivers to be involved and receive feedback. Due to the development of digital solutions for selfmanagement, technological information systems, and focus on patient involvement during the 21st century in Sweden [1], patients' and informal caregivers' behaviors have changed. Swedish healthcare is nationally regulated, tax-funded, and locally administrated. Inhabitants of all social groups are entitled to a strong safety net since everyone have the same benefits and a maximum out-ofpocket cost [12,13]. There is a high use of the Internet in the Swedish population [14], and, in line with other Scandinavian countries, longterm experiences of end user involvement in digital technology [15]. This might increase a movement towards more empowered patients and informal caregivers in Sweden. However, studies researching healthcare professionals' knowledge and perception of these patient behaviors are scarce. Some studies describe healthcare professionals' concerns about patients' online information seeking which might lead to stressful encounters with unrealistic expectations from patients [16,17], as well as the concern of patients or informal caregivers finding harmful information [17]. Other studies show that healthcare professionals believe ideas and questions raised from patients' online information searches could improve diagnostic decision making [18] and the patient-professional relationship [19]. There seem to be opportunities for healthcare professionals to ensure quality-controlled information access of their patients, for example by providing recommendations of websites, referrals to different peer-communities , explaining or discussing research results with patients, or including patients' own self-generated data into medical decision making [20]. Still, research indicate a continued concern from healthcare professionals regarding patients and informal caregivers searching medical information online [17,20]. More knowledge is needed regarding the experiences and attitudes of healthcare professionals when encountering empowered patients and informal caregivers in clinical settings, and to what extent clinical workplaces offer strategies and support for this. Could empowered patients and informal caregivers contribute to improving the quality of health care? Therefore, the aim of this study was to investigate healthcare professionals' attitudes towards and experiences of working with empowered patients and informal caregivers, and perception of workplace support with regards to collaborating with empowered patient and informal caregivers. --- Method --- Study and survey design This cross-sectional study had a descriptive design [21] with a multi-centre web survey conducted in Swedish. Using descriptions of empowering behaviors from a previous study as a guide [11], a study specific questionnaire was developed by the research team. The survey questions are presented in an additional file [see Additional file 1]. Before distributing the web survey, six cognitive interviews were performed to validate the questions. The web survey was pilot tested among 30 healthcare professionals before the final distribution. The final web-survey consisted of a demographic section with five closed multiple-choice options, nine sections containing a total of 42 statements and nine free text options, as well as an additional section with four open-ended response questions. The open-ended questions and free text options were used as follow-ups to the closed multiple-choice options and non-mandatory. All other questions were mandatory to answer. --- Data collection Non-probability sampling was performed, acknowledging the consequence of an exploratory sample not reaching the whole study population [22]. The link to the survey was first distributed to operational managers at different specialist healthcare settings managing oncology, diabetes, neurology, psychiatry, and rheumatic patients. These settings were representing the five patient driven innovations facilitating co-care within the research programme entitled "The patient in the driver seat", that this study is part of. Managers distributed the link to their employees' e-mail addresses. Two reminders were sent out. However, this approach resulted in a low response rate . As a second distribution, an advertisement strategically targeting healthcare professionals with interest in health care, speaking Swedish, and age over 18 years old, was posted in social media for two weeks. The advertisement was distributed by Karolinska Institute's Communications and public relations office, and reaching healthcare professionals within specialized, primary, and secondary healthcare settings in Sweden. Inclusion criteria for both samples were being a healthcare professional, older than 18 years, and Swedish speaking. The web survey was developed, distributed, and collected through Microsoft Teams, and data were collected between April 2021 and February 2022. --- Data analysis A descriptive quantitative approach was used to analyse demographic characteristics and survey responses from a frequency perspective [21]. Similar responses were dichotomized according to being positive or being negative . The analysis was performed using Microsoft Excel © [23]. Survey responses were summarised into three main topics based on the nine sections in the survey. The nine sections are displayed in an additional file [see Additional file 1]. A brief Thematic analysis was performed of the qualitative data from the open-ended responses, to see if the narrative changed. An inductive coding was performed when grouping codes generated from the data into categories. From this a deductive approach was performed using the three main topics to locate the categories [24]. --- Results --- Demographics In total, 279 healthcare professionals completed the survey. In the first distribution, 478 surveys were sent out, resulting in a response rate of 18% . The second distribution, through social media, resulted in 193 completed responses , equivalent to a response rate of 36%. Respondents from both distributions included healthcare professionals working in different specialist settings, as well as primary and secondary healthcare . In sample one 26% were physicians compared to 13% in sample two. Respondents in sample one answered the free-text questions to a greater extent in comparison to sample two. Further, respondents in sample one reported to be more aware of the concept of patient lead users compared to in sample two. No other differences between the samples were found. The demographic characteristics of both samples are presented in Table 1. They show a fair distribution across age, gender, workplace, and occupation. However, specialist healthcare professionals are somewhat over-represented, as well as women, reflecting the female dominance in Swedish health care. The data from both data collections were summarized under three main topics, based on the nine sections from the web survey [see Additional file 1]: Patient knowledge , Innovative patient self-care behaviors , and, Patients navigating the healthcare system . Results are presented per topic, listing respondents' attitudes, experiences, and perceived workplace support for each. Free-text responses are summarized under each topic, illustrated by a quote. --- Patient knowledge The respondents showed a positive attitude towards knowledgeable patients and informal caregivers and stated the importance of and their interest in effective knowledge exchange. The respondents were also positive towards patients and informal caregivers sharing knowledge and experiences with each other, since it has the potential to lead to better everyday life Fig. 1 Flow chart for the data collection process experiences. However, the free-text responses revealed that individual patient knowledge might also be irrelevant if patients and informal caregivers lack an understanding of the overall situation. Further, respondents mentioned the risk of incorrect knowledge being shared, which might lead to unrealistic expectations and demands from healthcare professionals. This was described by some respondents to make it harder to build a mutual trusting patient-healthcare professional relationship. --- Intellectually, it is generally positive with knowledgeable patents and informal caregivers, however, emotionally there is one negative aspect. There is a concern of getting out of balance in my profession if they are highly competent and at the same time crit-ical… If so, they might not feel confidence in me as a healthcare professional and this will aggravate our partnership. Most respondents had experiences of learning from patients and informal caregivers . Respondents also described experiences of discussing existing knowledge with their patients and the informal caregivers . Such discussions often included the latest standard of care, patient symptoms, and the complexity of the condition. However, few respondents stated having regular follow-ups regarding what they learned from individual patients and informal caregivers . In the free-text responses, some respondents described not having the time and support from their workplace to do so. Half of the respondents stated having experiences of encouraging patients and informal caregivers to share their experiences with their peers, and almost the same number of participants confirmed that their workplace supported them to do so . Overall, the free-text responses showed that respondents had great support from their co-workers regarding existing challenges with demanding patients and informal caregivers, but less support from existing structures within their workplace and from operational management. --- Innovative patient self-care behaviors Most participants reported a positive attitude towards patients and informal caregivers creating innovations for their own health conditions. Similar results were reported for patients and informal caregivers that used digital solutions to manage their own condition. The respondents also showed a positive attitude towards patients that use alternative ways to interact with health care. Patients who perform selftracking on their own initiative were considered positive by the participants. A majority of the respondents confirmed having given feedback various times to patients own collected data . More than half of the respondents had experiences of using digital solutions together with patients to some degree, however it was less common together with informal caregivers . The respondents also reported some experience from patient innovations , but less from innovations performed by informal caregivers . However, only 34% of the respondents perceived that their workplace provided the right conditions for them to manage innovations created by patients or informal caregivers. Just over half of the respondents stated to have a workplace that regularly followed up patients' needs for alternative ways to interact. Half of the respondents considered their workplace to be encouraging towards patients performing self-tracking to some degree . Some concerns were raised related to patients using non-validated methods for Other 0 30 self-tracking which was considered being too subjective. Even though the healthcare professionals' attitudes were positive regarding use of digital solutions, they reported that their organizations were not encouraging to the same extent . In the free-text responses, some respondents described negative experiences such as increased anxiety for patients and informal caregivers using digital solutions, since they were not considered evidence-based. Some respondents also expressed concerns that digital solutions potentially could lead to less face-to-face interactions, increased inequality, and unnecessary additional work for healthcare professionals. Overall, the respondents reported that the main challenges were lack of time and flexibility when engaging with empowered patients and informal caregivers. To "google" your symptoms and condition might be very misleading. Since treatments can be individually adapted, these are often questioned which leads to long discussions that we do not have time for, unfortunately. --- Patients and informal caregivers navigating the healthcare system Most respondents had positive attitudes towards patients engaging in the development of health care . It was not considered equally positive when informal caregivers engaged in the development . A majority of the respondents showed positive attitudes towards patients and informal caregivers coordinating their own healthcare contacts between different healthcare units. Some respondents raised a concern that not everyone can or should coordinate their own health care. The free-text responses also indicated that the respondents considered themselves helping patients and informal caregivers coordinating their healthcare contacts. However, only 58% confirmed to be doing so in the closed survey questions . Even though it was considered positive with patients and informal caregivers navigating the healthcare system in different ways, relatively few respondents stated that they had experience of encouraging patients to engage in the development of the clinical workplace. Even fewer respondents reported having encouraged informal caregivers to do so . Respondents' experiences were often exclusively connected to engagement with patients, but less with informal caregivers . However, in the free-text responses some of the respondents reported trying to involve informal caregivers in the patients' care. Some respondents saw possibilities for patients and informal caregivers to engage in the development of the clinical workplace, because of existing inadequacies in the healthcare system. These inadequacies could be lack of organizational guidelines, routines, resources, and forums for engagement. It is difficult to encourage this when there is no existing plan for it within our organization. --- Discussion Overall, the healthcare professionals participating in this survey revealed a positive attitude towards empowered patients and informal caregivers, despite lack of experiences of working together with them. Only a small proportion of respondents considered their organizations to provide the optimal conditions to involve patients and informal caregivers as well as support the respondents when difficult situations occurred. The questionnaire developed for this study was based on the self-empowering behaviors found by Scott Duncan et al. [11]. Meeting the influencing self-empowering behaviors [11] showed that our respondents had positive attitudes towards patients and informal caregivers sharing their knowledge with other peers as mentors or communicators. The respondents had encouraged this to some degree, and it existed some organizational support. It was as well positive attitudes regarding patients innovating for their needs, as well as for informal caregivers . The respondents were to some degree involved in patient and informal caregiver innovations. However, the support from their organization regarding managing these innovations was rather low . Being engaged in healthcare development as activist or healthcare partner was considered positive if being a patient and somewhat positive being an informal caregiver . However, workplace support was rather low for involving patients and informal caregivers in healthcare development. Even though there are overall positive attitudes from healthcare professionals, it was reported in Scott Duncan et al. [11] that patients and informal caregivers considered having low support for their efforts and wished to do more than was expected from them by healthcare professionals [11]. The respondents also considered it to be positive when patients and informal caregivers coordinated their own health and care, whereas patients and informal caregivers considered coordinating their own care as a burden [11]. The respondents did as well have less experience of helping out coordinating the care , even though patients and informal caregivers have reported a need for better support [11]. The participants' positive attitudes are contradictory to other studies [25,26], where rather negative attitudes are displayed from healthcare professionals regarding patient involvement. Less experiences of working together with empowered patients and informal caregivers among our respondents could be the result of different barriers for patient involvement. One barrier reported in the literature is lack of communication and confidence of physicians [27,28]. Other barriers reported are the paternalistic structure within the healthcare organization and lack of time and encouragement [28,29]. It seemed to be even more difficult to involve informal caregivers for our respondents. The literature expresses lack of involvement, support, and being acknowledged by healthcare professionals as informal caregivers [30][31][32] as a confirmation of these barriers. A weakness of our study is the low response rate. Still, the sample included a total of 279 participants in an online survey despite a demanding time in society and health care because of the COVID-19 pandemic. Web surveys also tend to generate lower average in response rate than paper and telephone surveys [33]. Not using a validated survey could be a weakness, however, both cognitive interviews and pilots were performed to ensure the relevance of the questions. A non-probability sampling was used, which might have caused a selection bias and a skewed sample [23], since we do not know if the respondents were more positive towards empowered patients and informal caregivers in general. Given the healthcare professionals' positive attitudes, future research needs to investigate how healthcare systems can better meet the willingness to involve empowered patients and informal caregivers to a larger extent. --- Conclusion The healthcare professionals who responded to this survey lacked experiences of working with empowered patients and informal caregivers, and workplace support for involving patients and informal caregivers. It was considered difficult to engage empowered patients and informal caregivers in the way the respondents preferred, and could lead to less active patient participation. The overall positive attitudes of healthcare professionals suggest however possible implications for health care and policymakers to provide better structures for involving patients and informal caregivers. These positive attitudes are a fundamental prerequisite towards a transition of the healthcare system into recognizing empowered patients and informal caregivers as partners. --- Data Availability According to the EU General Data Protection Act personal data is restricted and anonymized. Data in Swedish is available by the corresponding author. --- --- Supplementary Material 1 --- Author Contribution All seven authors have contributed to the study design, reviewed, and endorsed the final manuscript. The first author collected the data, and T.SD, S.R, and A.B analyzed the data. T.SD and S.K wrote the manuscript. --- Declarations Ethical approval and consent to participate An ethical approval has been applied for, however, according to Swedish Ethical Review Authority has ruled that the legislation regarding ethical review is not applicable for this study, according to decision 2020-05805. We do not have an institutional review board, it is a review authority in Sweden. Written information about the aim of the study, description of the data management, and the possibility to opt-out at any time were provided to the participants before starting the survey. All participants gave informed consent to participate in the survey, and all methods were conducted according to relevant guidelines and regulations in the Declaration of Helsinki as well as the European General Data Protection Regulation . --- --- Competing interest None to declare. ---

Background More knowledge is needed regarding the perceptions of healthcare professionals when encountering empowered patients and informal caregivers in clinical settings. This study aimed to investigate healthcare professionals' attitudes towards and experiences of working with empowered patients and informal caregivers, and perception of workplace support in these situations. Methods A multi-centre web survey was conducted using a non-probability sampling of both primary and specialized healthcare professionals across Sweden. A total of 279 healthcare professionals completed the survey. Data was analysed using descriptive statistics and Thematic analysis.Most respondents perceived empowered patients and informal caregivers as positive and had to some extent experience of learning new knowledge and skills from them. However, few respondents stated that these experiences were regularly followed-up at their workplace. Potentially negative consequences such as increased inequality and additional workload were, however, mentioned. Patients' engagement in the development of clinical workplaces was seen as positive by the respondents, but few had own experience of such engagement and considered it difficult to be achieved .Overall positive attitudes of healthcare professionals are a fundamental prerequisite to the transition of the healthcare system recognizing empowered patients and informal caregivers as partners.

In 2018, I moved from inner-city Cardiff to Mountain Ash, in the South Wales valleys. Having been a GP for 35 years in suburban Nottingham, inner-city Manchester, and inner-city Cardiff, I was not expecting to be as surprised as I was by the clear demonstration of poverty's effects on health that I found there. Sited in a beautiful Welsh valley, with the Cynon River running along its course, a walk down Mountain Ash's high street shows a line of fast-food outlets, nail and tattoo bars, betting shops and beauty parlours. The Post Office closed last year, there are no banks, greengrocers, bakeries, or butchers' shops. Many of my patients have BMIs well over 35, and the concomitant list of chronic diseases that go with that. Two weeks in, I said to my new partners over coffee: 'Aren't we missing a trick here? Although we have treatments for these chronic conditions, wouldn't it have been better to have helped patients avoid getting them in the first place?' They looked at me in horror and said that there was far too much that needed to be done within the surgery walls for time to be spent in schools, community centres, sports centres trying to prevent the development of chronic disease, with outcomes that were uncertain and difficult to measure. Many GPs would agree with that, and certainly our current workload does not leave much headspace for anything more than the patient in front of us GPs over the same period of time 1,2 ). So, is that correct? Should we be keeping our noses to the appointment grindstone, day after day, mopping up the problems created by unhealthy lifestyles and adverse childhood events? In South Wales, our health legacy includes the amazing Dr Julian Tudor-Hart. Julian was a GP who spent his life serving a small Welsh Valleys town and conducting seminal research with that community on blood pressure management, salt and dietary intake, among other trials. In 1971, he coined the famous and widely quoted 'Inverse care law' 3 -that those who need care the most are the ones least likely to get it. 50 years later, we see that happening today, across the globe, but also in our own practices. Julian also said that 'medicine without social justice, is not medicine at all' -would we agree? And he said that research was important -not only to help us decide what the best treatments and management plans were, but also because that data brought power with it. Power to demonstrate to those who make the funding and planning decisions what effect their decisions will have on the lives of the communities that make up the UK. Why should we worry about the health inequalities gap? In addition to knowing that poorer communities and vulnerable groups have a shorter lifespan 4 and more chronic illness, 5 we also know that those countries where the gap between rich and poor is wider have worse outcomes in eleven different health and social problems, including physical and mental health, drug abuse, education, imprisonment, obesity, trust and community life, teenage pregnancies and child wellbeing. 6 And this affects both rich and poor in these countries. In the UK, the health inequalities gap has been widening since the publication of the Black Report in August 1980, 7 so we should be worrying about what is happening in the UK. And as Sir Michael Marmot says so succinctly in his book 'The Health Gap', 8 'What is the sense of sending people back to the same conditions that made them sick in the first place?' When the COVID-19 pandemic struck, it cast a harsh light on the health and wider inequalities that persist in our communities. The virus and efforts to mitigate its impact had a disproportionate effect on certain sections of the population, including ethnic minority communities, the most socio-economically deprived communities, people with learning disabilities, those with severe mental illness and inclusion health groups. Even before the pandemic, the issue of health inequality had already been identified as major issue to be addressed across the NHS and the Long-Term Plan 9 made tackling health ABSTRACT LEADERSHIP Kamila Hawthorne and Bola Owolabi prevalence and impact of long-term conditions and how other risk factors that cause and exacerbate some of these LTCs, such as smoking rates, have remained stubbornly persistent in poorer areas despite significant declines more generally. 12 When I started in my role as director for the National Healthcare Inequalities Improvement Programme in NHS England just over 2 years ago, I made it my mission to speak extensively with people and organisations across the country to gain insights into why, despite decades of publications on health inequalities, measurable progress remained elusive. I was struck by the refrain 'Don't boil the ocean'. In essence, there was a general view that in an effort to tackle all the causes and manifestations of health inequalities at the same time, energy and resources were being distributed too thinly to be able to make demonstrable impact on any dimension. I surmised from these discussions that we needed focus. Focus, in service of gaining traction and making demonstrable impact within a reasonable time frame. My mantra therefore becamefocus, traction, impact. With wider conversations with fellow GPs and many other groups, and further reflection, I began to see the real value of distinguishing between 'health inequalities' and 'healthcare inequalities'. The former are unfair and avoidable differences in health across the population, and between different groups within society. These include how long people are likely to live, the health conditions they may experience and the care that is available to them. The conditions in which we are born, grow, live, work and age can impact our health and wellbeing. These are sometimes referred to as wider determinants of health. 13 Healthcare inequalities by contrast are indicated by unequal rates of access to healthcare services , and inequalities in the experience and outcomes that different social groups obtain from inequalities a clear national priority. Black women in the UK are four times more likely to die in pregnancy or childbirth, 10 the healthy life expectancy gap between the most and least deprived communities of England is 19 years 4 and the disparity between average age at death in people with learning disability and the general population is 22 years for males and 26 years for females. 11 Given this context, our health inequalities work does not stop with the COVID-19 response. While GPs can't individually influence the social and economic factors that result in the health inequalities, there is plenty we can be doing in our own surgeries and practice areas. GPs in Scotland banded together in 2009 to form the 'Deep End' movement -covering the 100 practices that served the most deprived communities. Deep End GPs work together to advocate on behalf of their patients across a wide range of issues, using a swimming pool analogy as its title. Since then, the movement has grown across the UK and internationally, with 14 Deep End groups across seven countries. Developing that supportive network has proved valuable to individual practices who are otherwise struggling with their workload and complexity of patients they are seeing. And we don't have to do it all ourselves! As we explain below, there is much that can be done across professional boundaries and in integrated care systems. A perspective from Bola Owolabi, director of the National Healthcare Inequalities Improvement Programme Kamila's experience mirrors mine as a GP working in one of the ex-mining villages in Northeast Derbyshire. I see firsthand how socio-economic deprivation drives the incidence, Health inequalities in practice It defines a target population cohort -the 'Core20PLUS' -and identifies '5' focus clinical areas requiring accelerated improvement. There is an adult version as well as one for children and young people. --- Core20PLUS5 The approach has three components . Core20: The most deprived 20% of the national population as identified by the national Index of Multiple Deprivation . The IMD has seven domains with indicators accounting for a wide range of social determinants of health. PLUS: ICS-determined population groups experiencing poorer than average health access, experience, or outcomes, but not captured in the 'Core20' alone. This should be based on ICS population health data and will typically include ethnic minority communities, other groups that share protected characteristic groups, coastal communities, people with multiple health conditions, inclusion health groups such as people experiencing homelessness, drug and alcohol dependence, vulnerable migrants, Gypsy, Roma and Traveller communities, sex workers, people in contact with the justice system, victims of modern slavery and other socially excluded groups. --- 5: The final part sets out five clinical areas of focus, with smoking cessation as a cross cutting theme given the strong social gradient associated with smoking prevalence and the impact of smoking on the severity and impact of the 5 other clinical areas defined. These clinical areas and the specific interventions related to them are: > Maternity: ensuring continuity of care for 75% of women from BAME communities and from the most deprived groups. > Severe mental illness : ensuring annual health checks for people living with serious mental illness , bringing SMI in line with the success seen in learning disabilities. > Chronic respiratory disease: a clear focus on chronic obstructive pulmonary disease driving up uptake of covid, flu and those services. Action to address these inequalities falls squarely under the NHS's control. 14 This distinction is a powerful enabler for us as GPs. It liberates us from the paralysis that arises from a feeling of helplessness in effecting improvements to the wider determinants of health . A focus on healthcare inequalities means we can work within our sphere of influence -healthcare access, experience, and outcomes. We can lean into our role as clinical and professional leaders focusing on quality improvement as a tool with a real emphasis on data for improvement, strengths-based approaches and co-production with communities, patients, and service users. --- Addressing the challenge of health inequalities in general practice --- Core20PLUS5 The Core20PLUS5 approach was developed to help frame and streamline our actions in tackling healthcare inequalities. As an approach, it does not purport to be a silver-bullet solution to the enduring challenge of health inequalities. To use a mechanical analogy, I often describe Core20PLUS5 as a chassis: a structural framework which supports our efforts in taking specific and targeted action to narrow the healthcare inequalities gap. Core20PLUS5 offers each integrated care system in England a focused approach to enable prioritisation of energies and resources as they address health inequalities. We present this approach as the NHS contribution to a wider system effort by local authorities, communities and the voluntary, community and social enterprise sector to tackling healthcare inequalities -and it aims to complement and enhance existing work in this area. --- Kamila Hawthorne and Bola Owolabi Reimbursement Scheme can play a bridging role between general practice and those responsible for the wider determinants, such as housing liaison officers. In doing so, we will be expanding our sphere of influence beyond healthcare inequalities and reducing our sphere of concern in relation to the social determinants. --- Health inequality as a strategic priority for the Royal College of General Practitioners There is more that we can all do collectively. As chair of the RCGP, I [KH] have had the privilege of selecting four strategic priorities for my 3-year term of office. 15 With my working background, you won't be surprised to know that one of these is indeed health inequalities. We are a large membership body with over 54,000 members and there is much happening across the UK that we need to share so that we are not all re-inventing the wheel. I've designed our RCGP approach in three tiers: > What we can all be doing in our individual practices -we will collect examples of good practice and design a toolkit to help practices work out what they feel able to do in their own contexts. Practices vary hugely across the UK in terms of their list sizes, types of communities they serve, and urbanisation/ rurality. Each must find its own level to work in comfortably. All information will be posted on our Health Inequalities Hub online, free to access by anyone. > What we can be doing in groups of practices, working across teams in both health and social care, with third sector organisations, local councils and local community groups. Again, there are many examples of innovative and exciting work that is proving successful on a locality level and again, the RCGP will be showcasing as many as we can find. > What we should be doing at a national level, in our daily work with other national bodies , the other royal medical colleges, academics with special interest in health inequalities, and the NHS. This includes gathering best evidence and best practice and using the data to lobby government and political parties to influence their thinking and planning for the years ahead. This is especially important and urgent as we come up to a general election in 2024. As GPs, it is surprising how often you might find yourself in the local or even national spotlight, and able to put your lived experience of health inequalities and its human effects on people to a wide audience. So be prepared! Think about what you might want to say, as your patients' advocates, if you get the chance. If you have experience of working on health inequalities in your communities, be prepared to speak up about the work, to reach the widest possible audience. Together we are much stronger than when we are working alone. ■ pneumonia vaccines to reduce infective exacerbations and emergency hospital admissions due to those exacerbations. > Early cancer diagnosis: driving towards 75% of cases diagnosed at stage 1 or 2 by 2028. > Hypertension case-finding, optimal management and lipid optimal management: to minimise the risk of myocardial infarction and stroke Core20PLUS5 for children and young people The approach for children and young people has a similar structure, again with three components . Core20: the most deprived 20% of the national population as identified by the national Index of multiple deprivation . The IMD has seven domains with indicators accounting for a wide range of social determinants of health. PLUS: the PLUS population groups includes children and young people from ethnic minority communities, coastal communities, inclusion health groups and protected characteristic groups and children and young people with learning disabilities, autism and multiple health conditions, amongst others. Specific consideration should be taken for the inclusion of young carers, looked after children/care leavers and those in contact with the justice system. 5: the final part sets out five clinical areas of focus. The five areas of focus are part of wider actions for ISCs and integrated care partnerships to achieve system change and improve care for children and young people. Governance for these five focus areas sits with national programmes; national and regional teams coordinate local systems to achieve aims. > Asthma: address over-reliance on reliever medications and decrease the number of asthma attacks. > Diabetes: increase access to real-time continuous glucose monitors and insulin pumps across the most deprived quintiles and children and young people from ethnic minority backgrounds; and increase proportion of those with Type 2 diabetes receiving recommended NICE care processes. > Epilepsy: increase access to epilepsy specialist nurses and ensure access in the first year of care for those with a learning disability or autism. > Oral health: reduce tooth extractions due to decay for children admitted as inpatients in hospital aged 10 years and under. > Mental health: improve access rates to children and young people's mental health services for 0-17-year-olds, for certain groups related to ethnicity, age, gender and deprivation. The hope is that, by differentially focusing our efforts and energies on those experiencing the most disadvantage and those clinical areas making the most contribution to both life expectancy and healthy life expectancy gaps between the rich and poor and across other domains of disadvantage, we can gain some traction and demonstrate measurable impact within a reasonable timeframe. Our communities are justifiably impatient to see this happen. Beyond the scope of healthcare inequalities, we can be influential members within the ICS and make meaningful contributions to materially improve the lot of our patients and communities through closer links with colleagues in local authorities, the VCSE sector, and faith and other community outreach organisations. This is where our social prescriber link workers and other roles coming under the Additional Roles

Kamila Hawthorne, chair of the Royal College of General Practioners Council, and Bola Owolabi, director of the National Healthcare Inequalities Improvement Programme, offer a personal perspective on how their professional experiences have led them to see health inequality as urgent priority, and introduce initiatives that can help general practitioners make a difference individually and collectively.

Introduction Adolescence is the period of age ranging from 10 to 19 years, is one of the critical transitions in the lifespan that occurs after childhood and before adulthood, and is characterized by a tremendous pace in growth and change that is second only to that of infancy [1]. According to census 2011, age group from 10 to 19 years constituted 24.19 percent of the total population where 3207821 were male and 3199583 were female [2]. Psychosocial problems, such as behavioural, emotional, and educational problems are highly prevalent among children and adolescents [3]. Adolescents are vulnerable to psychosocial dysfunction when they suffer from physical injuries, psychological trauma, or major changes in their environments especially in the absent of strong support system [4]. Adolescence period is critical times for developing good mental health [5,6]. Mentally healthy adolescents enjoy a positive quality of life; are free of symptoms of psychopathology; and function well at home, in school, and in their communities [7]. Lifetime psychiatric disorders usually have their first onset at a young age: half of them by 14 years and threequarters by 24 years [5,6]. Psychosocial problems have emerged as a threat in their overall development of adolescents [8]. A study in Dehradun showed that the overall prevalence of psychosocial problems was 40.5% [9]. Similarly, another cross-sectional study in Dehradun revealed that the overall prevalence of psychosocial problems among the adolescents was found to be 31.2% [10]. Similarly, a study in Nepal also revealed that 17.03% of adolescents were suffering from psychosocial dysfunction [11]. Adolescents have very special and distinct needs, which can no longer be overlooked. It is essential to invest in adolescent, as they are the future of the country. Therefore, this study was initiated an attempt for 2 Psychiatry Journal identifying the psychosocial problems among school going adolescents in Nepal. The results of this study are expected to contribute to design preventive programs, primarily focusing on psychological intervention for improving mental health of the school adolescents. --- Research Methodology --- Study Design. In December 2016, cross-sectional descriptive study was carried out for identifying the psychosocial problems of school going adolescents in Nepal. --- Study Setting and Population. Study settings were three schools, namely, Nabingram Secondary School, Chapabot Higher Secondary School, and Shree Kalika Saran Higher Secondary School. These schools are located at Jarsingh Pauwa, ward no. 5 of Shankarapur municipality, Kathmandu, Nepal. The study population was all students from class 8 th , 9 th , and 10 th including boys and girls of that selected schools. --- Sample Size and Sampling. Based on the students' presence at the time of data collection, 292 students were approached. Excluding the questionnaire having missing data, a total of 287 responses, i.e., 71.3% of target population, were used for the final analysis. Convenient sampling technique was adopted for the selection of those samples. --- Data Collection Tools. The instrument used in this study was composed of two parts. Part one was related to questionnaire regarding sociodemographic information. Name of school, age, grade, religion, ethnicity, types of family, parent's educational status, occupational status, and marital status were included in this part. Part two was a tool of the youth report of Pediatric Symptom Checklist [12] that was used to screen the psychosocial problem among adolescents. It was completed by the adolescent aged 11 and up. It consisted of 35 items that are rated as "Never," "Sometimes," or "Often" present and scored 0, 1, and 2, respectively. The total scores were calculated by adding together the score for each of the 35 items. Items which were left blank were simply ignored . If four or more items left blank, the questionnaire was considered invalid. The cutoff score for the Y-PSC was 30 or higher. For adolescents of all ages, a score of 30 or higher indicated significant psychological impairment. A positive score on the Y-PSC suggested the need for further evaluation by a qualified health or mental health professional [13]. For ensuring content validity, these instruments were further evaluated by consulting subject expert, advisor, experienced professionals, academics in child health and development, and psychologist to determine whether the instrument reflects the known content area. Reverse translation of the instruments was done. Pretesting was done among 25 students of Sarada secondary school by using Nepalese version of the questionnaire. It was done in order to establish the conceptual/linguistic and functional equivalence before the administration of the instruments to the actual sample of the study. Those students were similar in characteristics with actual samples of this study. Reliability coefficient of the Y-PSC was 0.808. --- Data Collection Procedure. The duration of data collection was 2 weeks. It started from 18 th December 2016 to 29 th December 2016. Firstly, appointment for data collection was taken from the principal and class teacher. By considering ethical procedure, data collection was done from students in their respective classroom. Before distributing the questionnaire, orientation about the tools was given to the students and they were informed about the importance of responding to each statement of the questionnaire very carefully. Then, structured self-administered questionnaires regarding sociodemographic information and Youth Pediatric Symptoms Checklist were distributed to each respondent separately. Respondents took 20 minutes time in average for the completion of the questionnaire. Field editing and central editing were done for identifying its completeness, correctness, and accuracy at that time. 2.6. Data Analysis Procedure. The obtained data were edited, classified, and the different variables of the questionnaire were coded and double checked. Double data entry as well as data cleaning was done using Epi Data software. Then, the data were transferred to statistical package for social sciences software version 16. Descriptive statistics was used to describe the sample characteristics. For analyzing the association between sociodemographic characteristics and psychosocial problems among adolescents, chi-square test was used. For this test, significance was considered at p ≤ .05 for 95% confidence interval. --- Results Completed responses were received from 292 respondents. However, only 287 responses were used for the final analysis because of the missing data. This study shows that 50.9 percent of respondents were in age group of middle adolescents , 36.6 percent of respondents were studying in class nine, 49.5 percent of respondents were Hindu, and 53.3 percent of respondents were Tamang. Regarding types of family and parent's education, 52.3 percent of respondents were living in nuclear family, and majority of respondent's father and more than half of respondent's mother were literate. Related to parent's occupation, majority of respondent's father were farmer and nearly half of respondent's mother were involving in household work. Most of respondent's parents were married. Table 1 displays that majority of respondents did not have significant psychosocial impairments and only 12.9 percent of respondents had significant psychosocial impairment. Table 2 displays that regarding internalizing problems, 44.6 percent of respondents had significant impairment in internalizing problems. Related to attention deficit hyperactive disorder , 25.8 percent of respondents had Note. Score of 30 or higher = significant psychological impairment . significant impairment in ADHD. Regarding externalizing problems, only 4.2 percent of respondents had significant problems with conduct. Table 3 shows that there is association between age group and psychosocial problems and there is no association of respondents' school grade, religion, and ethnicity with psychosocial problems. Table 4 reveals that there is association between respondents' parent's marital status and psychosocial problems . However, there is no association of respondents' types of family, father's educational status, mother's educational status, father's occupational status, and mother's occupational status with psychosocial problems. --- Discussion The present study showed that 12.9 percent of respondents had psychosocial problems. A previous cross-sectional study in Hetauda, Nepal, showed that 134 were suffering from psychosocial problem [11]. Another descriptive comparative study in Bharatpur also revealed that school children of nonworking mothers 11.7% had slightly more psychosocial problems than working mothers 8.3% [15]. A similar finding revealed a cross-sectional study in Pune which showed that 328 children were found to be at risk of psychosocial problems [16]. Findings are shown by a cross-sectional study in Muzaffarnagar, India, which revealed that the overall prevalence of psychosocial problems among adolescent was found to be 41.43% [8]. Another study in Dehradun, India, revealed the psychosocial problems among children [9]. Yet another study findings revealed by a cross-sectional study conducted in Dehradun which showed that the overall prevalence of psychosocial problems among the adolescents was found to be 31.2%, among them [10]. Differences in prevalence may be associated with respondent's contextual differences in national and international settings. Therefore, proper action should be taken for addressing these issues. The current study showed the internalizing problems , ADHD , and externalizing problems among respondents. Findings of previous descriptive comparative study in Bharatpur showed that internalizing problems were found equal in children of working and nonworking mothers whereas attention problems were slightly higher in the school children of working mothers than school children of nonworking mothers . Externalizing problems were found slightly higher in the respondents of nonworking mothers than children of working mothers [15]. Findings of another study in urban and rural areas of Dehradun showed that anxiety and conduct disorders were more common among adolescents in rural area while depression was more common among adolescents in urban area [9]. Respondent's contextual differences in national and international settings might influence for this dissimilarity in findings. The present study yielded that there is association of age group and parent's marital status with psychosocial problems among respondents. Similar findings which were revealed by a crosssectional study in Hetauda, Nepal, revealed that age was significantly related to psychosocial problem. Among family factors, frequency of family dispute was highly associated with psychosocial problem [11]. A cross-sectional study in Pune also showed a similar finding that statistically significant difference was observed as per age group [16]. However, the present study revealed that there is no association of school grade, religion, ethnicity, types of family of respondents, father's educational status, mother's educational status, father's occupational status, and mother's occupational status, with psychosocial problems among respondents. Contradictory finding was revealed by a cross-sectional study in Pune which showed that statistically significant difference was observed as per class of student and similar finding was also revealed by this study that no significant difference was observed as per the type of family [16]. The considerably such prevalence of psychosocial problems among adolescents suggests the vulnerability of school going adolescents. Therefore, these findings indicate a need for national survey and launch awareness programme for preventing psychosocial problems. --- Conclusions The prevalence of psychosocial problems is evident among school going adolescent. Adolescents' age group and parent's marital status are associated with psychosocial problems. Therefore, health care policy maker and school authority Psychiatry Journal should create awareness program on psychosocial problems among adolescents, develop strategies for health promotion of adolescents, and plan for prevention of psychosocial problems among adolescents. --- Limitations of This Study This study adopted cross-sectional descriptive design covering three schools of Jarsingh Pauwa of Shankarapur Municipality, Kathmandu; therefore, the findings of this study may not be generalizable to all the school going adolescents of Nepal. Adolescents were asked about their problems that may be best fitted to them in the past days. Therefore, there may be a chance of recall bias. As this was a crosssectional study, no causal relationship could be inferred. This study did not analyze the association between gender and psychosocial problems. Therefore, this study recommends for the replication of this study on a larger sample, covering a wider geographical area and gender-based analysis for better generalization to all school adolescents. --- Data Availability The datasets used and/or analyzed during the current study will be available from the corresponding author on reasonable request. Individual responses will not be shared. --- Ethical Approval Administrative approvals from Padma Kanya Multiple College and school authority of the selected schools were taken before collecting data. This study was approved by the Ethical Review Board of Nepal Health Research Council, Kathmandu, Nepal . --- Consent The respondent's rights were protected by taking verbal informed consent from the parents by using informed consent form and informed assent was taken from the respondents by using written informed assent form before data collection and keeping the collected information confidential. Anonymity was assured by requesting respondents not to write their names or any identifiable marks on the questionnaires. For, confidentiality, all the information that was provided by the respondents were treated with the strictest confidentiality while the study was conducted and when the study is being reported or published and only the aggregate data were analyzed. The privacy was maintained by keeping the collected information confidential and respondents were advised to respond the questionnaire individually by covering the answered question by using paper. --- Disclosure Mina Timalsina is a first author and Mana Kafle and Rekha Timalsina are coauthors. --- Conflicts of Interest The authors declare that they have no conflicts of interest. --- Authors' Contributions Mina Timalsina conceived the research and collected and analyzed data. Mina Timalsina and Rekha Timalsina helped to analyze the data. Mana Kafle supervised research process and manuscript writing. Rekha Timalsina wrote the manuscript and all the authors read and approved the final manuscript. --- Stem Cells International

Background. Psychosocial problems refer to the difficulties faced by adolescents in different areas of personal and social functioning. Adolescents are vulnerable to psychosocial problems because of physical and physiological changes that occur in their body during this developmental stage. Therefore, this study was conducted to identify psychosocial problems among school going adolescents in Nepal. Methods. A cross-sectional descriptive study was adopted. Nonprobability convenient sampling technique was used for selecting 287 adolescents. Ethical approval was taken from Nepal Health Research Council and self-administered structured questionnaire was used for data collection. Data collection was done in 2016. Descriptive statistics and chi-square test were used to analyze the data. Results. The findings of this study show that 12.9 percent of adolescents had psychosocial problems. While categorizing psychosocial problems, the adolescents had internalizing problems (44.6%), attention deficit hyperactive disorder (ADHD) (25.8%), and externalizing problems (4.2%). There is association of age group and parent's marital status with psychosocial problems. Conclusion. It is concluded that psychosocial problems (i.e., internalizing problems, ADHD, and externalizing problems) were prevalent among Nepalese school adolescents. Adolescents' age group and parent's marital status are associated with psychosocial problems. This study recommended that school authority, health professionals, and other professional related to child health and mental health should play an important role for the prevention and earlier recognition of and intervention for psychosocial problems.

Introduction Africa has the highest HIV prevalence rate globally, averaging nearly 4% of their population [1]. The World Health Organization recommended male circumcision and HIV testing to become part of a comprehensive approach to HIV prevention [2]. Circumcision is a longstanding tradition involving the surgical removal of the foreskin from the penis [3][4][5]. The suggestion stemmed from findings in three randomized controlled trials carried out in Kenya, South Africa, and Uganda during the early to mid-2000s. These trials demonstrated a 50-60% decrease in the risk of female-to-male transmission of HIV among circumcised men when compared to their non-circumcised counterparts [6][7][8]. Additional evidence indicates that circumcision is associated with a decrease in HIV infection in some eastern and southern African countries where the public health policies have integrated voluntary male medical circumcision and HIV testing [9][10][11]. Studies suggests that 75 to 85% of HIV infection are likely acquired through sexual intercourse and majority of men living with HIV might have been infected through the penis [12]. Therefore, circumcision could eliminate the inner layer of foreskin enriched with HIV-1 target cells and receptors for the HIV virus [12][13][14][15]. This removal may confer protective effects against HIV infections. Both epidemiological and biological studies offer compelling evidence supporting the protective benefits of male circumcision in reducing heterosexual HIV transmission among males. These studies also help to explain the notable geographical variations in the prevalence of HIV infection within sub-Saharan Africa. The increasing focus on circumcision practices is also driven by empirical evidence suggesting a lower incidence of certain sexually transmitted infections and HIV among circumcised individuals [14,16,17]. Some research even suggests that circumcision could reduce the HIV infection rate by up to 70%, marking a significant contribution to public health [18,19]. Overall, fifteen WHO designated priority countries in Sub-Saharan Africa have implemented the integration of Voluntary male medical circumcision with HIV prevention practices such as HIV testing as a comprehensive HIV prevention strategy with measurable implications on HIV prevalence rates [20]. Experts warn that male circumcision should not replace HIV prevention practices [21]. Nigeria is the most populous country in Africa and the seventh most populous nation in the world. Most recent estimates indicate that more than three million Nigerians live with HIV/AIDS out of an estimated total current population of 223, 804, 632 million people [1,22]. The HIV burden in Nigeria ranks fourth worldwide but highest in West and Central Africa with alternative conservative estimates suggesting that about 1.8 million people are living with HIV in Nigeria [23,24]. In 2021, the HIV prevalence in Nigeria reached 1.9 million people averaging 74,000 new HIV infections at a 1.3% adult HIV prevalence rate, with 51,000 deaths attributed to AIDS, encompassing both adults and children [23,25]. Nigeria unenviably ranks the lowest in life expectancy among countries in the West African sub-region, with HIV mortality significantly contributing to this problem. Currently, Nigeria has to its credit, a unifying national HIV and AIDS strategic framework 2021 -2025 which plans to increase access to HIV testing services so as to enable 90% of people living with HIV know their status by 2025 [23]. However, Nigeria has not integrated male circumcision and HIV prevention practices as part of its comprehensive health services despite compelling evidence shown in doing so by other countries in the control of the HIV/AIDS pandemic. Apparently, Nigerian health authorities reckon that circumcision rates are already high, and that subsisting national HIV/AIDS framework may be sufficient. The challenge lies in accurately assessing the implementation of this HIV/AIDS framework with a view to quantifying the impact of the thematic areas of focus on reduction of new HIV infection. Granted, the 2018 Nigeria Demographic and Health Survey indicate that male circumcision rate is indeed very high but other reports indicate that new HIV infection rates remain significantly high coupled with the high HIV burden [23]. Given that experts warn that male circumcision should not replace HIV prevention practices [21], and male circumcision rates are already very high, there is crucial need to explore the differentials in the knowledge of AIDS, other sexually transmitted infectious , contraception methods and their association with selected HIV prevention behaviors among circumcised males in Nigeria. Understanding the health seeking behaviors of circumcised men will provide valuable insight on alternative thematic areas for focusing health education and intervention resources. --- Methods --- Data Source The 2018 Nigeria National Demographic and Health Survey is a comprehensive, nationally representative sample survey designed that offers current insights into demographic and health trends [26]. The NDHS was conducted by the National Population Commission , and the Department of Health and Human Services Inner City Fund International with support from United States Agency for International Development , Global Fund, Bill and Melinda Gates Foundation, World Health Organization , and the United Nations Population Fund . Approximately 42,000 households were included in the representative sample. All men aged 15 to 59 residing permanently in the chosen households, as well as temporary visitors who spent the night before the survey in those households, were eligible for interview participation. Data collection for the NDHS took place from August 14 to December 29, 2018. The survey encompassed various topics, including sociodemographic characteristics, awareness and attitudes towards HIV/AIDS and HIV testing, and other health-related issues. The study utilized a male questionnaire to gather information from eligible male participants aged 15 to 59, focusing on modules covering HIV, condom use, aspects of male circumcision, and demographic characteristics. --- Sampling Design and Sample Size This study is based on a secondary dataset known as the 2018 NDHS which marked a significant milestone as it introduced computerassisted personal interviewing for the first time [26]. This innovative approach facilitated a faster and more efficient provision of data compared to previous survey methodologies. Employing a stratified, twostage cluster design, the survey utilized enumeration areas as the first-stage sampling units. In the second stage, a complete listing of households was conducted in each of the 1,400 selected EAs. A sample size of 13,311 male participants were interviewed for this study. In addition to presenting national estimates, the report provides key indicators for both rural and urban areas, Nigeria's six geopolitical zones, all 36 states, and the Federal Capital Territory . This extensive survey serves as a valuable resource for understanding and addressing demographic and health issues across the country. Measures were solely derived from the Man's questionnaire of the NDHS which was administered to all men aged 15-59 in the subsample of households selected for the men's survey. --- Measures Male circumcision is measured by asking participants. "Some men are circumcised, that is, the foreskin is completely removed from the penis. Are you circumcised?" This variable was measured dichotomously as "yes" and "no" . Knowledge of HIV/AIDS was assessed by asking the participants: "Now I would like to talk about something else. Have you ever heard of HIV or AIDS? This variable was measured dichotomously as "yes" and "no". Knowledge of any contraceptive method was measured by asking the participants if they had knowledge of any contraceptive method. This was dichotomously coded as "yes" or "no". STIs knowledge was measured by asking participants: "apart from HIV, have you heard about other infections that can be transmitted through sexual contact?" Also, this variable was measured dichotomously as "yes" and "no" . HIV prevention behaviors was assessed by asking the following items used condom every time they had sex with the most recent partner in the last 12 months, "was a condom used every time you had sexual intercourse with this person in the last 12 months? sought advise/treatment for last sexual infection, "The last time you had a sexually transmitted infection, did you seek any kind of advice or treatment?" This was dichotomously coded as "yes" or "no". The socio-demographic characteristics included the following variables: region, ethnicity, resident type, age group, marital status, education, employment, wealth status, and religion. These were used as control variables. --- Data Analysis The statistical package for the social sciences were utilized for data analysis. For the descriptive statistics, the frequencies and proportions of sociodemographic variables were computed and tabulated. Chi-square was used to determine the association and prevalence estimates between male circumcision and knowledge of HIV/AIDS, other STIs, contraception methods, and HIV prevention behaviors. To determine whether knowledge of HIV/AIDS, other STIs, and contraception methods predict the sexual health seeking behaviors of Nigerian men aged 15 to 59 years, we conducted multivariate logistic regression while controlling for sociodemographic characteristics. Afterwards the new knowledge of contraception methods variable, knowledge of HIV/AIDS, and STIs were used to fit a logistic regression to predict the odds of practicing either consistent use of condom or seeking for advice/treatment . The statistical significance was set at the p-value of 0.05. --- Ethical Considerations The --- Results --- Descriptive Characteristics of the Participants This table provides descriptive characteristics of the participants. The total sample size included 13,311 Nigerian men aged 15 to 59 years. By region, the participants were distributed across different regions of the country, with the highest representation in the Northwest , followed by the Northeast and North Central . In terms of ethnicity, most participants belong to the "Others" category which represents minority tribes, followed by the Hausa/Fulani ethnic group , Igbo , and Yoruba . More than half of the participants lived in rural areas, were between 15 to 34 years old and were married . About 63.5% of the participants had at least secondary or higher education, and only 36.2% were currently working, and belonged to the "rich" by wealth status . Slightly more than half of the participants identified as belonging to as Islam , and the overwhelming majority of the participants were circumcised . The Chi-square results showing an association and prevalence of male circumcision by HIV knowledge, condom use, and seeking advice/treatment for last STI infections are shown in Table 2. The results indicated that knowledge of HIV/AIDs, including ever heard of HIV/AIDS , ever heard of other STIs , as well as seeking advice/treatment for the last STI infection were significantly associated with male circumcision. However, consistent condom use in the last 12 months was not significantly associated with male circumcision . Particularly, 94.8% of the men who reported having ever heard of HIV/AIDS were more likely to be circumcised compared to 5.2% who never heard of HIV/AIDS. Second, 74% of men who had ever heard of STIs were more likely to be circumcised than 26% who had never heard of other STIs. In terms of seeking advice/treatment for the last STI infection, the prevalence of having sought advice/treatment for the last STI infection was only 2.5%, with only minimal difference for male circumcision compared to 2.4% who never sought advice/treatment for the last STI infection. Finally, although consistent condom use was not significantly associated with male circumcision, surprisingly, 70.7% of the men who used condoms during sexual intercourse in the last 12 months were more likely to be circumcised compared to 29.3% who never used condoms during the last 12 months. The logistic regression results of HIV/STI knowledge and method of contraceptive predicting consistent condom use in the last 12 months and seeking advice/treatment for STI infection are shown in Table 3. Men who ever heard of HIV/AIDS had higher odds of consistent condom use with their most recent partner in the last 12 months compared to those who never heard of HIV/AIDS . However, having heard of other STIs was not significantly associated with consistent condom use every time among men had sex with the most recent partner in the last 12 months compared to those never heard of other STIs . On the other, having heard of other STIs was significantly associated with higher odds of seeking advice/treatment for the last STI infection among men . However, having ever heard of HIV/AIDS and knowledge of contraceptive methods were not significantly associated with seeking advice/treatment for last STI infection among men in Nigeria. --- Discussion Our aim in this study was to determine the prevalence and examine the differences between HIV/AIDS knowledge, other STIs, contraception methods, and the potential odds of such knowledge impacting the practice of selected HIV prevention behaviors amongst circumcised men in Nigeria. Hence, this study has made a significant contribution to the literature by providing critical insights about high rates of male circumcision that appear not to result in a commensurate reduction in the HIV prevalence rate and HIV/AIDS burden among Nigerian men. In particular, in our prevalence findings, we found that among 96.8% of men who were circumcised, 94.8% reported having knowledge of HIV/AIDS, 74% knowing other STIs, 94.8% having knowledge of contraceptives, and 70% reported having used condoms in the past 12 months in Nigeria. At the same time, the significant association we found between reported knowledge of HIV, and consistent condom use with male circumcision is consistent with findings from previous studies indicating a high knowledge of medical circumcision for HIV prevention among male samples in other countries [27][28][29]. This suggests that knowledge of HIV/AIDS may serve as both a protective and prevention mechanism with numerous benefits of reducing the potential HIV or infection from other STIs among men in Nigeria. Thus, the continuation of intervention or programs in some parts of the country lagging HIV awareness is paramount to successfully control HIV/AIDS infection rates in Nigerian society. In the present study, we found a highly significant association between other STIs and male circumcision status. This aligns with the previous study, which indicated that men who were circumcised, regardless of their HIV status, exhibited a lower prevalence of syphilis compared to their uncircumcised counterparts in Nigeria [30]. Additionally, circumcised men demonstrated a reduced prevalence of curable STIs among their female partners [30]. In contrast, one study conducted in Uganda posits that there was no association between male circumcision and other bacterial STIs with the exception of syphilis and chancre [31]. This study also found that male circumcision was associated with variations in awareness of HIV/AIDS, other STIs, and contraceptive methods, as well as seeking advice/treatment for the last STI infection but not for consistent use of condom for every sexual intercourse in the past 12 months. Similar studies conducted in eastern, southern, and central Africa support this finding [29,32,34]. Furthermore, knowledge of AIDS was significantly associated with higher odds of consistent condom use among men in Nigerian. Specifically, those having knowledge of AIDS were 2.4 times more likely to use condoms relative to those who were not aware of AIDS in Nigeria. Conversely, rather than knowledge of AIDS, some prior studies found that perception of condom use by friends, self-efficacy , and general impulsive attitudes were predictors of condom use [35][36][37]. Another study suggests that having more social network members was a significant predictor of consistent condom use [38]. A more recent study found that condom use were reportedly high during anal sex than in vaginal sex [39], while another study identified the predictor for condom use as initiating condom use during one's first sexual experience and engaging in a conversation with one's initial sexual partner regarding the use of condoms [40]. These studies did not specifically examine the knowledge of AIDS in relation to condom use in their sample. On the other hand, evidence from the current study suggests that knowing other STIs was associated with higher odds of seeking advice/treatment. Precisely, Nigerian men who have heard of other STIs were three times more likely to seek treatment or advice if they had a sexually transmitted infection. Although other studies have suggested that age and number of sexual partners are correlates of HIV and STI status in similar sub-Saharan African demographics [35], this study presents unique differentials in knowledge about other STIs examined through HIV prevention evidence. It is also possible that having knowledge about STIs could serve as a reminder about the potential consequences that are associated with undiagnosed STIs. As such, men who engage in unprotected sexual risk behaviors may be compelled to seek treatment or advice regarding STIs. In this case, knowledge of STIs is a potential protective factor and can prompt immediate response to seek support, treatment, and information vital to living healthy and productive lives among men at risk of STIs. --- Implications of the Study The findings from this current study present a unique perspective, but like other related studies [35], proposes the crucial need to escalate HIV/AIDS preventive measures in favor of those who are behaviorally vulnerable to HIV/AIDS such as adolescents, young adults and the uneducated who may not have been exposed to family life and HIV education. Sex education have been the thrust of prior studies with similar demographics, although it is not often emphasized in the Nigerian society due to sociocultural, educational, and religious inhibitions [41][42][43]. In this sample, 22.1% have no education at all. In comparison, 14.4% have only primary education, which indicates that 36.5% of Nigerian men may likely be uneducated about sexual health beyond knowledge of the existence of HIV/AIDS. Also, young men aged 15 to 34 years who may be most behaviorally vulnerable to HIV/AIDS account for the highest group with the least possibility of sexual health education, assuming they fall within the 36.5% of Nigerian men with little or no education. Given the possibility of the lack of family life and HIV education in the younger demographic prior to potential exposure to sexual risks, more concerted efforts at sexuality education should be prioritized early from 10 years old. --- Limitations of the Study The limitations of the current study include the lack of HIV testing data which may also contribute to understanding the relationships between male circumcision and HIV prevention behaviors. Also, the study design suggests that the current study cannot establish or infer a causal relationship between male circumcision and HIV prevention behaviors or knowledge of HIV/AIDS, contraception, and STIs. Considering that HIV is still a disease rooted in stigma, social and economic inequities, the selfreport survey design may be prone to response bias. Participants in the study may have potentially underreported or experienced recall bias, influencing the outcomes. Despite these limitations, our models incorporated variables chosen for their relevance in extensive existing literature, particularly in studies related to male circumcision and HIV. Notably, the study employed a nationally representative sample, enhancing reliability and generalizability to other settings in Nigeria and across sub-Saharan African countries. --- Conclusions This study provides valuable insights on the differentials in HIV/AIDS knowledge and behaviors and male circumcision with the intent to contribute to the HIV prevention literature. Evidence from this study suggests that knowledge of HIV/AIDS is a significant predictor of consistent condom use during sexual intercourse. Similarly, knowledge of STIs is also a significant predictor of seeking advice or treatment following sexual infection. Besides, male circumcision status has been associated with differences in awareness of HIV/AIDS, other STIs, and contraceptive methods, as well as seeking advice/treatment for the last STI infection. These findings provide thematic areas of focus for the national response towards ending HIV/AIDS in Nigeria. Considering that male circumcision status does not account for reduction in HIV burden and prevalence among Nigerian men, concerted efforts should focus on alternative approaches such as expanding access to HIV prevention, treatment, and linkage to care services at the community level through relevant formal and informal community structures. Current efforts addressing HIV prevention in Nigeria should integrate sexuality education into all existing community health services, focusing on building the knowledge and capacity of community stakeholders and service providers. Moreso, evaluating the progress of the national HIV/AIDS strategy should be carried out quarterly in order to sustain progress in ending the HIV epidemic. Finally, this study provides direction for further studies that could utilize ethnographic qualitative approaches to integrate community engaged HIV prevention strategies such as HIV testing and linkage to care services into research. --- Conflict of Interests The author declare that they have no competing interests.

Compelling evidence regarding the HIV/AIDS reduction potentials of male circumcision abounds in sub-Saharan Africa, but men aged 15 to 59 years in Nigeria are mostly circumcised already. Yet, Nigeria continues to experience tens of thousands of deaths due to HIV/AIDS yearly, accounting for one of the highest HIV burdens worldwide. Differentials that may account for new HIV infections and high HIV prevalence rates need to be investigated. The purpose of this study was to examine the prevalence of male circumcision and determine the association between HIV/AIDS/STIs knowledge, contraception methods, and HIV prevention behavioral practices amongst men in Nigeria. This study utilized the 2018 Nigeria Demographic and Health Survey (n= 13,311). Chi-square tests were used to determine the prevalence of male circumcision. Two binary logistic regression models were performed to predict the relationship between independent and dependent variables. The prevalence of male circumcision was higher among men with knowledge of HIV/AIDS (98%) and other STIs (74%), used condoms (70.7%), and only 2.5% of men who sought advice/treatment were circumcised. Further evidence suggested that knowledge of HIV/AIDS was associated with higher odds of consistent condom use (AOR = 2.37; 95% CI =1.12 -5.02), and knowledge of other STIs was significantly associated with higher odds of seeking advice/treatment for the last STIs (AOR= 3.06;. This suggests that comprehensive education regarding HIV/AIDS and STIs if prioritized may serve as both a protective and preventive mechanism to achieve HIV epidemic control among men in Nigeria.

A. INTRODUCTION ndonesia is a unique country in the world. Considering that Indonesia has a large number of islands, and has high biodiversity and cultural diversity . The Indonesian nation is a pluralistic nation because its society consists of a collection of people or groups with ethnic characteristics who have various cultures with different ethnic backgrounds. Cultural diversity Indonesia has more than 1,128 ethnic groups living in areas spread across thousands of islands stretching from Sabang to Merauke . Indonesia has around 300 ethnic groups , each ethnic group has a cultural heritage that has developed over the centuries . Every Indonesian citizen has the right to uphold culture. Because basically National culture refers to the superior values of local cultures which are the cultural heritage of the Indonesian nation . This is confirmed in Article 32 paragraph 1 of the 1945 Constitution that, "The state advances Indonesian national culture in the midst of world I civilization by guaranteeing the freedom of the people in maintaining and developing their cultural values" According to Ki Hajar Dewantara, culture is the result of people's struggle against nature and times which proves the prosperity and triumph of people's lives in responding to or facing difficulties and obstacles to achieve prosperity, safety and happiness in their lives . Culture is a habit or tradition that develops in an area which is passed down from generation to generation . Culture is everything related to creativity, taste, and the results of people's work. Culture can be distinguished based on its form, namely object culture and non-object culture . Material culture refers to all the creations of society that are real, concrete, produced and used by society. While non-object culture abstract creations passed down from generation to generation as guidelines or references for community groups, followed with full awareness in the form of values, norms, morals and beliefs . Based on the Law on the Advancement of Culture Number 5 of 2017 there are 10 objects for the promotion of culture including oral traditions, manuscripts, customs, rites, traditional knowledge, traditional technology, art, language, folk games and traditional sports. One of the 10 Objects for the Advancement of Culture that needs to be preserved and maintained is oral tradition, because this oral tradition has a strategic role in traditional events and for tourism . Preserving culture is closely related to what this nation's independence has aspired to, namely the aspiration to "educate the nation's life", educating the nation's life is not a meaning based on the concept of "Science and Technology" or the "Genetic Biological Concept", but a conception culture . Preservation of culture is a large system, having various components related to the subsystems of life in society. Culture is the forerunner of society. Culture is made by society, there is no society without culture, which means that almost all human actions are culture . In Regulation of the Minister of Culture Number 10 of 2014 it is explained that the preservation of culture as the preservation of traditions is an effort to protect, develop, and utilize a custom from a group of people supporting culture whose transmission and inheritance take place from generation to generation. Based on Law Number 32 of 2004 which contains the principle of decentralization, it allows local governments to formulate local regulations regarding the implementation of cultural preservation in an area. Local governments must have a commitment to protect and preserve various forms of culture in their area . West Sumatra is one of 34 provinces in Indonesia which has challenges of cultural diversity. This is marked by the diversity of social identities, the variety of ethnic groups living in this area and the lifestyle of a globally oriented society . The Regional Government of West Sumatra Province is one of the regions in Indonesia that formulates policies related to culture and customs as a regional symbol and characteristic which are then legally stipulated to become a Legislative Regulation, namely West Sumatra Province Regional Regulation Number 7 of 2018 concerning Nagari. The biggest culture in West Sumatra is known as the Minangkabau Culture. Minangkabau is known as one of the subcultures or cultures in Indonesia . The Minangkabau tribe is one of the largest ethnic groups in Indonesia . Minangkabau is an ethnic group in Indonesia, which inhabits the Central Sumatra region, namely West Sumatra Province, Kerinci Jambi Province, Kampar Bangkinang Riau Province, known as the Minangkabau tribe . Law Number 17 of 2022 explains that Minangkabau Culture is based on philosophical values, adat basandi syara', syara' basandi Kitabullah in accordance with the applicable customary rules of Salangka Nagari, as well as a wealth of history, language, art, traditional village/nagari, traditional ceremonial rituals , cultural sites, and local wisdom that shows the religious character and height of the customs of the people of West Sumatra. West Sumatra is a place known as the origin of Minangkabau Culture. West Sumatra is a region that has a variety of cultures, traditions, ethnicities, tribes and languages. This is certainly a valuable asset for the local community, regional government, and also the central government. West Sumatra Province has culture and traditions which are still maintained and carried out today, such as the Tabuik Festival in Pariaman City, Pacu Jawi, Tanah Datar Regency, Ronggiang West Pasaman Regency, Batombe, South Solok Regency, and others. These cultures and traditions are in the form of religious rituals, language arts, dance arts, and even as sporting activities. Dharmasraya Regency has a multiethnic population including Minangkabau, Javanese, Sundanese and Batak ethnicities. And Dharmasraya Regency is an area in Minangkabau which still upholds culture. As it should be, local governments have a commitment to protect and preserve various forms of culture in their area . So based on this statement, the Regional Head of Dharmasraya made "culture" his vision and mission. Based on Government Regulation Number 38 of 2007, cultural affairs are one of the obligatory matters handed over by the central government to regional governments. The implication of this government regulation is to provide an opportunity for local governments to make strategic choices related to cultural policies in the areas that are their authority . The Regional Head of Dharmasraya makes "Culture" his Vision and Mission. Vision and Mission which will become Government Policies, Programs, Activities and Sub-Activities for 5 years . Regional Policy or also known as "Public Policy" is defined by Thomas R. Dye as "whatever governments choose to do or not do. Public Policy is What government do, why they do it, and what difference it makes". . Harold Laswell and Abraham Kaplan define Public Policy as "a projected program of goals, values, and practices . David Easton defines Public Policy as the impact of government activity . James Anderson defines Public Policy as a relative stable, purposive course of action followed by an actor or set of actor in dealing with a problem or matter of concern . So the conclusion of Public Policy based on the understanding of these experts is what the government does to achieve a goal. In this study, culture is a topic of local government issues that must be handled through a policy. Empirically the Dharmasraya Regency Cultural Preservation Policy is contained in the 6th Mission of the Regional Head namely "Increasing religious, customary and cultural values that reflect regional personality". Then Mission 6 was lowered into the Dharmasraya RPJMD for 2021-2026 and the Strategic Plan for the Department of Culture, Tourism, Youth and Sports for 2021-2026. Theoretically, the Regional Government of Dharmasraya has formulated a policy on culture. Furthermore, the formulation gave birth to a Program, activities and sub-activities that will be implemented. Winarno said that implementation is seen broadly as having the meaning of implementing laws where various actors, organizations, procedures and techniques work together to carry out policies in an effort to achieve policy goals or programs. . Lester and Stewart Jr. said implementation as a process and a result . The success of a policy implementation can be measured and or seen from the process and the achievement of the goals of the final result, namely whether or not the goals to be achieved . And according to Van Meter and Van Horn, Policy implementation encompasesses those action by public and private individual that are directed at the achievement of objectives set forth in prior policy decision. This include both one time efforts to transfrom decisions into operational terms, as well as continuing efforts to achieve the large and small changes mandated by policy decision . Then Merilee S. Grindle said that the implementation of state policy is actually not just starting with the mechanism for elaborating political decisions into routine procedures for bureaucratic channels, but also related to issues of conflict, decisions and who gets 'what' from a policy, so it's not wrong if an important aspect in the entire policy process is policy implementation . The implementation process is just starting, if the goals and objectives have been set, the program of activities has been arranged, and the funds have been prepared and distributed to achieve these goals. Based on Grindle's definition of "implementation", as well as the conceptual model and framework for implementing implementation as a political process and an administrative process. The successful implementation of a public policy is basically influenced by certain factors, both internal and external factors. Edward III identified aspects that were strongly suspected of contributing to policy implementation, namely: communication, resources, disposition or attitude of implementers, and bureaucratic structure. The four aspects influence the implementation of policies, either directly or indirectly, and each aspect influences the other aspects . Meanwhile, according to Van Meter and Van Horn, policy implementation runs linearly from public policy, implementor and public policy performance. According to Van Meter and Van Horn, several variables that influence policy are: 1) Implementation activities and communication between organizations, 2) Characteristics of implementing agents/implementors, 3) Economic, social, and political conditions, and 4) Disposition executor/implementor . Merilee S. Grindle explained that the successful implementation of a policy is determined by the content of the policy and the context of its implementation. The basic idea is that after the policy is transformed, then the implementation of the policy is carried out. Its success is determined by the degree of implementability of the policy. Then in more detail, the contents of the policy include: 1) Interests affected by the policy, 2) Types of benefits to be generated, 3) degree of change desired, 4) position of policy makers, 5) implements the program, 6) resources deployed . Meanwhile, on the other hand, the implementation context is: 1) Power, interests, and strategies of the actors involved, 2) Characteristics of institutions and authorities, 3) Compliance and responsiveness . The content and context of the policy is illustrated as follows. --- Figure 1. Implementation as Policy and Administrative Process The success of policy implementation is determined by two variables: the policy content variable and the policy context variable. The content variable relates to what is contained in the contents of the policy towards implementation. The context variables relate to how the political context and administrative processes are affected by the implemented policies . Furthermore regarding Policy Implementation, the Content Variable has six parameters/elements: 1. Interest Affected . The policies made have an influence on the political process which is "stimulated" by policy formulation activities. 2. Type of Benefits . Programs that provide collective benefits are easier to implement. Programs that only provide benefits and can be shared equally and are particular/specific can exacerbate conflicts. --- Extent of Change Envision . Programs in the long term, or demanding changes in actors, tend to experience difficulties in their implementation. --- Site of Decision Making . The more dispersed implementers , the more difficult it is to implement. --- Program Implementor (program executor ). The quality of program implementers affects the success rate of implementation. --- Resources Committed . Availability of adequate resources can support program implementation. Then, the Context Variable includes 3 elements, namely: 1. Power, Interests and Strategies of Actors Involved . The involvement of the parties is determined by the content and form of the program being administered. 2. Institution and Regime Characteristics . The capabilities or powers of the parties involved and the characteristics of the regimes in which they interact facilitate the assessment of the degree of opportunity for realizing policy or program objectives. --- Complience and Responsiveness . Implementing officials must focus on: how to realize the consistency of objectives in the policy. They must be able to change their opposing attitude from being harmed by a policy or program to accepting it. "Contents of Policy" basically has an influence on the possibility of the success of an implementation process. Whose interests will be affected, what benefits will determine the attitude and reaction of the community/target group to a program. It is the same with the Cultural Preservation Policy, because of the existence of rules regarding "synchronization of development planning to policy directions between the regional government and the central government", indirectly the Dharmasraya District Policy must be linear with the National Program on Culture. It is this synchronization that creates a symbiosis of Mutualism between the Regional and Central Governments. Regional government programs are implemented thanks to assistance from the Central government, and National Programs are achieved with linear regional policies. Likewise, to what extent a program will bring change, the level and number of decision-making, the role of implementers in the field, namely "street level bureaucrats" will determine the results of implementing a policy and the public's reaction to it. Of course available resources for a program also determine. In terms of the "context" of a policy, it must first be taken into account regarding the "power structure", as well as interest groups that are "stakeholders" in a policy, both in the sense of support and in the sense of rejection of a policy. In this case the Indonesian Ministry of Education and Culture has become a "power" that assists Dharmasaya Regency in implementing regional cultural preservation policies, especially assistance in terms of financial resources. Policy implementation actually also involves a network of political, economic and social forces which can directly or indirectly influence the behavior of the stakeholders involved, and which ultimately affect the impact, both intended and unexpected. . The term politics is not solely limited to political parties, political understanding is based on patterns of power and influence between and within the organizational environment. . The support of the political elite or the rejection of the bureaucracy at the national, regional or local level has an influence on the possibility of a program's success. Assessment of the "power capabilities" of the actors involved in the implementation process, as well as the nature and nature of the sensitivity of an organization and the existence of government institutions also determine the likelihood of the success of a policy. Political power as well as economic elite power, often affects the complexity of the policy implementation process objectively, because of the general tendency in the third world, where certain political forces and socio-economic forces prefer to wait for the implementation stage rather than before, as stated by James Scott that "A a large proportion of individual demands, and even group demands, in developing nations reach the political system not before laws are passed, but rather at the enforcement stage" . From a political point of view, looking at the dynamics within groups and the relationships between groups, the terms are found: implementers play, bargaining norms, involvement of interest groups, and sources of power and influence . Political linkages in influencing the success of a policy implementation indirectly are now known as the synchronization of development planning and policy direction of the regional government with the central government. Even the background of the regional head's political party will also determine the pattern of relations that will be established with the central government. Desa, 2021), making culture an issue and a regional problem that must be considered. This is even stated in the 6th Mission of the Regional Head, namely "Increasing religious, customary and cultural values that reflect the personality of the region" which will be achieved and translated into regional policies and programs for 5 years. Regional government program regarding Providing Operational Costs to Rajo, Nagari Traditional Meetings, and Traditional Apparatus in Tribes" which is part of the Regional Cultural Preservation Implementation process itself. This program is the only Regional Cultural Program and was first implemented in West Sumatra. Based on these empirical and theoretical phenomena, researchers are interested in analyzing and describing the Implementation of Regional Culture Preservation Policy in Dharmasraya Regency by using Merilee S. Grindle's Policy Implementation Theory. --- METHOD he approach used in this study is a qualitative approach. According to Bodgan and Taylor, qualitative research is a research procedure that produces descriptive data in the form of written words from the behavior of the people being observed . Researchers used interview and documentation techniques in this study, while the types of data consisted of primary data and secondary data. Primary data was obtained through interviews and opinions of informants, while secondary data was obtained from documentation and archives related to research. Data analysis was carried out by the process of organizing data consisting of field notes, recording results, documents in the form of reports by collecting, sorting, grouping and categorizing data so that it is easy to interpret and understand. According to Miles and Huberman, analysis activities consist of three streams of activities that occur simultaneously, namely data reduction, data presentation, and conclusion/verification . The Unit of Analysis is something related to the focus and components to be studied. The unit of analysis can be in the form of individuals, institutions or institutions, the whole program or the whole specific background according to the focus of the problem . The unit of analysis in this study is the Department of Culture, Tourism, Youth and Sports of Dharmasraya Regency.. --- RESULTS AND DISCUSSION harmasraya Regency has a multiethnic population consisting of Minangkabau, Javanese, Sundanese and Batak ethnicities. Dharmasraya has a long history, which is the origin of the birth of this area. This history has produced relics in the form of inscriptions, temples, gadang houses, mosques, statues, and others. In addition, there are also intangible cultural heritages in the form of Toga Dance, Randang Paku, Tanah Liek Batik, Silek Pingian Rantau Batanghari and others. These relics from past history prove that Dharmasraya has its own culture and traditions. In order to maintain the integrity of the regional culture and preserve it from generation to generation, the Dharmasraya Regency government has made culture a focus of attention that must be addressed and made it one of the government affairs in development planning and regional policy directions. The condition of the area prompted the Regional Head of Dharmasraya Regency to make "Religious, Customary and Cultural Values" one of the Regional Government Missions, namely the 6th mission. Based on Minister of Home Affairs Regulation Number 90 of 2019 "Religious, Customary and Cultural Values" entered into the Nomenclature of District Government Affairs in the Field of Culture. This nomenclature becomes the basis for regional development planning which is then translated into local government policies, programs, activities and sub-activities. In essence, this regional culture preservation policy has been formulated by following the rules of the central government, namely Law Number 5 of 2017 concerning the Advancement of Culture. Then based on the latest policies from the National Development Planning Agency and the Ministry of Home Affairs that the direction of regional policies and development must be in line and in sync with the Provincial Government and the Central Government. In this case the Dharmasraya Regional Cultural Preservation Policy must be in sync with the Policies of the Provincial Culture Office and the Indonesian Ministry of Education and Culture. The following is a picture of the flow of planning for the advancement of culture: --- Figure 2. Flow of Planning for the Advancement of Culture Law Number 5 of 2017 also stipulates a workflow for promoting culture which is supported by vertical integration between the central and regional governments as well as horizontal integration between the government and society. Both are managed through tiered planning starting from the district/city, provincial to national levels, namely in the order of preparation starting from the Main Thoughts of Regional Culture , Cultural Strategy and Master Plan for the Advancement of Culture . Meanwhile, in the preparation of PPKD at the district/city and provincial levels, various stakeholders at the same level must be involved. The Dharmasraya Regional Cultural Preservation Policy is basically a spearhead for the realization of the Vision of the State of Indonesia, namely the Vision of Indonesia 2040 namely "a happy Indonesia based on T D cultural diversity that educates, reconciles and prospers". That is, regional culture is part of the national priority program. To achieve these National Priorities, the Directorate General of Culture sets a goal of "Increasing the advancement of culture to optimize the role of culture in development" . As explained by Wahab that: "... The policy implementation function is to create a relationship that allows policy objectives or targets to be realized as an "outcome" of activities carried out by the government" . e. preparation of norms, standards, procedures and criteria in the field of cultural heritage preservation and cultural promotion; --- Purpose of the f. provision of technical guidance and supervision in the field of cultural heritage preservation and cultural promotion; g. management of cultural data collection system; h. implementation of evaluation and reporting in the field of preservation of cultural heritage and promotion of culture; i. implementing the administration of the Directorate General; and j. implementation of other functions given by the Minister. In carrying out a number of these functions, the Directorate General of Culture holds the mandate to succeed the fourth National Priority "Mental Revolution and Cultural Development" in the 2020-2024 RPJMN. The mandate is described in Chart 3 below. --- Figure 3. Mandate for Cultural Management in the 2020-2024 RPJMN Based on this explanation, the synchronization of regional policies with the central government will be well connected, vertically the goals set by the central government can be achieved and realized by the implementation of local government policies. This is in line with Ariyanta's opinion that "ideally, a regional regulation is not only an implementing regulation of regulations made by the central government, but also a reflection of the aspirations of the people in the region" . That is, the policies of the regional government and the central government must be in accordance with regional needs, solve regional problems, and provide benefits to the community. Ideally, the formulation of the Dharmasraya Regional Cultural Preservation Policy is the result of an evaluation of the performance of the previous regional government as contained in the technocratic document for the 2021-2026 RPJMD Draft. The evaluation results become regional issues and problems which are then determined as the direction of regional policy and development for the last 5 years. According to Adiyanta, with the existence of authority based on the principle of autonomy, district/city governments have the discretion to plan and manage the potential, sources of wealth and human resources in the regions. . However, in essence this authority is also indirectly limited by central government policies such as Permendagri Number 86 of 2017 concerning Procedures for Planning, Controlling and Evaluation of Regional Development, Procedures for Evaluation of Draft Regional Regulations concerning Regional Long-Term Development Plans and Medium-Term Development Plans Regions, as well as Procedures for Amending Regional Long-Term Development Plans, Regional Medium-Term, and Regional Government Work Plans Permendagri Number 90 of 2019 Concerning Classification, Codification, and Nomenclature of Regional Development and Financial Planning, as well as other regulations. With the empirical phenomenon of the Implementation of the Regional Cultural Preservation Policy which is influenced by political factors such as the Provincial government and the Central Government, the Dharmasraya Regional Cultural Preservation Policy will be analyzed and described using the Content and Context Model of Merilee S. Grindle Policy Implementation. This model is appropriate to use because there are resource, economic and political factors that influence the successful implementation of this cultural preservation policy. It is based on this reference that the Dharmasraya district government formulates issues related to culture, establishes a Policy on the Preservation of Regional Culture, and implements the Policy to achieve the expected goals. The following are strategic issues related to the culture of the Dharmasraya Region: 1. Lack of public knowledge about regional culture 2. Public ignorance about the protection of cultural heritage objects 3. Not all community leaders know and care about cultural heritage 4. The lack of facilities in the community for the development of regional culture 5. Lack of guidance on culture in the community 6. The influence of modern technology and the lack of cultural events. Then based on these issues the Regional government of Dharmasraya formulates and develops a Strategy in Preserving Regional Culture which includes the following: --- Regional Culture Preservation strategy The preservation strategy is determined by the regional government so that the noble values of culture that exist in a cultural tradition can be maintained even though they have gone through a process of changing the form of culture. The goal is to maintain the existence of the culture itself as a legacy from our ancestors to future generations. So as to increase the understanding and application of customary values and culture in social life. involve all related stakeholders such as BPCB, BPNB, provincial government, ministries and academics whose aim is to collect objects for the promotion of culture which include 10 objects for the promotion of culture and cultural heritage so that they can produce a strong regional identity.. 2) Socialization: To follow up on the results of the analysis carried out, we need to socialize the results, because with socialization we can convey the importance of developing and preserving local culture, with socialization we can invite the community to revive local cultural values that are starting to disappear. This can be done by forming a Cultural Community among the Young Generation. After the socialization is carried out, it is hoped that there will be implementation or application in the community. This implementation can be done in three ways. First, by introducing regional culture to the public which aims to revive culture which was previously known only to a few people or groups to become public information as a regional identity. Second, accustom the community to being active as actors in regional preservation. Third, preserving regional culture by reviving studios, associations, theaters, arenas to restore the functions of gadang houses and traditional institutions. 3) Promotion: The next step in developing regional culture is promotion which aims to disseminate regional culture through online and offline media and carry out regional cultural events. --- b. Protection The second strategy of Preserving Regional Culture is Protection. The first step taken is an inventory, namely the Recording, Data Collection and Documentation of Cultural Heritage. Second, security, namely the Registration of Objects for the Advancement of Regional Culture in the Integrated Cultural Data Collection System. The next step is saving in the form of Renovation and Revitalization of the Regional Cultural Heritage. c. Utilization The next strategy for Preserving Regional Culture is utilization. Utilization in question is the utilization of Cultural Advancement Objects to strengthen religious, customary and cultural values that reflect regional personality. It can be concluded that the steps to utilize local cultural wisdom in developing tourism in Dharmasraya Regency include several strategies. In developing local culture, local governments need to take steps that include monitoring practices that can damage or eliminate local cultural wisdom, because research that has been conducted shows that the factors that hinder the implementation of cultural preservation policies in Dharmasraya Regency are supervision and lack of available financial resources. --- Direction of Regional Cultural Policy Based on the Cultural Preservation Strategy, the regional government then determines regional policy directions consisting of: As conveyed by Grindle, the Context variable, the indicator "Power, Interests and Strategies of Actors Involved ". That the involvement of the parties is determined by the content and form of the program being administered, in this case the Implementation of the Dharmasraya Regional Cultural Preservation involves many relevant actors and stakeholders. As well as the indicator "Complience and Responsivenesi " in which the DPRD as an actor involved in the formulation of policies provides a positive response and responsiveness, this is evidenced by the implementation of the current Draft Regional Regulation on Preservation of Customs and Promotion of Culture. Apart from that, Strengthening the History, Customs and Traditions Sector also provides significant benefits for Indigenous Institutions and Stakeholders, as explained by Grindle that "Interest Affected" Indoikators explain that the policies made have an influence on the political process that stimulated by policy formulation activities. Then the indicator "Type of Benefits " that Programs that provide collective benefits, can be easier to implement. In accordance with the Grindle Implementation Model which states that "Policy Implementation is a Political and Administrative Process". This is explained in the Variable Content that the successful implementation of a policy is influenced by "Interest Affected " and Resources Committed ", where the Preservation of Regional Culture is also a part of the National Vision and Program , as well as the Strategic Plan from the Directorate General of Culture, the Indonesian Ministry of Education and Culture in this case as the Central Government. --- c. Strengthening Central government assistance in terms of budget is able to reduce the regional burden on the implementation of this policy, not only that the support and assistance of the central government accelerates and facilitates the process of implementing this policy. In addition, Grindle also explained that the success of a policy is influenced by several "Context" variables, namely Power, Interests and Strategies of Actors Involved and "Institution and Regime Characteristics" / regime)". The power possessed by the Ministry of Education and Culture facilitates the Implementation of the Dharmasraya Regional Cultural Preservation Policy because it has the same interests. This condition is exploited by the Regional Government as part of a strategy to succeed in policy implementation. Then the similarity of the political background or political interests of the Dharmasraya regional head with the highest Holder of Power in this case is that the Central Government makes the political process and administration of this policy run well. In other words, the implementation of the Regional Culture Preservation Policy is an urgent policy to be implemented both from the point of view of the Regional Government and the Central Government. This policy was implemented well thanks to the strategy formulated by the local government, the similarity of the vision and mission and goals of the local and central government, as well as the political support of the central government. Furthermore, the implementation of cultural preservation policies also requires support from various parties. For example, support from local governments, cultural institutions, local communities and the private sector. because the research results show that the implementation of cultural preservation programs really requires synergy between the government, local communities, the private sector and cultural institutions. This can be seen from the support provided by these parties in the form of financial contributions, procurement of venues, art performances, education to tourists, and active participation in cultural preservation activities. Based on the Strategy and Implementation of the Regional Culture Preservation Policy, the Regional Government will be able to achieve the targets and objectives of the District's Vision and Mission: Independent and Cultured Dharmasraya Maju. And in line with this, the dharmasraya regional government will also be able to contribute to achieving the Vision of the State of Indonesia, namely the Vision of Indonesia 2040 namely "A happy Indonesia based on cultural diversity that educates, reconciles and prospers". --- CONCLUSIONS he implementation of the Regional Culture Preservation Policy for Dharmasaraya Regency is going well thanks to complex policy formulation, appropriate Policy Strategies and Political Elite Support from the central government. The success of the implementation of this policy is strongly influenced by political factors, where the policy of preserving regional culture dharmasraya is synchronous and linear with the State Vision and National Program which has a positive impact so that the dharmasraya local government is able to work closely with the central government and even get budget assistance. Based on the phenomenon of the concrete implementation of the Dharmasraya regional cultural preservation, the researcher provides advice to local governments to make regional policies by: 1. Formulate policies according to regional issues, problems and needs 2. Establish policies that are able to involve and provide benefits to many relevant stakeholders 3. Make policies that are in sync with national programs or central government policies, so that cooperation is created that facilitates the implementation of regional policies.

This article analyzes and describes the implementation of the Regional Cultural Preservation Policy which is a form of effort to achieve the Vision and Mission of an Independent and Cultured Dharmasraya Maju Regency. Government Regulation No. 38 of 2007, cultural affairs is one of the mandatory affairs handed over by the central government to regional governments. This research is motivated by Regional Policy as a concrete form of overcoming strategic issues and regional problems in the Dharmasraya area, then expressed in the form of policy to achieve regional government goals. With this policy, this program became the only Regional Cultural Program and was the first to be implemented in West Sumatra. The approach used in this study is a qualitative approach, and uses interview and documentation techniques, while the types of data consist of primary data and secondary data. The results of this study explain that the Regional Government of Dharmasraya Regency has formulated regional cultural problems by elaborating on regional strategic issues which then determines regional culture preservation strategies and regional policy directions. The implementation of the Regional Preservation Policy is carried out systematically and in collaboration with several relevant stakeholders such as LKAAM, DPRD, Ministry of Education, Culture, Research and Technology, Ministry of PUPR, BPCN, and BNPB.

Introduction This is a story about Ron, Chloe and Polly. i They are disabled people from the North West of England, who, at the time of this research used the care services of a large UK non-profit organization. Controversy surrounds depictions of disability not least because descriptions are often grounded in normalising ideologies of ableism that discriminate through stigmatizing those identified as disabled . As a consequence, while recognizing that no language is entirely free of performative effects as discussed later in this article, we employ the terms disabled people and impaired bodies as what may be seen as a least worst usage. This care organization had initiated and funded a pilot project called CommunITy whose purpose was to provide 25 disabled people with reconditioned computers, necessary peripherals and, thereby, access to the Internet. The selection criteria revolved around the potential for disabled people who live on their own in geographically remote areas to benefit from technology , especially in relation to their participating more fully in the social world. Underpinning this was the belief that the way in which to facilitate this participation was to normalise disabled people through technology. The organizing principles of the CommunITy project, therefore, can be said to have emerged from ableist normative assumptions, where disabled people are seen as lacking something that can be 'fixed' so as to make their lives similar to those of the able-bodied . The research highlighted how filling in assessment forms is an ongoing embodied sociomaterial process in which the complexities of people/subjects and objects/materialities inherent in these processes are obscured as the administration is managed in practical, contextualized situations. The article focuses on the ongoing embodied, sociomaterial performances that are enacted as the assessment form ii is deployed as a framing device during selection interviews. Our research concerns the comparative neglect of discussions of disabled people in organization studies, although the specific problematic of this article revolves around how forms and their surrounding sociomaterial performances constitute, but are also transformed by, subjects, objects and organizational relations. As we proceed through our analysis, we identify how these assessments are informed by humanistic and normalising assumptions concerning individuals realising their potential to be sociable. In the conclusion, we explore how this can have the unintended effect of reinforcing, rather than diminishing, the marginalization of disabled people. We also address the over-reliance on semantics or texts within the literature on forms, which obscures the embodied sociomaterial performances of their enactment iii . In part, our research parallels that of Cooren, Kuhn, Cornelissen and Clark who argue that 'the vast majority of the work on organizational communication and discourse tends to focus on the textual aspects' at the expense of what human agents do and what is done by non-human agents. Seeking to extend beyond these textual or semantic analyses, our research questions are: what embodied sociomaterial performances are involved in the enactment of forms and how can they obscure the embodied experiences of disabled people? On an empirical level, we provide an ethnographic 'observer participant' account, which serves to confront the interactional entanglements of embodied subjects , their identities and expectations, and the objects , all of which constitute their sociomaterial accomplishments. With that in mind, we develop a theoretical framework through which to analyse the sociomaterial embodied performances that took place as the form was deployed within, and indeed beyond, the selection interviews. Our contribution on a theoretical level allows us to address the paucity of discussions on forms in the organization studies literature. This complements but also adds a distinctive dimension by highlighting how the form had certain agency and implications for disabled people but part of the analysis is about how human agents had to work on violating both the letter and spirit of the form in order to deliver on organizational demands to allocate computers. The article is comprised of five main sections. First we provide a brief examination of the literature on embodiment and materiality in order to develop a theoretical framework for our empirical study. This framework is drawn upon in the second section to explore our problematic, which is the paucity of analysis of forms or documents and/or the tendency to over-focus primarily on semantics at the expense of the embodied sociomaterial performances of their enactment. As our empirical data illustrates, these performances have a tendency to obscure the embodied, practical experiences of disabled people and those engaged in some capacity with their care. Thirdly, we offer an overview and critical engagement with our assumptions and some of the methodological practicalities encountered in the field. Fourthly we turn to our empirical material to discuss the performances that ensued as the assessment form was deployed in selection interviews for allocating computers. This is followed by a discussion section which draws on the theoretical framework and the literature of documents and organization more generally to analyse the empirical material. Finally we offer some concluding remarks to summarise the significance, contributions and implications of the article. --- Embodiment and materiality A critique of the tendency within certain strands of organization studies to privilege technology or humans, as if either could be seen as discrete entities that determined the other, facilitated space for a view of their sociomaterial entanglements. While clearly recognising that 'there is no social that is not also material, and no material that is not also social' , the embodied nature of these assemblages was not always given attention. Other literatures have, however, stressed the importance of bodies and embodiment in analyses of social and organizational life . A major strand of this literature is to challenge the dominant liberal and neo-liberal traditions that have prevailed in Western economies through an enlightenment philosophy underpinned by a Cartesian belief in the body as subordinated to the mind . By seeing the mind as an idea of the body , it becomes impossible to separate them let alone subordinate one to the other. Consequently, it is no longer exceptional to engage with individuals as integrated, embodied subjects, as occurs in everyday life by contrast with how academic representations of the body are often presented. Furthering this debate, in her discussions of embodied subjectivity, Dale argues that this Cartesian split often leads to assumptions of generalized 'normal' bodies as bounded entities, assumptions which obscure the particularities and processes of bodies. For her, '"Embodiment" is much more active, indicating the negotiation of everyday life in relation to the material world and to the creation of social life ' . This sociomaterial way of thinking sees events as comprising internally related, yet unbounded, elements that 'don't precede, but emerge from or through, their intra-action' and are thereby always in processes of becoming. In social life, bodies and objects are 'partial' in the sense that they are never separate and self-contained. Shoes, for example, are only realised as such when worn by bodies preparing to walk the streets and, at that point, bodies, shoes and pavements all merge into 'a dynamic network of convertibilities' that momentarily are defined 'through assemblage' with one another . This reflects an understanding of the entangled intradependence and undecidability of the boundaries between 'things' and 'body-agents' or how we represent the world of objects partly to arrest the instabilities of our bodies and selves . This representation of bodies in objects and their inverse helps us to explore how, in practical and situated circumstances, their attributes and the material world they inhabit involve dynamic, mutual exchanges. Within science and technology studies, there are a number of commentators who discuss exchanges in attributes between impaired bodies and the made world . Moser highlights the inter-and intra-related nature of impaired bodies and materialities in her discussions of an 'order of the normal' and the encounters of a man named Jarle who 'is paralysed from the neck down and needs a ventilator to breathe'. She discusses the materialities whereby sucking and blowing into a pipe activates a computer to scan and communicate 'all the alternatives aloud', thus facilitating his living at home. It was then possible for Jarle to select his options and manage his space , according to his requirements. These are not social events that can be separated in order to be located and analysed but emerge in practice, and are better seen as 'different moments in the unfolding biography of the artefact' . More importantly, there are complex inter-and intraconnections where impairment is managed through material artefacts against a background of embodied experience through various social and organizational constructions. Thus, embodiment can be seen as that which 'accommodates not simply the body, but the manner in which it is experienced and lived as an embodied subject' within a sociomaterial world. In addition, these exchanges involve 'heterogeneous engineering' and entanglements/intra-actions between different bodies and sociomaterialities . Take parking a car as an example. One day reverse parking in a tight spot, in between 2 other cars is relatively easy for the experienced drivera 'good' parking day. The next day, however, might generate difficulties for the driver -a 'bad' parking day. Neither the subjects or the objects may have changed but different particularities are being enacted to constrain the performance within specific, situated and practical situations. This example is a little simpler than the accounts concerning the allocation of computers to disabled people where a multiplicity of objects and subjects as well as their entanglements are involved. Bearing the preceding discussion in mind, we now turn to the focus of our research which is to theorize forms and their enactments to show how they are inseparable from the organizational and embodied world that their entanglement reflects and reproduces. --- Documents and Organization/Organizing As indicated at the outset, forms or documents are powerful devices that are not only a product, but also important in the constitution and reconstitution, of organizations. We illustrate this initially by drawing on different examples where the document is perceived as an object that ascribes, confines, describes and enacts particular versions of bodies, people and practices. Various commentators have highlighted the constitutive aspect of texts and textual agency in organization/organizing . For example, Lennie discusses the role of written plans in the constitution of organizational reality and argues that 'a plan, in giving direction, confers shape and substance on the world' and, even when there are deviations, arguably sets the parameters of action. Certain events are then selected out of given situations which lead to what may be seen 'as mappings' . These selection processes are examined by Smith in her discussion of an interview to decide whether or not a woman named 'K' had a mental condition. She indicates how in such interviews everyone involved carry with them pre-conceived ideas of what is and what is not a mental condition and whether a particular individual fits the category. As part of any interview, a 'cutting out' procedure takes place in which only certain features and events of the actual process are selected and documented. Moreover, during this process, the assumptions or pre-conceived ideas that precede the interview are no longer visiblethey are, in effect, written out. Directly in line with our focus on assessment interviews, the form serves to standardise categorisations as well as to constitute subjects and this is key to our understandings of how 'things' are performed in the reproduction of organizations. For McLean and Hoskin, , the scene of the form is like the staging of a play. 'At first before the curtain rises, the props are ready, and the script in place, but in suspended animation until the dramatis personae enter to play out the roles that text and context open up. Engaging with the script, moving between foreground and background across different moments and situations, they become readable in their truth, as this personae or that.' It needs to be recognised that by contrast with some documents , forms are only partially written and can be said to act 'on behalf of or in the name of the organization' . As a result, they await completion in being 'filled in', and it is this process that is our focus. Implicit in much of this discussion is how human agency/embodiment affects, or contributes to, the constitution of organization that takes place during processes of organizing; in other words, 'the moving between foreground and background across different moments and situations' and the adding on of bits, 'endowing with organs' . These studies usefully identify the part that all kinds of documents play in the constitution of organization. However, the explicit focus on the semantics of written texts has a tendency to obscure their conditions of possibility in embodied performances. This lack of focus on embodied performances has to a certain extent been addressed in other fields. Within sociological research, Bittner's study of the police on skid row, Garfinkel's studies of suicide and Reed's discussions of the warrant cover or prison intake document, for example, focus on the agential experiences and bodily performances taking place as documents are enacted within practical and situated circumstances. To illustrate, Reed examines these with reference to a prison in New Guinea where warders employ 'force to extract answers to the questions on the warrant cover' , whilst 'inmates do their best to conceal or disguise their own biographies ' . Closer to our field, Shildrick and Price discuss the claims procedure for Disability Living Allowance and how the self-assessment form is crucial. The questions claimants are asked to answer in order to complete the form imply that impairment is a 'fixed' category, in advance of it being observed. Completing the form produces objectifications that enable bodies to be recorded on the appropriate self-assessment form as 'broken' and fixed in this or that way. However, this is not the case for the impaired body, in so far as conditions are forever changing, often on a daily basis. For whether or not the body is impaired, it is 'always at risk of disruption' and of threatening the stability of organization but this is most evident when 'the body resists not just conscious control but predictability' . As identified above, the form can be seen as an organizational attempt to render impaired bodies predictable. These latter studies challenge the concentration on semantics by highlighting the importance of bodies and embodiment in discussions of documents. However, they are less explicit in revealing necessary 'intra-actions' of embodiment and materiality. Intra-action reminds us 'that "concepts"…and "things" do not have determinate boundaries, properties or meanings apart from their mutual intra-actions' . We now turn to a discussion of our research methodology before presenting the empirical findings. --- Methodology The purpose of the CommunITy project was to recruit and select 25 'suitable' recipients who would use the computers for social interactive purposes. Our empirical data is drawn from fieldwork undertaken by one of the co-authors in the North West of England. The stage of the fieldwork being discussed relate to the selection of 'suitable' users, which covered the period from 2006-2008 although reflections on this data are facilitated by longitudinal involvement which lasted for a period of almost 10 years. Some 18 months prior to the researcher's involvement, 25 disabled people had been identified, interviewed and accepted onto the project. However, during this time, the providers of the 25 reconditioned computers facilitating the project had become bankrupt and although a new provider had been identified, several of the identified users had either pulled out of the project, bought a computer of their own and, in one or two cases, had sadly died. When the researcher entered the organization, therefore, 13 new users were needed. At this time, the project co-ordinator was in the process of recruiting the remaining users needed to fulfil the remit of the project. Appointments were made by him for a visit to be made to each potential user's home for a selection interview. The researcher was involved with the 13 interviews iv . Jack and the researcher began an interview programme using an assessment form designed by Jack. There were a variety of reasons for the selection of the three users discussed in this article and for why we give much more space to our first case study where Ron was the applicant. Firstly, Ron's interview was the first that the researcher was involved with and the case study served as a model of the interviewing process for all three. Secondly, the time spent with these participants was longer and the visits more frequent than with many of the others. As a result, the interactions between the researcher and the various participants resulted in rather richer longitudinal data from before, during and after the interviews than was possible with the other applicants. Finally, the researcher's involvement with these participants exposes "'hot spots' that 'glow' for the researcher, whether encountered during the fieldwork, analysis or later" . These are the cases that best exemplify how the initial requirements were negotiated in order to complete the form and allocate the computers. These were negotiated in all of the 13 interviews. Of course, there were both positive and negative outcomes, for some recipients still use the computers saying "they would never be without one now" whereas others have either returned or never use them. The form to assess suitability was used to guide discussion in all of the interviews and was, in most instances, completed in part during the interview situation and then finalised after the interview. The interviews were recorded, with permission, using a digital voice recorder but, in addition, the researcher kept a diary throughout the project in which the details of each visit were noted as soon as possible. All interviewees were assured of confidentiality and anonymity should the data generated from their interviews be used in future academic or practitioner discussions of the project. At all the interviews, the researcher was introduced not only as external to, but also as a participant in, the project; she was seen and began to see herself as a member of the organization v since she was fully active and immersed in the setting. In relation to this, Wacquant suggests that we need to foster long-term, intensive, even initiatory, forms of ethnographic involvement to allow the investigator to master, in the first person, the social phenomena being investigated. In terms of data analysis, this was very much guided by the events that occurred in the field in relation to the sociomaterial complexities and the disruptions resulting from the enactment of the assessment form. As these evolved, through her own reflections, the researcher selected a range of these and recorded in the research diary. Accordingly these complexities, each of which differed in individual and embodied sociomaterial circumstances, were analysed in relation to the inscriptions on the assessment form, the aims of the project as inscribed on the initial proposal document, and the actual events that took place in the field. The nature of this methodology sometimes comes at a cost. Firstly, it is difficult to separate yourself from the lived reality in order to make sense of it. However, this is partially overcome through joint authorship with a colleague who has not conducted the empirical research. Secondly, as Coffey suggests, ethnographic methodologies might produce something of a 'passionate analysis' whereby the researcher's own ethical engagement plays a key role in the representation of data . Thirdly, and to illustrate the previous point more fully, Geertz discusses how these writings/representations could be classed as 'fiction' since they are 'crafted by their authors' . Indeed, it has been claimed that our empirical research often 'serves only as excuses for our theoretical works that are essentially fictions' . So, not only do the practical and situated circumstances come to be modified by the researcher's engagement in them, so too does the knowledge produced as a result of this engagement. However, to be overconcerned about these matters is usually a legacy from our positivist past where, even when committed to phenomenological or poststructuralist analysis, we cannot help looking over our shoulder, anxious that we might meet with the wrath of the oppressive scientistic master who denies us space for the body and embodiment , or the serendipity of random events and elements of disorder. Accordingly, a reflexive methodology has been adopted whereby rather than falsely denying or discounting the multiplicities of complex interactions and intra-actions between subjects, bodies and representations, we approach 'the process of subject formation in a distributive, dispersed and multiple manner ' through ontoepistemological thinking . This means that we regard not only the research participants, but also the researchers, as engaged in complex embodied sociomaterial relations of interpretation and representation that reflexively form the images, materials and meanings from which we articulate a research account. --- Constituting the User: The Form The project's co-ordinator had developed/designed an assessment form to be used for the interviews of all potential service users. The form consisted of seven sections and had inscribed within it a series of questions typical of which are the ones regarding IT requirements . --- Figure 1 -The Assessment Form vi The original template for the assessment form was taken from a project named Workability initiated 10 years earlier by the same organization aimed at providing disabled people with reconditioned computers to facilitate access to the workplace. Guided by this and the initial proposal document for the CommunITy project, Jack designed the form in relation to how the organization expected the 25 potential recipients to use the computer 'to maximise their involvement in social, recreational, leisure and educational development'. This is evident in Section 6 of the form, which asks the question 'what do you want the computer for?' There are a list of fixed yes and no response boxes to all the questions, arguably pushing respondents in a certain pre-defined direction of seeing the computer as facilitating more social interaction as a means of reducing their isolation. The final open question of other provides for some different responses but because of the closed nature of all the other questions, respondents were left apparently in some difficulty. For example, a number of the participants interviewed would ask for further clarification of an 'open' question and seemed more comfortable with those requiring fixed answers. This was partly because they had little or no prior experience of using a computer. The questions inscribed on the form have been selected out of a variety of situations and circumstances including, for example, the Workability project, its aim to reduce the social exclusion of disabled people from the workplace, the assumption that technologies can help to achieve this aim, combined with Jack's interpretations of the project, informed by the initial proposal document. A mapping/cutting out procedure has taken place and, in terms of selecting suitable users, the scene has been set enabling certain sociomaterial and embodied accomplishments. We now turn to the three empirical case studies that we have selected as a way of illustrating our arguments, first of which is the full selection interview script of the allocation to Ron after which we briefly consider Chloe and then Polly. --- The Selection Interview --- Case Study 1 -Ron What follows is a discussion of an interview, which took place as Ron was being assessed as a potential CommunITy user whilst lying in bed at home. Ron was in his early fifties and an expoliceman. He had multiple sclerosis and as a result of the deterioration of his body, was categorised as quadriplegic. He was only able to move from his neck upwards. He had a possum vii to operate his telephone, his television, his lights and his front door. Ron lived in a supported living scheme and was supported by staff on a 24-hour, daily basis. He did therefore have quite a high degree of social interaction with his support workers. He also had a lot of interaction and care from various family members. He did get out of the house but this was only with help and these times were limited to approximately once a month. Present at the interview were Ron, his sister/carer, a key/support worker, Jack and the researcher. Below is a transcription of the interview process in which the assessment form was used as a guide to discussion although there was not always a strict adherence to the questions inscribed on the form. The form was completed some days following the interview process. Jack outlined to Ron that the aim of the project was to allow people more access to the outside world in an attempt to increase their social interactions. Jack introduced the researcher to Ron as someone who was to spend a period of time involved in the project, with a view to writing an evaluation report. Researcher: As Jack said, I'm working on the evaluation side of the project in the hope that we can see how or whether expectations have been filled further down the line. etc, resulting in an increase in their social interactions . Ron's responses in relation to not wanting to use the computer for socialising in this way could have disqualified him from being allocated a computer. However, this discrepancy was ignored when the form was completed since it was felt that Ron complied with the spirit, if not the 'letter', of the project even though he did not strictly meet the criteria inscribed in the form. So, what Ron as a physical and social being found appropriate and what was being offered to him socially were incompatiblehis physicality and sociality were arguably entangled in complex ways that were not consistent with the aims of the project. In the open question, respondents are encouraged to think about other uses for the computer, and Ron's carer pre-empted his response by asking, 'Can you play games on it?' Ron demonstrates an interest only in logical, not social, games but this enabled his carer to assert, 'That'll keep your mind active as well -it'll be good for him', thus providing some appropriate material for completing the form. To this final response of 'other ' the completed form reads, 'To play games online'. Therefore, although the main reason Ron gave as to why a computer would be useful to him was the ability it would give him to deal with his finances, this was not indicated on the assessment form. The question needs to be asked as to why this was the case? We would suggest that 'to play games online' was a better fit for the notion of increasing social interaction on the basis of which the project was developed/designed. This was selected/cut from the discussion and inscribed onto the form as it was seen as a more suitable response than the one given by Ron i.e. 'buying or selling shares'. The reason for this selection was that it followed the researcher's daughters positive social experience of playing an online game. Thus, in this particular instance, it indicated a possible or more likely way than buying and selling shares to increase Ron's social network and thus reshape in the written document the actual embodied sociomaterial performances that had taken place. viii --- Interview inconsistencies and frustrations The conversation turned to what, where, and how Ron could set up a computer in order for it to be useable. Jack asked him about his IT requirements. Ron responded that this is what he thought the meeting was about. At this point there was some confusion as none of those present were IT specialists and had little knowledge of what adaptations etc. were needed or could be bought. Ron's carer mentioned a laptop but the project had been provided with 25 reconditioned computers and a relatively small budget for adaptations but also Ron had indicated that a laptop was not feasible because of his spasms. We see evidence here of what Cooper might call the mis-representation of bodies in technologies for there was no way he would be able to use a laptop. There were problems of accessibility and space. [Ron lives mainly in a small bedroom in which there is his large hospital type single bed, his large flat screen TV which is mounted on the wall and just enough room at either side of his bed for one person to be able to walk and care for him]. He spoke about the support staff/carers being able to move the computer out of the room [the idea being that this would be next to his bed on a work station with wheels] when they needed to get access to him. Observing Ron's bodily expressions during this interview, the researcher was able to identify the building up of frustrations. For example, he started to roll his eyes, to shake his head and it was visually obvious that the conversation was beginning to annoy him. His frustrations were made explicit when he stated, angrily, "no one is listening to me". In some sense, Ron was voicing a concern of being ignored by professionals, whose organizationally and form driven preoccupations with managing the allocations, were seemingly given precedence over the wellbeing of their clients. --- Managing the inconsistencies Directly after the interview, Jack and the researcher discussed how Ron was not really a suitable candidate in terms of the formal aims of the project since he was neither socially or geographically isolated. ix This they reasoned was a result of the amount of interaction he had with his carers and the support staff whom he sees frequently throughout the day. Another potential reason for refusing an allocation was his apparent unwillingness to see the computer as a tool for improving his social interactions since he was only interested in games of logic or playing the stock market. Despite these reservations, it was eventually agreed that Ron was someone who would benefit from having a computer with Internet access. This, it was felt, could lead to an increase in his independence as long as he was able to use the computer without too much difficulty. They discussed also how once he got online the likelihood was that he would use email and communicate with others on an informal level. In their thinking the chances were that his social networks might also increase so Jack and the researcher decided that Ron should have a computer. This was not just because computers needed allocating . More substantively, he was accepted because he had simply been stuck in a bed for the past few years. He was accepted because the only real entertainment to which he had 24 hour access was his widescreen television. Jack and the researcher were made aware that Ron had difficulty sleeping, as he was in pain a lot of the time and that during the night he found it particularly difficult when he could not sleep. All he could do at these times was watch television . An additional factor was that Ron's sister had approached them in the hallway of his house and asked how long they thought he would have to wait for a computer, if accepted. She told them that her reasons for asking and seeming so "pushy" were that he had been promised many things from the organization in the past and that she would hate for him to have to suffer another disappointment. All of these issues were discussed by Jack and the researcher in the conversation that took place following the interview and were later reflected on and recorded in the researcher's diary. Jack formally completed the assessment form some days later and Ron was subsequently allocated a computer. --- Case Study 2: Chloe Some weeks later Jack and the researcher went to interview a 25-year old woman who used a wheelchair and lived in a supported living scheme. When they arrived Chloe was with two carers and the researcher guided by the questions inscribed on the form began the interview. Questions such as 'what would you use the computer for?', 'would you use it for communication?' were put to Chloe who was struggling to respond. Shortly after the interview started, her mother entered the room and shouted "you're a week late, it was meant to be the 14 th and we're going out now"the interview was subsequently re-arranged. Chloe's mother was present at the next interview. Jack and the researcher were quite anxious having the mother there, given her anger at the previous meeting so were happy to let the interview proceed as the mother pleased. It was learned that Chloe had had leukaemia when she was 7 years old and as a result of medical treatment, she had been left fully paralysed down the right hand side of her body, including one side of her brain. Chloe had difficulty communicating as a result of her impairments and her mother who was the main carer responded to the questions being asked that were inscribed on the form. Chloe was asked whether she would like to use the computer for online shopping. Her mother responded that she would not let her shop online as she liked to take her out as much as possible to do her shopping . She added that Chloe also went out of the house a couple of days a week to visit day centres where she would meet up with many of her friends. After the interview, Jack and the researcher had a conversation in relation to her suitability for the project. They felt that because she gets out quite a lot, she did not fit the social isolation criteria of the project. However, they agreed that although she might not be socially isolated, she lacked independence largely because of the way her mother was highly protective of her. Thus the criteria of suitability were broadened to include independence since this enabled Chloe to be allocated a computer. So, as with Ron, the form set the scene but the embodied sociomaterial performances determined the outcomes. --- Case Study 3: Polly Another case involved Polly who was in her mid-sixties and had been diagnosed as having Lupus, Angina and Arthritis. Although able to walk, she tired very easily and got breathless with over-exertion and extremely uncomfortable if her body was positioned in a way that affected her impairments. Prior to the interview, the co-ordinator told the researcher that she "experiences depression from time to time as a result of being at risk of social isolation". At the interview, Polly had seemed like a dream in comparison to Ron and Chloe, in the sense that she responded positively to the questions about using a computer for purposes of social interaction. This allowed all the boxes to be ticked on the assessment form without having to manipulate the responses, or needing to persuade her that this was a potential solution to her social exclusion. Following, what could be termed, the successful nature of Polly's selection interview, she was allocated a computer. Two months after the installation, Polly contacted Jack to inform him that she did not want the computer anymore. She told Jack she could not afford broadband and that she found the computer uncomfortable to sit at and to use. She told him how it was making her miserable, worrying about the discomfort and the cost and that really she had never wanted the computer in the first place. It can be seen here how the form served to constitute or sustain the welfare organization partly through Polly's concern to be socially accepted by responding positively to the interviewer's questions, thus allowing a computer to be allocated. Clearly in this case, however, while the organization was sustained through embodied sociomaterial relations, it was at the unanticipated expense of Polly's bodily and mental wellbeing. Polly's computer was returned and she was removed from the organization's database. --- Discussion As we have sought to argue, the problem for us is the paucity of discussion in the literature on the part that the form plays in the constitution of subjects, objects and organizational relations. In addition, when there are discussions of forms/documents more generally, the almost exclusive focus is on the texts and semantics rather than their implications for organizational relations. Our research has attempted to re-dress this imbalance by drawing attention to the embodied sociomaterial performances that take place as documents are enacted. In relation to these matters, the empirical data has demonstrated how at the beginning of Ron's interview it was becoming clear that he wanted to use the computer for 'buying and selling shares'. At the beginning of Chloe's interview, it was clear that her mother did not want her to shop online as she liked to take her out shopping, and at Polly's interview it seemed certain that she wanted the computer for interactional reasons which fitted well with the ideal of the project. In respect of all three cases and their potential suitability, one might say that the form was shaping the limits and possibilities of the selection process . In order to become part of the CommunITy project and be allocated a computer, there was an assumption that candidates presumed a standardised subject position whereby they would want to use the computer for social interactional purposes. However, Ron and Chloe wanted the computer for different reasons and made that quite clear. From the very outset, then, Ron and Chloe's mother were not performing in accordance with expectations that were inscribed on the assessment form and this was becoming problematic. The form was designed, therefore, with the purpose of overcoming the marginalization of disabled people by increasing their sociability through digitalization. However, once enacted in these interview situations we see the instability of these expectations which became reconfigured in each and every empirical example discussed throughout this article. Drawing on Smith we suggest that the performance and indeed outcome of all of these interviews is conditioned by the different interpretations of those involved in relation to the normative framework underpinning the project and the form as an inscription device for determining and justifying outcomes. Yet these interpretations and the procedure of filling in the form became subordinate to achieving a material outcome of digitalizing disabled people through distributing computers to them. In Ron's case, a range of interpretations of the different participants existed prior to the interview so that they were all 'singing from slightly different hymn sheets'. However as the questions on the form were addressed and became part of the performance, attempts were made to align the different interpretations, including those of Jack and the researcher, with the requirements of the project. In addition, the form symbolised the authority of Jack and the researcher as representatives of the non-profit organization and the presumed participation of a network of materialities, Ron and his carers. Following this line of thinking, these attempts were aimed at making everyone 'sing from the same hymn sheet' in order to constitute Ron as a suitable user. This was evident in relation to the question of why Ron wanted the computer for it was simply assumed that Ron would provide fitting answers that could allow the boxes to be ticked and enable him to become a CommunITy user. As is seen in the transcript, Ron's negative responses to questions about email, online shopping and social forums created difficulties in administering the form. Ron's interview transcript revealed he was resistant to answering the questions in a way that would allow the right boxes to be ticked. Chloe's mother also, seemed adamant not to be pushed in the direction of, in her case, shopping online as she wanted to continue to take Chloe shopping during the week. Ron and Chloe were refusing to fit in with the rationale behind the project as inscribed in the form. None of these refusals to fit are documented on the completed assessment formsthey are indeed glossed over . Polly was different in the sense that she responded in a positive way to all the questions being asked. Usually particularly in public welfare, resources are limited and a competitive process takes place in order to distribute them to deserving cases. In such circumstances, we can assume that processes of allocation might run relatively smoothly. Here, however, because of the previous failure to allocate computers, resources were not scarce, yet the façade of a public display of competition and allocation on the basis of merit had to be maintained, as evident throughout the interviews. Polly was unaware of this failure and, we would suggest, saw this as a competitive process which led to her complying with requirements in an unproblematic manner in order to get a free computer. Perhaps had the process been informal and less focused on filling out the form, Polly may have intimated her reservations in advance of being allocated a computer and thereby not suffered the mental and financial anxieties it provoked. This discussion serves to demonstrate the power of the form in relation to the constitution of the CommunITy project and how it ascribes, confines, describes and enacts particular versions of bodies, people and practices. It highlights certain normative assumptions that are inscribed in the form; namely, of disabled people somehow lacking the ability to fully participate in the social world and therefore in need of a technological 'fix'. It also illustrates how these assumptions underpin the actions of Jack and the researcher when they seek to repair the performances. In addition, it alerts us to how these assumptions serve to marginalise disabled people as evidenced in the glossing over of Ron's wanting to buy stocks and shares, his frustrations that 'no-one was listening' to him, and Chloe's mother not wanting her to shop online. We have evidence of the contribution of human agency/embodiment and materialities and their different particularities in accomplishing and sustaining organizational practices through processes that are enacted before, during or after official selection interviews. For example, for Ron, the interview was about matching the IT to his impairment, for his carer the interview was about getting him something out of the organization which often in the past had not happened , for Jack the interview was about allocating resources and about helping disabled people increase their sociability and for the researcher the interview was about generating data to write a thesis and helping Ron to get a computer which she saw at the time as offering him a way of improvinghis life. Following the interview, Jack and the researcher discussed how even though Ron deviated from the expected responses that would have helped complete the form and identify him as a suitable user, he should have a computer anyway as this would likely increase his independence. The same was true of Chloe who was identified by Jack and the researcher as warranting a place on the project in order to increase her independence. This illustrates how the form and the form filling process was powerful in constituting not only the users but also the interviewers and occasionally marginalizing the agency of these disabled people by glossing over the sociomaterial, embodied relations. However, like rules in a bureaucracy, the form was perhaps more about legitimizing rather than directly determining outcomes, for our research has shown how participants of the interview process negotiated their way around the inscriptions and prescriptions on the form. To illustrate, Jack and the researcher had their own interpretations of the projectconditioned arguably by the form, their own morality in dealing with disabled people, and their own pragmatic concernsall of which led to the additional inclusion of independence eventually almost displacing sociability as a criterion of suitability, especially in the case of Ron and Chloe. The research provides evidence of the emergent, entangled and embodied 'ongoing materialising performance of the world ' which contributed to the 'repair work' that continually took place throughout these processes of selection. It reflects how the goals of the organization of distributing computers to reduce social exclusion as inscribed on the form were interpreted to accommodate the discrepant behaviour of Ron and Chloe. We illustrate these ongoing sociomaterial performances through some further empirical data since the researcher visited Ron again twelve months after the original research. Ron had bought a workstation with wheels and this enabled him to use the computer and voice recognition software but only when his carer or someone else was available to give assistance. They were needed to wheel the workstation into his small bedroom, to switch on the computer and to put on his headset, which he needed in place to use his voice recognition software. Ron also needed someone to be there when he had finished using the computer in order to take off his headset, turn off the computer and wheel the workstation out of the room. At that time, Ron was suffering from a great deal of neck strain which was a result of having to turn his head to the left hand side and look down slightly so that he could see his computer screen. These continuing implications are also evident in the discussion of Polly and the events that took place two months following the installation of the technologies in her home. This is because Polly was suffering discomfort from sitting at the computer, and was worrying about the costs of broadband, and this eventually resulted in stress and the onset of depression. These were some of the events that led to Polly contacting Jack and telling him she did not want the computer anymore and had never really wanted it in the first place. In both examples we can identify notions of 'active embodiment' where bodies cannot be seen as fixed entities independent of time and space. For bodies are continually changing as is the material world which they inhabit, and different rationales are created and deployed to account for the diversity and the failure of the form to reflect this. The performance of Ron, Chloe and Polly in the selection process for the CommunITy project were conditioned in part by the assessment form/document, their bodies and the available technologies. They should therefore be identified as an ongoing and emergent embodied sociomaterial accomplishment in which multiple entitiesin this instance the form and the particularities of bodies and of materialities are involved in complex entanglements. --- Conclusion This article has presented an embodied sociomaterial analysis of how an assessment form facilitates and obstructs the selection/organization processes of fitting three users into the category of 'suitability' for the CommunITy project. We have demonstrated how documents such as the form and their social and organizational context generate limits and possibilities for applicants who seek to be 'suitable' users for the CommunITy project. However, these limits and possibilities are in turn conditioned as well as reproduced by the embodied, sociomaterial performances that surround the deployment of the assessment form. For example, Ron had suggested that his views were being ignored by the professionals who seemingly defined reality for him and, in this sense, we can see how the form was having the unintended effect of reinforcing a degree of marginalization. This was partly a consequence of how the CommunITy project embraced humanistic beliefs in disabled people realizing their potential through social communication, so that Ron's disinterest in the computer as a means of social interaction was a challenge to the charity's ethos and therefore had to be ignored. The computer was seen as a 'technological fix' in facilitating communications, regardless of whether it is the most appropriate way of doing so and of the extent to which this is even of any concern to particular disabled people. Such an imposition as a norm can have the effect of marginalizing disabled people even further, either by making them feel inadequate in not realizing themselves in this way or in feeling deviant by not complying. It cannot be ignored how language and, in our case, the very concept of marginalization is 'performative' in the sense that it is active or consummates an action and thus has an effect in bringing about the very 'thing' to which it refers, or reproduces what it claims merely to describe . This occurs not merely because language and concepts are rarely innocent politically but also by virtue of simply marking disabled people out as in 'need' of special attention. This is one of the unintended consequences of welfare where there is a fine line between care and patronising power, which our case study illustrates quite sharply. It is partly why within the recent literature on ethics, traditional deontological, consequentialist and even virtue approaches have been criticised in favour of a return to the ancient Greek focus on caring for others through first caring for the self . For care is most effective when it is inscribed in everyday sociomaterial, ethical and embodied relations and not the end product of external interventions that have difficulty avoiding the charge of patronage, especially when they are informed by unreflexive humanist assumptions or are driven instrumentally to meet organizational targets and goals. Fortunately, Ron was a strong enough character not to be intimidated by this exercise of power implicit in the agency of the form, and to a lesser degree in the power of welfare agents. Also we should acknowledge that in everyday life this kind of ethical care is routinely practiced despite organizations and institutions imposing particular beliefs and procedures that can contradict it. We have analysed the part that the assessment form plays in constituting and organizing the CommunITy project. In doing so, we have shown also how the assessment form is distinct insofar as it is incomplete until filled in whereas other documents relating to, for example, plans or strategy have an immediate effect upon action. Of course, the assessment form guides action, but does not actually determine it for, as we have seen in all three of the selection processes discussed, the organization of ongoing and emergent, embodied sociomaterial relations, as a form of 'heterogeneous engineering' , facilitated an effective CommunITy project outcome. Our stories of selection and how the assessment form is embedded yet transfigured or transgressed in the process allows us to display the bodies/materialities/documents and their intra-action with selection/organizing processes that were involved in the allocation of a computer to disabled people in this empirical setting. In addition, our empirical research and analysis on the deployment of assessment forms in allocating computers to disabled people draws explicit attention to the enactments of agents, technologies and their representations in interviews concerning the sociomaterial and embodied relations and the social organization of charitable work. What we have demonstrated is that the form to be written and the written form always continues to 'matter' throughout processes of selection, but that to explore 'how' it continues to 'matter', there is a need to understand the enactment of forms as a complex embodied, sociomaterial performance whose outcome cannot be wholly predicted. In summary, our contribution has been to address the paucity of discussions of forms in organization studies, which results in a neglect of the part they play in the constitution of subjects and organizations generally, and specifically in our example of the charitable distribution of technology to disabled people. In particular, it has sought to compensate for the limited research of embodied sociomaterial performances that ensue as forms or documents are enacted within situated and practical circumstances. Theorizing these enactments has enabled us to identify how forms and form filling processes shape our organizational realities and how an overfocus on forms can lead to a glossing over/obscuring of the ongoing performances that are taking place in situated practice. We have identified how these performances are themselves entangled with forms and form filling and shown how organizational and indeed social reality is a complex outcome of engagement with these entanglements. An important implication of this research, which has become increasingly evident throughout the review process, is how disabled people can readily become marginalised. This occurs not just through conventional processes of stigmatization but also because of normalising procedures that define disabled people as lacking something, such as 'good' social and communicative relations, and in our research, seeking to deploy technology as a way of restoring it regardless of whether or not this is their priority. A significant conclusion of our research is to argue that insofar as humanistic assumptions about sociability as well as realising human potential are prevalent within charitable organizations, they may have the unintended effect of normalising, rather than embracing difference, thus contributing to, rather than eliminating, the marginalization of the recipients of welfare. While this is of practical concern for the disabled people in our study, it may have much wider implications for welfare in general. --- Researcher: What about meeting new people on line? Ron: No, not interested. Just want to research info, buy stuff, sell stuff. Researcher i All names of participants and organization are pseudonyms. ii We use the term document and form interchangeably in the article although the former term takes precedence in the title. The limited literature tends to use the term form. iii Mol suggests 'that in the act, and only then and there, something is-being enacted' . iv The researcher was also involved with the installation of all 25 computers and visited the 25 selected users to evaluate their use of the computers some 3 to 4 years after the allocation. She also kept in contact with a number of users for several years, one of whom she last visited and re-interviewed in 2016. v This is evident in the transcript taken from Ron's interview. vi Other sections omitted questions such as name, address, registered doctor, convenient times for volunteer to visit, etc. vii A possum is an environmental control system, which in this case was operated by blowing into a tube which is fixed in a position close to Ron's mouth in order to make it easily accessible to him. viii Although buying stocks and shares involves some social interaction this is rather minimal unless employing a stockbroker ix These aims were inscribed on the proposal document that had been discussed at length, and as a result underpinned much of their thinking.

This article examines the selection process in a pilot project aimed at distributing computers to disabled people to allow for digitalization. Particular attention is paid to the complexities generated by an allocation assessment form, designed to help these people improve their social interactions through electronic media. There is a paucity of discussions on forms in the organization studies literature but when studied, an over reliance on semantics such that their enactment in embodied sociomaterial performances is easily glossed over. Our problematic revolves around how forms and their surrounding sociomaterial performances constitute, but are also transformed by, subjects, objects and organizational relations. The contribution of this article is, therefore, to address the embodied enactments and sociomaterial practices that are embedded within these allocation processes. So, for example, assessors in the project deviated from a strict interpretation of the questions on the form and sometimes ignored clients' responses so as to prevent formal allocations of computers from being seen as illegitimate, and potentially disruptive to the organization's objectives of distributing digital devices. This enabled us to focus on the sociomaterial and embodied relations that are enacted within the selection process and how these place limits on, but also possibilities for, those allocating and those seeking to be allocated computers. The case study shows how distributing computers to disabled people is a complex sociomaterial process that is conditioned by the embodied performances and textual devices deployed. At the same time, the process was informed by humanistic and normalizing assumptions about sociability that are inscribed on the assessment form as criteria for allocating the computers but one implication, we found, has a tendency to reinforce the marginalization of disabled people.

baCkground Social systems intersect and structure children's educational achievement. Urie Bronfenbrenner's ecological model is premised on the idea that both distal and proximate factors influence children 's development . This framework emphasizes the interrelationships between individuals at the micro level and their environments at the macro level . The model sheds light on factors that influence educational risk , the experiences of immigrant families , and family well-being in the COVID era . Bronfenbrenner's model calls attention to the idea that family members' lives and destinies are linked. --- Immigration Policy Federal, state, and local policies have increasingly targeted Mexican immigrants, the largest foreign-born group in the county, affecting their ability to live and work in the United States . For example, immigrant families and their U.S.-born children are excluded from important social safety supports that buffer the effects of poverty . At the same time, Section 287 of the U.S. Immigration and Nationality Act promotes partnerships between Immigration and Customs Enforcement and local law enforcement officials that request them. Once 287 partnerships are established, ICE can formally train and supervise local immigration enforcement activities . Since its inception, the 287 Program has experienced varying levels of partnership interest. Counties with 287 partnerships are found to have higher levels of community instability and household mobility . The Great Recession and its associated economic downturn ushered in a wave of statelevel immigration policies . Data collection for this study was contextualized by ICE arrests reaching a high of 232,796 in 2009 and l o W -i n c o m e Fa m i l i e s i n t h e t W e n t y-F i r s t c e n t u r y r s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s federal funding for 287 peaking at $68 million in fiscal years 2010 through 2013 . Despite the U.S. Supreme Court case Trump v. State of Hawaii et al. that ended President Donald Trump's "zero tolerance" and family separation policies, ICE's deputy director recently wrote that "ICE continues to use 287 partnerships to assist state and local agencies in ensuring the safety of their communities while working to expand the program consistent with EO 13768, Enhancing Public Safety in the Interior of the United States" . 1 Former Maricopa County Sheriff Joe Ar paio's change from "tough on drugs sheriff" to "ruthless immigration enforcer" was brought about by powers granted through the 287 program . In 2007, ICE trained more than 150 Maricopa County sheriff officers. Sheriff Arpaio relentlessly pursued immigrants until abuse, racial profiling, and civil rights lawsuits ended the partnership between ICE and the Maricopa County's Sheriff's Office in December 2011. This termination did not halt Arpaio's efforts, however . In 2010, the year before this study began, Arizona enacted Senate Bill 1070 , the first law in the country that made it a state crime to be an undocumented immigrant . Built on the enforcement mechanism authorized through the 287 program, SB 1070 deputized sheriffs to enforce federal immigration laws. SB 1070 had four key provisions. First, police could demand and investigate an individual's documentation status, referred to as the "show your papers" provision. Second, police could arrest individuals without a warrant on the presumption of their being undocumented. Third, not carrying federal registration papers indicating one's authorization status in the United States became a crime. Fourth, it also became a state crime for an unauthorized immigrant to seek or accept work in Arizona . In June 2012, coinciding with the end of data collection for this study, the Supreme Court struck down all provisions of SB 1070 except for the first "show your papers" provision. Although Phoenix was a hot spot for immigrant deportation during this study, immigration policy partnerships were not uniform across geographical contexts. Data from San Antonio's ICE field office suggest a less punitive approach to immigration enforcement: of the sixty-eight thousand immigrants who were deported from Texas in 2012, almost 70 percent were convicted criminals or repeat immigration violators; the remaining 30 percent were recent border entrants . In this study, we examine a national policy context hostile toward immigrants. Using a socioecological framework, we examine how variations in state and local immigration policies structured parents' broader set of social relations and environmental influences; we focus on implications for children's development. --- Work and Home Immigration policies and their strict enforcement have changed the nature and experience of work for Mexican immigrant households: first, by restricting employment and criminalizing those without valid documents; second, by using workplace raids to deport workers. Work restrictions constrain immigrants' access to employment by making it a crime to work without proper documents and imposing fines on companies that hire unauthorized workers . The end of legal employment for undocumented workers, however, enables employers to more easily exploit their labor. For example, Elizabeth Fussell finds that in the aftermath of Hurricane Katrina, when demand for low-skill construction workers to clean and rebuild the damaged city was high, Latinx day laborers were subjected to wage theft by unscrupulous employers who knew they would not report the theft to authorities. Ruth Gomberg-Muñoz similarly describes em-ployers who prefer undocumented workers because of their exploitability. She cites one employer who stated, "if the immigrant employees had papers, they might not be such hard workers" . She also explains how after worksite raids, employers struggled to replace undocumented laborers with those willing to endure the same working conditions. For example, when a poultry-processing plant in Georgia lost two-thirds of its workforce in an immigration raid, local African American workers expressed concerns about dangerous work conditions and questionable labor practices, leading the plant to fill the positions with prison workers, Hmong refugees, and the homeless. The result of criminalizing immigrant labor has resulted in a downward trend in wages and higher unemployment rates for undocumented and documented Mexican immigrants . The second way punitive immigration polices intersect with work and family life is through family separation. Between 2010and 2013, Jodi Berger Cardoso and colleagues estimated that "300,000 parents of U.S. citizen children were deported." As Jennifer Green points out in her study of mixed-status families in an era of mass deportation, mundane routines such as driving to the grocery store or going to work could result in deportation and fathers being separated from their families. Fathers are more likely to work outside the home in dual-headed Mexican immigrant households, putting them at greater risk of deportation . As the primary wage earners, fathers' loss of employment may also shape various aspects of family life for their families. For example, uneven ICE enforcement in some communities may force fathers to seek other employment opportunities, causing families to leave together or remain separated . For Mexican immigrant fathers, workplace restrictions not only threaten forced return to their country of origin, but also separation of the family unit . Although the role of Mexican immigrant fathers within the home is less understood than that of mothers, fathers' experiences with work outside the home may manifestly shape aspects of life inside it. Thus a contribution of the current study is understanding how strict immigration enforcement shapes parenting roles and practices in underresourced, structurally disadvantaged communities. --- Home and Schools As families buffer the distal effects of immigration policy on their communities, schools play a critical role in providing social support services, regardless of a family's documentation status. During the COVID-19 pandemic, 2020 Coronavirus Aid, Relief, and Economic Security Act stimulus payments were not available to undocumented immigrants, lawful permanent residents, and U.S. citizens in immigrant families . In response to the growing numbers of food-insecure families when schools closed due to the pandemic, the U.S. Department of Agriculture announced that it would make meals freely available to all students for the 2021-2022 academic year through the National Food Lunch Program . Schools also address the health and well-being of their students in other ways. For example, school-based health centers, funded through the U.S. Department of Health and Humans Services, provide a range of health-care services to families, including primary medical care, dental and oral care, and health education. To increase vaccination rates for COVID-19, the Centers for Disease Control and Prevention provided guidance to district administrators for hosting vaccination clinics to tackle disparities in vaccination rates. As schools increasingly provide critical supports for low-income and immigrant families, strong family-school relations are central to maximizing students' potential. Parents' involvement in their children's education inside and outside the home is associated with positive learning behaviors, higher test scores, and improved literacy skills . At the same time, building and maintaining strong family-school relations may be especially challenging in lowincome, Mexican immigrant families because of language barriers and parents' worries about their immigration status. Although for many Mexican immigrant families, schools embody the hopes and aspirations for their children's futures , parents' involvement may reflect the broader constraints of their environment. Because research on parenting practices and children's educational success focuses on mothers, fathers' role remains underexamined. The current literature suggests that Mexican immigrant fathers' involvement may be unique because of their multiple jobs, fear of deportation, greater English-language skills, and gender stereotypes about their role inside and outside the home . Drawing on a nationally representative sample of Latinx parents, Veronica Terriquez finds that Latinx fathers' participation in school-based events varies as a function of their immigration status and English-language proficiency: Spanish-dominant fathers and more recent immigrants were less likely to be involved in school-based activities. Additionally, Robert Moreno and Susan Chuang report that although Latinx fathers hold strong beliefs regarding their children's school participation, these beliefs are not reflected in their level of involvement at school. However, qualitative studies have pushed back on the narrative that fathers are uninvolved in their children's education; rather, teachers overlook or misinterpret fathers' participation and provide few avenues for deepening their engagement . Building and maintaining strong familyschool relations is especially important in lowincome, immigrant communities. However, research suggests that linguistically diverse parents are often excluded from their children's education . Although state education agencies and school districts are mandated to support English-language learners and their parents, the U.S. Department of Justice reports numerous compliance issues . Thus, the role of language on the formation of home-school relations among Mexican immigrant households warrants further study. Here we examine how immigration policy structures family life inside and outside the home. --- me Thods This article emerged from an examination of how parents' relationships formed and developed at home, in schools, and within their communities . Parents were selected for interviews from a broader pool of more than three thousand first-grade families and fifty-two Title I schools participating in a cluster-randomized controlled trial that examined the effects of parents' school-based relationships on children's early educational outcomes in low-income, predominantly Latinx schools. Overall results from the randomized controlled trial yielded null effects . To recruit families, we made cold calls from lists of parents at eight study schools who consented to participate in the randomized control trial. We aimed to interview four families in each school. Families in one school in Phoenix were oversampled because of unique circumstances that warranted further exploration. Thus, our final interview sample consisted of fifty-seven parents from thirty-four families. We restrict our sample for this article to thirty parents from seventeen families in which at least one parent was born in Mexico. --- Procedures The article draws on interview data and field notes collected by the first author and another researcher from March 2011 to May 2012. With rare exceptions, all interviews occurred at the family's residence on a weekday evening. Before every interview, the researchers purchased food at a local restaurant and shared the meal with the entire family. The interviews were conducted in English or Spanish, according to parent preference, and lasted between 60 and 150 minutes. Total time spent in each household was between two and four hours. The interviews were recorded, professionally transcribed, and translated when necessary. Across the seventeen immigrant households, thirteen couples were interviewed. Four interviews were with mothers only, one of whom was divorced. When two parents were present, interviews with mothers and fathers were conducted separately and simultaneously to capture parents' individual accounts of their experiences with their children's schooling. Interviews were conducted jointly when only one parent was present. --- Analytic Approach The analysis of the interview data proceeded in multiple stages. First, the second author and a research assistant open-coded all thirty interviews and wrote summaries outlining maternal and paternal involvement. During this process, the first and second authors met weekly to discuss thematic findings within the context of families' broader policy environment. Second, the interviews were indexed based on levels of parental involvement. Third, the second author and a research assistant coded the indexed data using NVivo 11 qualitative software. Analytic codes identified emerging factors that structured parental involvement. For example, after we created an index of parents who described intervening in a school-based issue, themes of distrust and fears of deportation emerged as salient. Last, parents' experiences with various domains of work, family, and community were contextualized based on the differing political environments in Phoenix and San Antonio . --- Findings In ecological systems theory, the mesosystem is where multiple microsystems intersect . In Phoenix, strict immigration enforcement throughout the community intersected with parents' ability to find work and participate in their children's schooling. In contrast to San Antonio, where strict immigration enforcement was largely absent, the effects of these policies reached into the home as parents adapted to the climate of fear and instability that pervaded their community. Despite differing levels of immigration enforcement in San Antonio and Phoenix, inadequate language supports structured Spanishdominant parental school involvement in both cities. In Phoenix, families' experiences with hos-tile immigration policy in their community provided a backdrop for schools' academic performance. In the best performing school in the sample, 26 percent of third-grade students were proficient in math and 32 percent were proficient in reading. In the lowest-performing school, 15 percent of third-grade students were proficient in both math and reading. In San Antonio, performance was marginally better: 56 and 39 percent of third-grade students were proficient in math and reading, respectively; in the lowest-performing school, 24 percent were proficient in math and 22 percent in reading. Drawing on survey data collected from families and schools in the randomized control trial , average parental involvement scale was compiled across ten items. Parents were asked to rate their level of involvement from 1 to 5, where 1 is "Never" and 5 is "11 or more times." Surveys included items such as "I helped my child with homework," "I went to a school program," "I asked my child to tell me about school," and so on. Across the eight schools, average parent involvement was 3.6, suggesting that involvement was neither high nor low. Among Mexican immigrant households in the interview sample, the average was higher . Some families reported scores of 5 across all items; the lowest average was 2.10. Principals were also surveyed and asked the extent to which they agreed or disagreed with the statement "Parents are actively involved in the school's programs," 1 being "strongly disagree" and 4 being "strongly agree." The average across the eight schools was 2.875, suggesting that they mostly agreed that parents were involved at the school. Schools' academic performance and the parental involvement scale provide context for parents' descriptions of how immigration policy served as de facto family, labor, and education policy. Immigration Policy as de Facto Family Policy Bronfenbrenner suggests that the macrosystem represents the cultural context in which families are embedded and serves as a blueprint within a given society. For this study, we consider federal and state-level immigration policy as the blueprint for the way immigrant a All names are parent-selected pseudonyms. b Children living in the home at the time of interview and relative to focal child . families experience life within their communities. At the time of data collection , national deportations were at an all-time high . San Antonio and Phoenix are separated by 982 miles, and the political environment in each city made it feel even farther. In San Antonio, state-level policies did not single out immigrant households and none of the parents expressed fears of deportation. However, in Phoenix, billboards and radio announcements promoting Arpaio's "illegal alien hotline" asked community members to call and report the location of suspected undocumented immigrants. Arizona's attitude toward immigrants was clear: almost every parent we interviewed described a personal experience with deportation, whether family members, friends, or neighbors. SB 1070 was part of Arizona's broader policy effort targeting Mexican immigrants. Local law enforcement frequently conducted status checks without cause. One consequence of the measure was that parents were constantly afraid of deportation because failure to carry proof of residency was cause for arrest. As an undocumented mother in Phoenix said, "Since the law was established, in almost every community, everybody is afraid of it. Many people were deported" . Another father explained how even social relationships with undocumented immigrants could result in arrest: "Let's say you want to get water, you go to the store, and the police stop you, and you tell him, 'Here is my license' You show your I.D., and they say, 'Arizona I.D.' If you do not have it, who has the problem? I do, because if I have a person in my car who does not have an I.D., that is breaking the law here, and you could go to jail" . In Phoenix, parents frequently discussed how SB 1070 created a climate of fear within their communities. At the time of the study, the policy context in Phoenix was distinct from that in San Antonio. Unlike Sheriff Joe Arpaio's Maricopa County, Bexar County did not have a 287 partnership . One father from Phoenix who had recently traveled through San Antonio described the differences between the cities: "A year ago, I passed by San Antonio. You don't see immigration [ICE officials] everywhere. In Arizona, you drive a little bit, and wow. Here you drive, and before you hit the road, immigration stops you, 'Let me see your I.D.' There you drive, and they are nowhere to be found. They used to say that the law was much stricter in Texas. To tell you the truth, the law is stricter here [in Phoenix]" . To better understand this contrast, we asked parents in San Antonio about the news of mass deportations in Phoenix. One immigrant father responded, "Texans are calmer and more tranquil. I know people who have been expelled, but not because they are Mexican, because of drugs or something like that. I have not been to Phoenix, but it seems more difficult. Here in San Antonio, it's good, you can go wherever you want at any time, and it is okay" . Relative to their counterparts in Maricopa County, immigrant families in Bexar County mentioned fewer interactions with law enforcement. Follow-up survey data from the randomized control trial revealed that four of the twelve Phoenix families had moved and dropped out of the study relative to one of five families from San Antonio. In Phoenix and San Antonio, immigration policy functioned as de facto family policy by structuring families' experiences in their communities. Fears of deportation, social distrust, and worries of family separation restructured the lives of many families who had called Phoenix home. --- Immigration Policy as de Facto Labor Policy In ecological systems theory, the mesosystem is where multiple microsystems intersect . Strict immigration enforcement in Phoenix targeted immigrant families by restricting fathers' access to work. Unlike in San Antonio, immigration policy punished employers and employees alike. As the economy faltered, families left the state in search of income for their family. In Arizona, SB 1070 made it illegal to work without proper documentation and penalized businesses that knowingly hired unauthorized workers by suspending or revoking their business license. Parents suggested that SB 1070 caused many workers to lose their jobs. Workplace raids were a defining feature of Sheriff l o W -i n c o m e Fa m i l i e s i n t h e t W e n t y-F i r s t c e n t u r y r s f : t h e r u s s e l l s a g e f o u n d a t i o n j o u r n a l o f t h e s o c i a l s c i e n c e s Arpaio's regime. As one immigrant father explained, "He [Arpaio] does his raids, he goes to big businesses where many people are working. They just arrive without letting anybody know. They arrive and take people out, people who have worked there for many years but do not have documents. They are working like many people who come from Mexico" . Employers were also required to use E-Verify, a federal website from the U.S. Department of Homeland Security and the Social Security Administration that would make it more difficult for workers to submit fake social security numbers. By comparing workers' employment authorization forms with federal databases, the system can verify employees' identification photos by matching the one in the system . These work restrictions were consequential for undocumented immigrant parents. One father explained: There are jobs, but you can't apply without a social security number. I have a cousin who does not work and lives in Mexico, so I told him, 'Let me borrow your social security number. I will work, and you do the taxes.' Before, you were able to do it, but now you can't. The thing is that they have E-Verify or something like that. So, when they started all of that [enforcement], we [those without valid work permits] could not do anything about it. Many people lost their jobs for the same reason, they did not have work permits. Another father described how E-Verify not only affected workers but also their employers: "Before, a business could get a job and finish it in a few months. Why? Because there were people to work. Now, they have to hire documented workers, but where are they? They are not around anymore. There are still some without papers, but if the sheriff shows up [at the worksite], he fines the company and takes all the Mexicans that do not have papers. Legal papers do not make a worker. Papers do not do jobs. Jobs are done by people who want to work" . As a result, employers continued to hire undocumented workers despite the potential consequences. We asked one immigrant mother, who worked at a recycling plant, whether her employer was aware of her status. She responded, "The boss knows everything. He knows we have someone else's papers, and he said that he would pay the fine for all of us. He is making a lot of money; why not?" . As parents experienced limited work opportunities and the increased policing of job sites, Arpaio's workplace raids became synonymous with immigration policy. Fear of deportation at work was a consistent theme in the data and restructured the lives of many families who had called Phoenix home for more than a decade. Community instability ensued as families left the state to look for work. One mother explained: "You had to be legal, or you would not survive in this place. The problem was work. So, the laws affected the employers because they were afraid of hiring people without papers. People say, 'You can live here,' but if you do not have a job, you cannot pay for anything. The first thing you lose when you do not work is the ability to pay rent and utilities, so you have to move" . The decision to leave Phoenix, however, was not an easy one. An undocumented immigrant mother explained the dilemma: "Many people moved away when they lost their jobs. My husband didn't lose his job, but we would not be here if he had lost it. But for many people, their children were born here and have lived here for more than 16 years; they had to go back [to Mexico] because they lost their home" . Although families had raised their children and bought homes in the community, as families decided to leave the state searching for work opportunities, long-standing community relationships fractured. Decisions to leave Arizona were incredibly complicated in mixed-status homes where one household member was undocumented. Parents discussed contingency plans with friends and family if they were picked up in a raid. One documented father explained: "Many families have made me in charge of their children. If, one day, something happens-if they get deported-someone will bring their children to me. Otherwise, the government takes them and gives them up for adoption. Can you imagine how sad it would be that another family adopted your children because you did not have papers and were deported to Mexico?" . In contrast with San Antonio, where strict immigration enforcement was largely absent, 287 partnerships and Senate Bill 1070 led to fears of deportation, social distrust, and worries of family separation. Labor shortages caused by employees moving or being deported presented employers a choice: either follow the law and their businesses might suffer or break the law and pay a fine. For fathers able to maintain a source of income, families adapted to the uncertainty of deportation as punitive immigration policies reverberated through Mexican immigrant communities, regardless of documentation status. --- Immigration Policy as de Facto Education Policy Bronfenbrenner describes the critical role of microsystems like the household in affecting children's development. In the context of Mexican immigrant households, we considered how macro-and meso-level factors influence mothers' and fathers' role within the home. In Phoenix, strict enforcement at parents' workplaces affected key domains of their children's lives, suggesting that immigration policy serves as de facto education policy. Parents' fears of deportation and family separation were central in their children's lives. As one father described it, children were "worried they would not see their fathers again" . As parents left the community to protect their children from the broader hostile context in which families lived, instability in schools directly ensued. Parents reported that school enrollment dropped dramatically. One mother explained: "[The school] sent me an application asking if I planned to return the following school year. I said no but because I was moving [to a new school], but many wrote no because they were returning to Mexico. In the last month of school, many students left. We were getting papers from the school telling us how we needed to talk to our children so the change would not be so hard" . When we asked her about the impact of SB 1070 on home-school relations, she replied, "[SB 1070] affects the family's mood and relationship with the school because children saw that their par-ents were worried. Parents were worried because they did not have jobs; spouses were fighting. At the same time, the children were afraid because of what they heard on TV. They were going to school crying, and they were fearful of going to school. That affected our children." Although many children of immigrant parents are U.S. citizens, parents reported that the emotional toll of deportation was often felt in schools. Parents who remained in Arizona were distrustful of their children's schools and worried that schools might report their documentation status to law enforcement. One mother described this distrust: "Right now, we heard that parents with kids starting school this year would need to present legal documents. So, there will be another exodus of people leaving the state. I remember that at the beginning [when SB 1070 passed], I did not want to go to school. My children were born here, but the school never told us, 'Calm down, nothing is going to happen to you.' Now, the excuse they are saying [when registering children for school] is that they want to verify the address, but they are saying we need legal documentation" . Due to the hostile policy environment in the community, the task of registering children for school, a standard process to ensure that children live within the catchment zone, caused this mother to believe her children's school had ulterior motives. She suggests that schools, rather than addressing her concerns about documentation status, did little to support immigrant families. In addition, Proposition 203, a ballot initiative passed in the early 2000s, mandated English-only instruction . One mother described the impact of this law on students: "Both of my children were in the Head Start program. At that time, it was a bilingual program, but in Arizona, they approved a law where they only teach English in schools. Before, it was not like that. I think it is bad for the children who are just arriving from Mexico or any other country because they start school not understanding anything. It is also bad because they do not speak to [the students] in Spanish, but [the students] do not understand English" . Despite recent efforts to overturn the law in response to declining educational outcomes for English-language learners, Arizona remains the only state in the country with educational language mandates . To better understand the context of parents' experiences, the first author visited an ELL classroom in Phoenix to observe parent-teacher conferences. In this district, parent-teacher conferences occurred in groups rather than one-on-one meetings. The author met the school interpreter informally before the conference and discovered that she was from the Midwest and neither a native-Spanish speaker nor a trained interpreter. Once the conference began, the teacher presented the students' academic progress to parents on a graph. She explained that although the grade standard was ninety to one hundred words a minute by the end of the year, some students were reading only six to eight words a minute. The teacher spoke for several minutes before looking to the interpreter to highlight the main points in a few seconds. At the end of the presentation, the teacher asked whether there were any questions. Through the interpreter, one parent asked whether the teacher could send reading homework in Spanish so that she could help her child at home. Exasperated by the question, the teacher exclaimed, "No, that would be way too much work!" The interpreter translated these remarks back to parents: "No, that is not possible." Observations of parent-teacher relations and interviews with parents in this Phoenix classroom suggest that parental involvement was structured by language barriers as well as the teacher's prejudice against Spanishdominant parents. One bilingual mother described how this teacher had reprimanded her son for being late because of a doctor's appointment. When the child knew he would be late the following day because they needed to pick up his prescription, he told his mother he was afraid his teacher would yell at him again. The mother assured her son that she would talk to the teacher. She recounted her exchange: "I went, and I told her, if you have questions or if you want to know exactly what is going on, you have my number, or you can have the office call me. Her response was, 'well, now, it's nice to know that you speak English and that I can communicate with you.'" When we followed up by asking how she felt about the teacher's comments, she explained that the interaction made her feel "horrible" because "I have an advantage that most parents do not have just because I know the language." SB 1070 and Proposition 203 structured home-school relations in Phoenix, but parents in both cities reported that language barriers shaped their involvement in their children's educations. Guidance from the U.S. Departments of Justice and Education on the provision of translators and interpreters requires that "SEAs [state educational agencies] and school districts must provide language assistance to LEP [limited English proficiency] parents" and that "school districts should ensure that interpreters and translators are trained on the role of an interpreter and translator" . However, parents viewed interpreters at the school as ineffective and unreliable. One mother shared her concerns: "when you are speaking through an interpreter, they do not tell the teachers the same things we say, they [the interpreters] fix it up a little bit" . Underscoring this observation, one father noted that school translators would "only repeat what you say like a machine" . Although the extent to which greater parental involvement in their children's education would have improved students' educational performance is unclear, immigrant parents' autonomy to decide their level of involvement was removed when schools did not provide adequate language support. In our discussions with parents about how relationships formed and developed in their communities, mothers frequently discussed how language barriers structured home-school relations. Through our interviews, we learned that fathers' English-language skills, often learned from work outside the home, were leveraged by mothers to help with educational responsibilities. Primarily, mothers reported that fathers' language abilities facilitated homeschool relations. For example, one mother explained how she had grown accustomed to not communicating with her son's teachers: "every year when my son begins school, I ask him, 'does your teacher speak Spanish?' 'No, she does not speak Spanish?' It is better if I stay quiet and tell my husband that he has to go and get involved" . Similarly, another mother reported that she had strong relationships with a translator at school, which lessened the need for her husband's Englishlanguage skills. Exacerbating issues with translators, parents also reported concerns about the effectiveness of their children's teachers and the inadequacy of their children's schools. When his child was at risk of being held back a grade, one father explained that his wife "did not want to deal with the teacher" and asked that he intervene on her behalf. Fathers also intervened in school matters of their own accord. One father described his dismay when he reviewed his son's workbook and found several incomplete pages. After consulting with relatives who had children in the same school, the father confronted the teacher about the unfinished work and ultimately decided to enroll his son in a charter school. Fathers also helped when homework was sent home in English. One mother explained how her husband was always helping their daughter: "Mario speaks English, so if it is not something that I have to read, I can help her, but if she has to read, and it requires too much English, I tell [my husband], 'Mario, she needs your help.' He is always helping her. That is why I don't have a problem. If he did not speak English, then that would be a problem" . Similarly, after coming across a problem that he thought was too difficult for his sevenyear-old son, another father described marking "a large circle around it, so the teacher would see it," noting, "if [a parent] doesn't speak English, they won't notice [the difficulty of the assignment]" . As parents viewed interpreters at the school as ineffective and unreliable, fathers were a vital family resource given the schools' inability to support Spanishdominant parents' linguistic needs. For families in Phoenix and San Antonio, mothers and fathers worked together to ensure their children's basic educational needs were met. Fathers' language skills facilitated homeschool relations and were a key resource for their wives and children given schools' failure to support Spanish-dominant parents and their children. Fathers intervened when educational issues arose, sent correspondences to teachers, and helped with homework. In Phoenix, strict immigration enforcement intersected with children's home environment. Excluding Spanish-speaking parents from full participation in their children's schooling limited parents' abilities to support their children as they struggled emotionally and academically. Other than sending notes home asking parents to discuss the impacts of immigration on their children, the ambiguity around deportation risks structured home-school relations: parents distrusted teachers and feared that schools would expose their undocumented status. As families withdrew their children en masse, schools were unaware of the dangers families faced in their daily lives. Schools represent essential sites of social mobility and social resources for low-income adolescents and their families. However, immigration policy can become de facto education policy for children attending schools that fail to buffer the effects of the punitive immigration environment and language barriers. --- disCus sion a nd ConClusion In this article, we examine how strict immigration enforcement policies shaped work and family life inside and outside the home in San Antonio, Texas, and Phoenix. Drawing on interview data, we highlight how punitive federal and state immigration policies structure community relations through work and school and how these broader influences reach into the home. This analysis provides insights into the intersections between work and school in an age of increased political hostility toward immigrant families. Mexican immigrant fathers' role in family life has been overlooked, and these findings shed light on the unique ways in which father's work contributed to children's development both inside and outside the home. Our findings point to several broad conclusions. First, coordination among federal, state, and local law enforcement officials suggests that immigration policy serves as de facto family policy, structuring families' experiences in their communities. Although immigration enforcement has long been under the federal government's purview, 287 program partnerships have caused state-level immigration enforcement to change dramatically under different presidential administrations. During the Barack Obama administration, Maricopa County Sheriff Joe Arpaio's name became synonymous with worksite immigration raids and deportations for families in Phoenix. The vestiges of anti-immigrant rhetoric and punitive policies in Phoenix during our study can be seen in other state-level polices, such as Alabama's HB 56 . Although families reported vastly different immigration contexts in San Antonio than in Phoenix, Texas has adopted a similarly punitive approach in more recent years. After Donald Trump's election to the presidency in 2016, Texas enacted Senate Bill 4, making it a crime for local officials or public colleges and universities to serve as "sanctuary cities" or refuse to work with ICE . During the Joe Biden administration, Texas has also fought against executive action to end workplace immigration raids and deportation of undocumented immigrants by suing the federal government for not enforcing federal immigration law. 2 Second, our work shows how immigration policy functioned as labor policy in Arizona. Immigration policy limited work opportunities, fomented social distrust, and exacerbated deportation fears. Work authorization documents were more stringently checked, and worksite immigration raids were a key enforcement mechanism. Experiences with deportation threats and limited work opportunities were largely experienced through fathers. As fathers sought other work opportunities and feared deportation, communities fractured. For those who stayed in Phoenix, commonplace activities like registering for school heightened families' deportation fears. Third, our data show how immigration also functioned as de facto education policy. Although immigration policies did not explicitly target children, children were not spared from its effects. In Phoenix, parents described vanishing classmates and children crying in school because they feared losing their fathers. In both cities, parents' efforts to ensure their children's basic educational needs were met were constrained by inadequate language supports. Parallel to Sarah Halpern-Meekin and Adam Talkington's findings , our results show that families responded to these constraints by relying on fathers to play prominent roles at home and in schools. Although advancing federal immigration reform should be a priority, educational policies needs to focus on improving trust and communication between parents and schools rather than punishing children for their parents' immigration status . Results also suggest school districts must enforce adherence to mandated language access policies. To hold schools accountable, districts should provide parents with information to contact the Office of Civil Rights at the Department of Education to report violations. Districts would, of course, need to clarify that reporting violations would have no impact on parents, whether documented or undocumented. That forces beyond their desired level of involvement shaped immigrant parents' participation in schools have implications beyond their children's education. Although the active enforcement of strict immigration policies has subsided under the Biden administration, their effects are likely to persist for generations. Scholars have recently called for expanding the adverse childhood experiences framework, which considers how childhood trauma shapes later life outcomes, to include the threat of deportation and deprivation resulting from strict immigration enforcement . Although moving to another state may have offered temporary protection from deportation threats, as the editors note in their introduction to this issue, low-income workers are suffering on a national scale: job opportunities for low-skilled workers have diminished, laborforce participation has declined for those with low educational attainment, and wages have stagnated . Finally, our study is not without limitations. Although our data succinctly capture the experiences of Mexican immigrant households from mothers' and fathers' perspectives, we cannot generalize to other populations or other communities. Policies identified in the study are not exhaustive, and factors not mentioned by parents may also have shaped their involvement. In addition, most of the families we interviewed were dual-headed, a typical household structure among working-class and poor Mexican immigrant families, but atypical for most low-income families, which tend to be single-headed households. Thus future research should consider the experiences of single-headed Mexican immigrant households.

Brown University. Elizabeth Peck is a social determinants of health researcher at Transcendent Endeavors, a Small Business Innovation Research (SBIR) startup.

Introduction People living in low-income countries and poor people living in affluent countries tend to suffer disproportionately from environmental disasters. The last two decades saw over 7,000 major environmental disasters that caused trillions of dollars in damage and killed more than 1.35 million people worldwide [1]. In this same time period, more than three times the number of people died per disaster in low-income countries than in high-income countries [1]. Even within countries that are more affluent and experience fewer disasters, the impacts of those disasters that do occur can be strikingly unequal. For instance, when Hurricane Katrina struck New Orleans, the impacts to life and property were disproportionately borne by the African American community-damaged areas comprised 46% Black people versus 26% in undamaged areas, [2] and 84% of missing people were Black in a city that is only 68% Black [3]. When it comes to disasters, differences in vulnerability can affect the magnitude and duration of impacts like the loss of property, livelihoods, or services. A widespread understanding of environmental disasters is that they are beyond human control-hurricanes occur to unsuspecting coastal communities, earthquakes strike without warning, and so on. In recent decades however, scholars from a range of disciplines have argued that natural disasters are not "natural" [4,5]. Rather, it is the social, political, and economic context that makes an environmental hazard become a disaster [6,7]. Just as a hurricane can leave one city in ruin for years to come while a similar storm leaves another city unscathed, so too do various human factors generate unequal exposure and susceptibility to wildfires. For example, a strategy to aggressively suppress wildfires near homes promotes increased intensity of wildfires in the future due to fuel accumulation [8][9][10]. At a household level, families without financial means cannot afford tree trimming, brush removal, or other fire mitigation services that could mean the difference between a low severity under-burn and a severe wildfire [11]. Additionally, families who rent are ineligible for much of the federal assistance available to homeowners for rebuilding after a fire event. [12]. In these ways, a sole focus on biophysical wildfire hazards like fuel and weather conceals the root causes that turn fire, a natural process, into a disaster. Geographic segregation of fire-prone places in the U.S. does not operate identically to other hazard-prone areas around the world. In developing countries, poor and ethnic minorities are more likely to live in hazardous areas, like flood-prone farmland in India or volcanically-active slopes in Guatemala [13,14]. This is not so in the U.S., where incentives like response aid, environmental amenities, and wildfire insurance facilitate the settlement of economically advantaged groups in fire-prone landscapes, oftentimes in second homes [14,15]. This seems to complicate traditional disaster narratives associating place-based hazards with poverty, but the important point is not just that socially vulnerable populations may be more exposed to environmental hazards, but that those hazards become disasters specifically when they affect vulnerable populations. Fire is a basic, necessary, and unavoidable component of many landscapes and underlies the delivery of a number of key ecosystem services in forests and rangelands around the globe [16]. Even so, wildfires since 1984 have affected nearly 6 million people, directly caused over 1,900 deaths, and generated more than $52 billion in economic costs [17]. While fire-prone places in the U.S. are more likely to be populated by higher-income groups, this fact threatens to overshadow the thousands of low-income individuals who also live in fire-prone places but lack the resources to prepare or recover from fire [15]. In California, for example, many individuals in rural areas, low-income neighborhoods, and immigrant communities do not have access to the resources necessary to pay for insurance, rebuilding, or continual investment in fire safety, thereby increasing their vulnerability to wildfire [18]. These disparities became very clear after the 2017 wildfires in Sonoma County, California, where price gouging on rentals worsened an already dire housing shortage [19]. Conceptualizing wildfire disasters as the product of the complete social-ecological system can provide novel insights into management. By understanding who is vulnerable and why, management can expand beyond technical fixes to socioeconomic and political solutions. Here we use a vulnerability assessment framework to identify places across the coterminous United States that are disproportionately imperiled by wildfire. This study can be seen as building on wildfire vulnerability scholarship by Wigtil et al. , Collins , and others while explicitly exploring the relationships between race, geography, and wildfire from an environmental justice perspective [14,15]. --- Methods --- Overview of approach We use a vulnerability assessment framework to identify census tracts that are disproportionately imperiled by wildfire . We define wildfire vulnerability as a combination of the exposure of social-ecological systems to a hazard, such as wildfire, and the adaptive capacity of a place to absorb, recover, and modify exposure to the hazard [6,20]. This framework captures both the potential of wildfire as well as the socioeconomic adaptive capacity of census tracts exposed to wildfire. --- Estimating U.S. fire vulnerability Wildfire hazard potential. We analyzed the potential of wildfire and adaptive capacity for each of the 71,901 U.S. census tracts . The United States Forest Service estimated Wildfire Hazard Potential for the continental U.S [21]. The WHP is a 270-meter resolution raster product derived from the Large Fire Simulator [22], incorporating fuels, vegetation, weather, historical fire occurrence, and resistance to control data to estimate the probability that an area would experience extreme fire behavior under conducive weather conditions [21]. The WHP values are binned such that two-thirds of U.S. land area is classified as having low or very low wildfire potential and only one-third classified as having moderate, high, or very high wildfire potential. We rescaled WHP in this analysis from 0-1 to match the adaptive capacity index , with 0 being unburnable land and 1 being very high potential for wildfire that is difficult to control. Higher WHP values correspond to higher potential for more extreme wildfires. We then estimated the WHP for each census tract as the average of underlying WHP raster cells. Vulnerability to fire requires at least a minimal exposure to wildfire, so for the purpose of this analysis we constrained our estimates to the 6,304 census tracts that have at least moderate fire potential. This effectively eliminates the most urbanized areas that lack fuels and other conditions necessary for wildfire. Census tract adaptive capacity. We next estimated the adaptive capacity of each census tract using data from the 2014 U.S. Census American Community Survey. Census data are an imperfect but commonly employed proxy for communities and neighborhoods. They are imperfect because neighborhoods and communities are socially-constructed based on both shared meaning and geography which are difficult to capture [23]. Despite this, census data are ubiquitous in the social science literature because of their coverage and availability [23]. We employ census tracts as the unit of analysis in this study because they are available at the national level and are groupings at which wildfire policy and management are often prescribed and implemented. With that in mind, we acknowledge that local results may differ with alternative geographies. By adaptive capacity, we are referring to the ability of a census tract to absorb and adjust to disturbances, like wildfire, while minimizing damage to life, property, and services [6,24]. The ability to adapt to hazards is influenced by a number of social and demographic factors, including age, income, the strength of social networks, and neighborhood characteristics [25]. For example, economically disadvantaged families, the elderly, disabled people, and residents of high-rise apartments or mobile homes tend to be less adaptable to hazards [26]. The least adaptable groups are likely those whose needs are insufficiently considered in the planning of local response and relief organizations [27]. Following Flanagan and colleagues, we created an adaptive capacity index based on 13 metrics from the U.S. Census American Community Survey [25]. The index was comprised of four domains: socioeconomic status, language and education, demographics, and housing and transportation . Socioeconomic status was characterized with three metrics: Persons below poverty level, the number of people unemployed, and per capita income. Poor households often cannot afford to pay for fire mitigation services like tree cutting and removal of fine fuels [11]. While wealthier residents may be disinclined to remove environmental amenities like trees, they are more likely to have fire insurance and the community firefighting resources needed to extinguish a fire [11,28]. Additionally, lower income households face more obstacles rebuilding or finding new housing after a fire [29]. Higher educational level also appears to improve adaptive capacity. In general, education improves access to relevant information, enlarging social networks that can facilitate recovery, and aiding in the navigation of bureaucratic hurdles [25,30]. Limited proficiency in English has also been linked to difficulty recovering from disasters [25,26]. Housing quality and transportation both covary with wealth, with economically disadvantaged people often living in poorly constructed housing or mobile homes [31]. Multi-unit housing poses an increased risk to health as escape routes can be overcrowded and building-owners are less likely to pursue fire mitigation on their properties [18,32]. Additionally, renters are eligible for less federal housing assistance than homeowners [12]. Finally, transportation out of an evacuation zone may be challenging for those without access to a vehicle, and for some, fuel costs may prevent vehicle use [33]. To construct the adaptive capacity index, we adopted the framework of Flanagan and colleagues [25]. Briefly, we ranked all 71,901 census tracts across each of the 13 indicators such that each census tract had 13 rankings from 1 to 71,901. We then converted these ranks to percentile ranks from 1 to 100, summed across the 13 values for each census tract, and then calculated the percent rank of those sums. The result is an index from 0-1 for each census tract, with 0 being those with the greatest capacity to adapt to a wildfire and 1 being those with the least. That lower numbers of the index correspond to greater adaptive capacity is unintuitive, but this was done because the census variables we use capture inherently negative properties that worsen as the derived index increases. To combine this intuitively with WHP, which denotes greater fire potential at higher values, we retain this scale so that a higher number denotes a census tract with lower adaptive capacity. Estimating census tract vulnerability. In the risk assessment literature, wildfire vulnerability has been defined as the combination of wildfire likelihood and intensity in a particular place with the propensity of that place to experience a change in value [34]. This is a purely biophysical perspective of vulnerability which does not incorporate the socioeconomic factors responsible for turning wildfire events into wildfire disasters. Here, we adopt a definition of vulnerability from the disaster and hazard literature that also incorporates the social characteristics of affected inhabitants [35]. Following this scholarship, we define the relative vulnerability V to wildfire of a census tract as the Euclidean distance of the census tract to the graphical origin in a space defined by the WHP and adaptive capacity index [36], or V ¼ ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ðAC À AC min Þ 2 þ ðWHP À WHP min Þ 2 q Under this framework, the WHP and adaptive capacity received equal weight and the vulnerability of a census tract increases with distance from the origin of the plot. --- Analysis We examined the relationship between vulnerability to wildfire and ethnicity in census tracts using quantile regression. Simple linear regression neglects change in the relationship at the edges of the distribution. This is particularly important in situations where some factor acts as a constraint on a variable. In this case, the estimated effects of a factor will not be well represented by changes in the mean of the response variable distribution. Because a number of historical, economic, and social issues may constrain the relationship between wildfire vulnerability and ethnicity we used quantile regression from the R package quantreg [37] to explore race and wildfire vulnerability . Quantile regression performs regression on subsets of the data divided at different levels and estimates the conditional quantile Q τ given the predictor variable. Thus for two quantiles, τ = 0.05, 0.95, we estimate two equations for each racial/ethnic group of the standard form Q t ðx i jWHP i Þ ¼ b t 0 þ b t 1 x i where x i is the population of a racial/ethnic group i in a census tract. The regression yields the 95 th and 5 th quantile weights for racial and ethnic populations given a wildfire vulnerability in a census tract. All spatial visualizations and analyses were conducted with ArcGIS v10.4 and R v3.3.2. --- Results The potential for high-intensity wildfires is spatially heterogeneous, with higher potential in the U.S. west and southeast . However, by incorporating adaptive capacity into our estimation of the vulnerability of communities to wildfire we generate a different perspective of the threat of wildfire. For instance, if we consider only WHP, the Southeastern U.S. generally exhibits moderate scores, with few places having very high potential for extreme wildfires. In contrast, when we consider the threat of wildfire from a social-ecological perspective, the Southeastern U.S. stands out as a region of high vulnerability. At a smaller-scale, similar shifts are evident: affluent exurban regions east of the San Francisco Bay and rural areas of the eastern Sierra Nevada Mountains in California have similar wildfire potential, but relatively poorer socioeconomic conditions in the Sierra Nevada Mountains make those communities far more vulnerable to fire disaster than their exurban counterparts. These areas are labeled in both Figs 2A and3. When we examine the association between wildfire vulnerability and race, quantile regression analyses reveal that some ethnic groups experience very different vulnerability to wildfire than majority-white communities . In particular, the minimum vulnerability to wildfire experienced by communities increases as the proportion of Native Americans and Blacks increases . A similar trend occurs in Hispanic communities. In contrast, as the proportion of Whites and Asians/Pacific Islanders increases, the minimum wildfire vulnerability that these communities experience declines. Consequently, census tracts experience an upwards compression of the minimum possible vulnerability score as they become more Hispanic, Black, or Native American, while the opposite is true as the proportion of Whites or Asians/Pacific Islanders increases. We divided our estimates of wildfire vulnerability shown in All non-white populations are underrepresented in the most secure census tracts. This quadrant , with moderate fire potential but a low adaptive capacity and thus greater ability to respond to wildfires, is the most populous in our analysis with nearly 12.1 million Americans. Nearly 1.4 million people live in communities that are more prone to fires but have a similarly low adaptive capacity . While Blacks do not tend to live in the areas with the highest fire potential, they are overrepresented in communities somewhat prone to wildfire but that would likely not respond or adapt well if one were to occur . Strikingly, Native Americans are highly overrepresented in all of the most vulnerable areas, especially communities with extreme fire potential but high adaptive capacity where they are nearly six times more likely to live than is expected . --- Discussion Nearly 29 million Americans live in census tracts with a moderate to very high potential for high-intensity wildfires. White Americans, who constitute 72% of the United States, make up 76% of these census tracts and the population is not, for the most part, socially vulnerable. This is congruent with other research that suggests that environmental amenities and fire insurance facilitate the settlement of more advantaged families in these areas [14,15]. However, an emphasis on only the fringes of fire hazard neglects the 12.4 million people living in census tracts with poor adaptive capacity and lower, but still significant, potential for wildfires. We argue that inhabitants with the lowest adaptability even in these moderately fire-prone landscapes are particularly vulnerable to wildfire and should be central to our understanding of fire disasters. There is some correlation between social vulnerability and race/ethnicity , and lower real-estate prices in some fire-prone areas may help explain higher numbers of socially vulnerable populations in these locations [38,39]. But it is not just economics-historical patterns of settlement, displacement, and migration have led to the spatial distribution of racial and ethnic groups in the U.S. that we see today [40]. For example, a likely constraint on the upward compression of wildfire vulnerability for Native Americans is their historical forced concentration on federal Indian reservations. Census tribal tracts in these reservations, particularly in the western U.S., have higher WHP scores on average than non-reservation census Differences between the spatial pattern of wildfire hazard potential and vulnerability are due, in part, to the relative scarcity of locations that have high or very high wildfire potentialonly 2% of census tracts have a WHP � 3. Consequently, the inclusion of socioeconomic, education, housing, and transportation metrics results in the elevation of locations with relatively moderate wildfire potential, but high adaptive deficit. It is also worth noting that by constraining our analysis to only census tracts with moderate to very high WHP, we largely excluded urban areas. Thus, our subset contains less than 7% of the U.S. Black population but over 20% of the U.S. Native American population-groups that, for historical reasons, dwell primarily in urban and rural areas, respectively. A presently unstudied question is how wildfire vulnerability will change as the geography of poverty shifts from cities with low wildfire potential to suburbs and exurbs that border the wildland-urban interface [41]. Emergency planning and mitigation strategies must be tailored to the diverse populations affected by fire, yet the engagement of socially vulnerable groups, particularly of non-white Americans, is still quite limited [42]. Indeed, in 2014, as a massive fire emerged in eastern Washington, language barriers prevented Hispanic farm-workers from receiving evacuation notification from authorities, and the only Spanish radio station in this region never received the emergency information [43]. Similarly, emergency departments and radio stations in Regression lines are calculated at the 5 th and 95 th vulnerability quantiles, respectively. Vulnerability coefficients of variation for census tracts with the lowest/highest proportions of a particular race/ethnicity are noted below each plots. Race "Other" includes those who identify as two or more races or a race not listed on the census form. As the proportion of Hispanics, Native Americans, Blacks, or Other increases in a census tract, the range of possible vulnerability scores for that tract becomes restricted to high values. Strikingly, vulnerability actually decreases as the proportion of White and Asian/Pacific Islanders increases. https://doi.org/10.1371/journal.pone.0205825.g004 Northern California and Santa Barbara struggled to release timely and correct bilingual information during the 2017 wildfires [44]. While this study does not directly assess multilingual evacuation warnings across the U.S., the vulnerability index does incorporate English-speaking ability and ethnicity. These results help make the case that correctly translating and effectively disseminating preparedness and evacuation materials is a prerequisite for equitably mitigating wildfire vulnerability. Cultural, historical or political experiences differ amongst racial and ethnic populations and these, in turn, also affect preferences for fire management of different communities. For example, in Washington, when the U.S. National Guard was deployed to provide emergency assistance, undocumented migrant farmworkers viewed them as government authorities and threats rather than as trusted helpers and messengers [45]. Similarly, lower than average trust in government by black Americans may underlie substantially greater reluctance towards fire mitigation practices than exhibited by white communities [46]. In contrast, many Native American cultures used fire to maintain or even enhance the value of landscapes. Indeed, Values are expressed as the ratio between the population of each race/ethnicity that lives in a quadrant and the total U.S. population of that race/ethnicity. The dotted red line corresponds to the expected share of the population if all groups were equally distributed across the United States. Native Americans are highly overrepresented in the most vulnerable areas while all non-white populations are underrepresented in the least vulnerable areas . In general, white Americans live in areas that are prone to wildfire in greater numbers than would be expected, but these communities are often resilient and better able to respond and adapt to fires. https://doi.org/10.1371/journal.pone.0205825.g005 many culturally important species require low-intensity fires, while recent severe wildfires resulting from suppression have had a detrimental impact on Native American cultural attributes [8]. We found that Whites and Asian/Pacific Islanders are less likely to live in the census tracts most vulnerable to wildfire. While the number of White residents in a census tract is correlated with WHP, it is negatively correlated with the adaptive capacity index, suggesting White Americans live in areas with better adaptive capacity . In other words, the Whites living in the WUI tend to have fewer of the marks of vulnerability than other ethnic groups living there. Asian/Pacific Islanders tend to live in neither fire-prone places nor places with a high adaptive deficit, settling instead in metropolitan areas or more affluent suburbs. However, this is a coarse census category with much intra-group variation, so it is possible that less affluent ethnic groups like Southeast Asians may be more vulnerable to wildfires occurring in places that both border the WUI and contain large Asian/Pacific Islander populations, like Southern California. Environmental justice issues tied to wildfire extend beyond property destruction and loss of livelihood. For example, smoke from wildfires has the potential to cause a significant health burden on nearby populations. A handful of studies on smoke inhalation from wildfires in Australia found divergent health outcomes between Aboriginal and non-Aboriginal populations, but there is little to no research on such racial disparities in the U.S. While this is an important topic, the wide geographic range of smoke exposure would likely require a different approach than our method of using census tracts or other small geographies [47]. Another question for further research is how we might increase the adaptive capacity of communities vulnerable to wildfire. Educational programs exist to create more fire-adapted communities, and some county and state agencies have cost-sharing programs to assist homeowners with reducing fuels on their properties. Whether these programs are reducing wildfire potential in the most vulnerable communities is an important question; one study found limited involvement of socially vulnerable populations in federal programs in Arizona, but more research remains to be done in other communities [42]. Wildfire disasters, which disproportionately disrupt the lives of the most socioeconomically disadvantaged, are as much products of social circumstances as they are ecological ones. The number of wildfires that are difficult to suppress is likely to increase as the climate becomes warmer and drier, wildland-urban interfaces see more development, and fuels continue to accumulate [48]. Therefore, wildfire management and alleviation of the factors that influence social vulnerability must be pursued in tandem to reduce the vulnerability communities to wildfires. Embracing a social-ecological perspective of fire-prone landscapes requires more than just accepting that fires will occur [35]-it forces us to consider variability in the capacity of communities to recovery from disturbance [49], cultural differences and experiences with wildfire [50], and disparate histories of exposure to wildfire. Approaching wildfire adaptation from this social-ecological perspective is a first step in creating safer, just, and more resilient communities. --- All relevant data are within the paper and its Supporting Information files. --- Supporting information --- S4 Table. Slope and standard error of quantile regression coefficients at 5th and 95th quantiles. A greater β at the 0.5 vulnerability quantile indicates that increases in the share of Hispanics, Blacks, Native Americans, and Other races is associated with a greater increase in vulnerability for the least vulnerable census tracts than for the most vulnerable tracts. In other words, increasing the population of these groups leads to more dramatic jumps in vulnerability for initially less vulnerable tracts. Levin.

Globally, environmental disasters impact billions of people and cost trillions of dollars in damage, and their impacts are often felt most acutely by minority and poor communities. Wildfires in the U.S. have similarly outsized impacts on vulnerable communities, though the ethnic and geographic distribution of those communities may be different than for other hazards. Here, we develop a social-ecological approach for characterizing fire vulnerability and apply it to >70,000 census tracts across the United States. Our approach incorporates both the wildfire potential of a landscape and socioeconomic attributes of overlying communities. We find that over 29 million Americans live with significant potential for extreme wildfires, a majority of whom are white and socioeconomically secure. Within this segment, however, are 12 million socially vulnerable Americans for whom a wildfire event could be devastating. Additionally, wildfire vulnerability is spread unequally across race and ethnicity, with census tracts that were majority Black, Hispanic or Native American experiencing ca. 50% greater vulnerability to wildfire compared to other census tracts. Embracing a social-ecological perspective of fire-prone landscapes allows for the identification of areas that are poorly equipped to respond to wildfires.

INTRODUCTION While many signs show that responsible investment has gained importance in capital markets , there has not been a significant global shift towards greater sustainability in finance. One potential reason for this paradoxical situation is the inconclusive debates on the performance of RI , despite more than 200 academic articles addressing sustainability's financial performance , including several metastudies . In this paper we propose another reason for this paradox, which we explore through a new theoretical lens. Building on the concept of collective beliefs , we argue that RI mainstreaming, i.e. the integration of environmental, social and governance issues into mainstream finance , is unlikely to happen as long as it is not supported by the collective beliefs, which are shared interpretations that guide investors' actions and decisions . We argue that the understanding of the collective beliefs around the activity of responsible investment can provide insights on equity market participants' decision making. The collective beliefs thereby inform the debate on sustainable development and financial markets. To support this proposition, we ask two questions: What are the collective beliefs for responsible investment and how have they evolved over time. Three areas of investigation flow from these research questions. We first want to identify the collective beliefs of RI in mainstream finance. Second, we address the evolution of the collective beliefs over time. Third, we discuss the implication of those collective beliefs for mainstreaming and RI in general. The analysis of the collective beliefs is based on the RI media coverage in the UK financial press between 1985 and 2010. During this period, we identified 3,462 articles to which we applied a bracketing method in order to decompose the history of responsible investment into successive RI periods. We then performed an in-depth content analysis of the press articles for a sub-sample of 89 articles. Our data show that five periods characterized the evolution of RI. Those periods are each marked with a very specific terminology and focus. It highlights the dynamism of the field but also the fact that RI is still in a process of institutionalization. The analysis has allowed us to identify collective beliefs around RI and revealed that the collective beliefs are not stable yet. The collective beliefs identified in our study can be classified in three main areas-'what is RI', 'why do RI' and 'how to do RI'. The content of the collective beliefs highlight the complexity of RI and tensions linked to RI mainstreaming. We primarily contribute to the literature on RI mainstreaming. Although mainstreaming has been widely discussed, very few studies, if any, have tried to theorize this phenomenon. We first provide a longitudinal study of RI supported with empirical data. Second, we focus on the meso-level, between the individual actors and the institutional level, by considering collective beliefs. And third we provide insights into the capacity of RI to become mainstream. Furthermore, our research on RI provides a new area of study for convention theory. Most importantly, we offer an empirical exploration of collective beliefs as well as a method to examine collective beliefs. Most of the studies on collective beliefs are indeed either conceptual or remain rather vague on how to empirically examine collective beliefs. Finally our study contributes to theory development by refining the notion of collective beliefs. We identify two types of collective beliefsjustifying RI and practicing RIand show that justification and action coexist and interact. Our paper therefore participates in an ongoing conversation about how institutions influence our thoughts and behavior. This paper is organized as follows. The first part outlines the theoretical framework and the context of RI mainstreaming. The second part presents the research design, data and methods used in the study. The third part provides the analysis and findings with a focus first on the RI periods, then on the collective beliefs. The results are discussed in the fourth and last part, including their implications and ideas for further research in the area of RI mainstreaming. --- COLLECTIVE BELIEFS AND RI MAINSTREAMING --- Collective beliefs The concept of belief is not common in economics and finance . The financial system is largely based on economic analysis and neo-classical financial theory that both give great importance to quantitative measures. However, by disregarding the beliefs and social context in which these numbers are produced, standard theories fail to explain anomalies such as speculative bubbles, confidence crises, excessive volatility, not the least of which is the latest financial crisis. They also fail to consider important dimensions of value . Value is a representation constructed by a group . This is true for religious value or aesthetic value but also for economic value, which neoclassical economics does not recognize. Even if economic value distinguishes itself from others because it is represented by a price, it still is largely a social construction. A well-documented illustration of collective beliefs is the social construction of price by the Black-Scholes-Merton options pricing formula. It gained exponential success in the 1970s among option traders, regardless of its accuracy in calculating option prices and of traders' personal belief in the accuracy of the model. Economic actors used this model to coordinate their actions under uncertainty, based on the collective belief that a majority of other economic actors used the model, with the unintended consequence of changing patterns of prices in the option market . A collective belief is a shared interpretation of the future evolution of financial markets, and plays a central role in Orléan's research . He defines collective beliefs as follows: "An individual I believes that the group G believes the proposition P if he believes that, in the majority, the members of the group G believe that the group G believes P" . A collective belief can therefore be disconnected from what individual agents believe: this is its self-referential nature. As a result, the market has its own autonomous belief, which is not the sum of individual beliefs. This becomes evident when investors make decisions based on their anticipation of the future behavior of "the market", and when we observe discourse such as "the market believes bonds are over-priced" or "the market does not believe the Federal Reserve's announcements". Under uncertainty, collective beliefs will help investors to make decisions, thereby influencing economic value and the adoption of new practices. A number of empirical papers have explored how financial market participants coordinate their actions based on collective beliefs . Bourghelle et al. and Guyatt address more specifically the case of RI mainstreaming, both suggesting that collective beliefs constitute impediments to mainstreaming. The former discusses this thesis theoretically, but is not backed up by empirical data. And though the latter refers to collective beliefs, these notions are presented in a behavioral finance perspective rather than according to Orléan's framework. ---Insert table 1 here ---During a period of instability, coordination based on collective beliefs increases stability. Everyone considers the same references, which reinforces their legitimacy. But since this coordination is based on beliefs and choices which could have been different, it is regularly challenged and may be put in peril. The studies therefore illustrate the content of collective beliefs and how they influence financial markets, but do not give much insight on how these conventions were formed or how they evolved. However, Bourghelle notes that since financial actors all read the same press and listen to the same experts, the financial press is an essential mediator in the formation of collective beliefs. The concept of collective belief is part of a theory, convention theory, which was developed by a group of economists and sociologists . Together they propose an enlarged model of rationality which becomes embedded in questions of coordination and values. Convention theory focuses on analyzing cognitive interactions and the multiplicity of equilibriums using discourse and conventions. Within this group of theorists, authors like Orléan, Bourghelle, and Dequech focus more specifically on understanding economic value in financial markets and highlight the self-referential nature of collective beliefs. This self-referential approach is the one that we adopt in studying collective beliefs in RI mainstreaming. There is no scientific basis to determine the mathematical expectations of the impact of ESG factors on the return of an investment. In this sense, expectations in the future of an asset's performance have a subjective component. They are informed opinions . Consider the following example: an individual fund manager believes a majority of other fund managers believe the market considers environmental, social and governance criteria as nonmaterial. This does not result from the fund manager's personal view on RI, or from other market players' personal views, but it will influence their investment decision. We argue in this article that collective beliefs play an important role in mainstreaming a new activity in financial markets. They can either support or hinder it depending on their content. --- Responsible Investment mainstreaming Although the definition of RI is the object of ongoing debate , scholars often agree that RI refers to the "integration of social, ethical, environmental and/or corporate governance concerns in the investment process" . However, both practitioners and the media use varied terminology when speaking about SRI. Non-exhaustive lists include "ethical investment", "green funds", "socially responsible investment", "sustainable investment" and "ESG investing" . This variety in terms points to a high heterogeneity which has developed throughout the history of RI . Practicing RI is neither easy nor unproblematic, because the heterogeneity of RI is not limited to its definition . The varied terminology goes along with a variety of investment strategies and practices ranging from exclusion based on screens, to shareholder engagement with companies, to selecting investments with a focus on having a community impact . RI mainstreaming is proclaimed by many RI practitioners and has become a central question in the RI research arena . But there is an implicit assumption within the field, in both academic and business communities, that everybody knows what mainstreaming is about, so it has never been quite defined. Mainstreaming of RI is sometimes considered as the maturation of RI . Lydenberg mentions the notion of professionalization, or even the notion of progress. These claims to RI mainstreaming recognize the evolution of the RI market in two directions: growth of assets under management and new investor categories. Mainstreaming is about the adoption or the practice of RI by major investors, consisting of the most important ownership groups of quoted companies, mostly represented by pension funds and insurance companies and referred to as institutional investors. Another element that we find in the literature on RI mainstreaming is the notion of RI spreading to every financial investment product category . Amaeshi et al argue that RI mainstreaming implies a fit between RI and the dominant financial market logic of calculation and singularization for profit, putting high emphasis on the financial performance of RI: "for the RI market to be mainstreamed, it has to be amenable to the mainstream financial market demands of objectivation and singularization " . All those factors lead us to say that RI mainstreaming means the integration of social, environmental and governance issues into conventional finance. Following these claims to mainstreaming, recent scholarship has focused on the impediments to mainstreaming. Various impediments to mainstreaming have been highlighted in literature, including at the institutional level, the organizational level and the individual level. These include, but are not limited to, opposing logics, lack of adequate products and tools, and lack of evidence of effectiveness in terms of social return . Juravle and Lewis state that "these views act almost as social paradigms within the finance community, facilitating or impeding the mainstreaming" of RI. A second type of impediment is the undemonstrated economic value of RI. This debate is important to legitimize ESG on the economic grounds that prevail in the business world and financial market . The question has been addressed from a management perspective and from an investor perspective . A few meta-analyses of these studies give an idea of the importance of this research topic: Orlitzky, Schmidt, and Rynes analyzed 52 studies, Allouche and Laroche analyzed 82, Wu analyzed 39 and Margolis et al. analyzed 251 studies . According to Margolis et al.'s meta-analysis, "after thirty-five years of research, the preponderance of evidence indicates a mildly positive relationship between corporate social performance and corporate financial performance", a result which neither satisfied the proponents of SRI nor its detractors. Some authors state that the results are inconclusive because the question was not correctly addressed. They criticize "several important theoretical and empirical limitations" , such as the sample size, insufficient historical data covering a short time span, aggregation of E, S and G issues which should be considered separately . A more recent type of criticism, made by Orlitzky and Vogel , is epistemological: financial markets are not efficient, financial actors are not rational, and there is no such thing as an intrinsic value which the market supposedly reveals. In that context, determining the value, and legitimacy, of sustainability may require looking away from neo-classical theories and financial return. With such views, this scholarship also challenges the dominant, simplified view of markets as fully efficient, transparent, and rational. Together, these studies reinforce the importance of understanding a financial activity in terms of discourse, representations, and social context. The point to be made here is that another perspective is needed to address how ESG signals affect financial markets and to understand how we collectively make sense of RI and its value. The approach we detail below links RI mainstreaming to collective beliefs, as crystalized in the financial press. --- RESEARCH DESIGN, DATA AND METHODS --- Media as a proxy for collective beliefs To research the concept of collective beliefs and to study their evolution in an emerging field with high uncertainty, we focus on the media coverage of RI in the financial press over time. Barkemeyer et al. suggest that "whilst it cannot be proven that there is a direct correlation between coverage of a specific event and change in behavior, there is no doubt that media coverage can influence the level of awareness of specific issues and could act as a general barometer of the contextual framing of issues such as business ethics, sustainable development, corporate citizenship, and accountability within society." There is a tradition of CSR studies as well as of finance studies using data from the media. Many authors are currently investigating the role and influence of the media in financial markets either in terms of the influence of media on companies' CSR , or in terms of the influence of media on share prices . We build on these studies to consider media as valid proxy for collective beliefs. However, the study of RI through media coverage is new to our knowledge. --- Data sampling and analysis The data we present come from the UK financial press and was selected in several stages. Table 2 outlines both stages of the research, with their respective sampling and analysis. ---Insert table 2 here ---We compiled articles by searching Factiva, a Dow Jones news database encompassing more than 28,000 sources 1 . We identified journals based on prior studies, limiting our search to the Financial Times, the Wall Street Journal Europe and The Economist, which all generally correlate closely with other sources of financial information such as Bloomberg, as suggested by Kaminsky and Schmukler and confirmed by Dyck et al. . The press articles provided non-sensitive, publicly available material, from which the researchers were independent. --- Stage 1: Determining the RI periods Sampling. We first constructed a large sample of all UK press articles addressing RI, using key word searches. The first set of keywords, selected based on their appearance and 1 On Factiva's content, see http://factiva.com/sources/contentwatch.asp?node=menuElem1522 . The data was collected in July 2010. frequency in RI academic literature, included: "responsible investment", "ESG", "sustainable investment" and "ethical investment". We completed the first set of articles with an iterative snowball process , using the names of key RI actors over time for new searches. EIRIS, FTSE and UN PRI are some examples of such names. The process was reiterated until searches using the names of field actors led to no new relevant articles: the saturation point. At the end of the first round of data collection, we had 3,982 articles on RI published between 1982 and 2010, as illustrated in Figure 1. The first articles we found in the database were published in March 1982. ---Insert figure 1 here ---Analysis. In this first stage of analysis we used a bracketing method for its descriptive utility in a longitudinal analysis, but also as a structuring process for analyzing and sensemaking. It allowed us to identify RI periods, which then constituted our new units of analysis for the exploration of the collective beliefs in the second stage of investigation. We identified the discontinuities at the frontiers between brackets based on the discourse, using word counts in a qualitative data analysis software. For example, between 2000 and 2001 the word "ethics" practically disappears and words such as "pension fund" and "pension manager" become salient in the data. This delineation based on content offers a stronger theoretical meaning than delineation based on coverage frequency, which can be influenced by economic cycles. Our approach is the first to provide empirical evidence of the evolution of RI over time, although several studies do trace the history of RI. We therefore verified the validity of our periods against periods proposed in literature , leading to similar conclusions with one exception: the early religious period of RI, which takes place before any coverage of RI in the financial press, that is before any sign of mainstreaming. Media studies tend to concentrate on this type of quantitative content analysis that is relatively easy to measure . But this type of analysis does not allow examining the symbolic meaning of the content, which leads us to our second stage of analysis, requiring a smaller sample. --- Stage 2: Identifying collective beliefs Sampling. To preserve the representativeness of the sample while reducing its size, we constructed theoretical two-week years, a recommended sampling method in media studies . Studies like those of Riffe, Lacy and Fico show that a qualitatively good newspaper sample should be based on at least twelve editions, where each day of the week is represented proportionally. For magazine articles in our data, such as articles from The Economist, we followed Wester according to whom an analysis of a weekly magazine can also be performed with a randomly selected issue per month, thus constructing a theoretical month, in some way similar to the process of theoretical weeks. We constructed our theoretical years after verifying that there was no seasonality in RI media coverage. The theoretical sub-sample is a selection of 89 articles distributed over the 18-year timeframe, a more manageable size for our second level of analysis which consists in identifying the RI collective beliefs. Analysis. The data analysis to identify collective beliefs was conducted in three main stages, a process which allowed us to move back and forth between the data and the emerging concepts to finally reach two abstract theoretical concepts. Our initial approach to code development was prior-research driven . We started the code list using literature on impediments to RI mainstreaming based on the idea that these impediments are issues around which there is uncertainty and confusion. These are typically situations where investors will need to rely on collective beliefs to coordinate their decisions. To ensure rater-to-expert reliability , the prior-research driven codes were discussed directly with the reference author. We read a first selection of articles searching for salient representations of RI, in the light of our prior-research driven list, and then completed the list with codes based on the data. Once all codes were named and grouped in categories, we followed an iterative process between data and categories to finalize our coding tree. Intra-rater reliability was achieved by coding the same text twice on different days, and inter-rater reliability was achieved by comparing the coding of a sample of texts by two researchers. The second step involved axial coding , where we compared firstorder codes with one another, looking for patterns and themes to create second-order constructs. This process consisted in trial and error constructions of models, regrouping different codes based on their characteristics in order to develop a set of more abstract, theory-rich constructs. The axial coding was done by one researcher and put to test by the other researcher in a series of meetings. Through these iterative discussions, three secondorder constructs appeared to have useful explanatory power in terms of collective beliefs. We tested the validity of the second order constructs quantitatively for the fifth RI period. The frequency of each second order construct was set as a hypothesis, which we tested by duplicating our analysis on a new random sample of 20 articles selected from within the universe of 1710 RI articles published during the fifth period . The results do not permit to affirm the existence of significant differences, with a confidence level of 1%. 2 Finally, in the third step we identified important dimensions from the sets of second-order constructs. For example, some codes like ethics or conflicts of logic appeared earlier in the history of RI. They were attempts to define and justify RI and they addressed more directly a critique, which was either implicit or explicit. Other codes, like engagement or innovation opportunity, appeared later in the history of RI, and were linked to success stories or challenges met by RI practitioners. We saw these as linked to the practice of RI and its challenges. Working in such a way through the relevant insights each construct provided, we consolidated the second level constructs into two broad theoretical dimensions: "justifying SRI" and "practicing RI." The theoretical dimensions resulting from the data resonate with convention theory literature, with a focus on the multiplicity of equilibriums and institutional maintenance or change. Most importantly, they provided guidance to understand the financial actors' collective beliefs around RI. A schematic overview of this process in figure 2 shows our first-order codes, second-order constructs, and derived theoretical dimensions which we use to study the mainstreaming process of RI among financial actors. ---Insert figure 2 here ---FINDINGS 2 When a confirmatory analysis is performed using a chi-square test, it appears that a difference exists between the proportions of occurrences of the three second order constructs "How", "Why" and "What". The confidence intervals constructed for the percentages of these occurrences indicate that the occurrences of "How" may be slightly over-estimated in the tested two-week-year sample, and the occurrences of "Why" may be slightly under-estimated, while the percentage of "What" falls within the constructed confidence intervals when the significance level was under 5%. Furthermore, with a 1% significance level, the data collected does not sanction the assertion of significant differences between the pi proportions of these three occurrences for Period 5 and the p*i estimations of the theoretical two-week-years. A first finding is the five periods of RI. They are both a result for the first level of analysis, and a step in the methodology providing units of analysis for the second level of investigation. A second finding is the emerging collective beliefs that affect the mainstreaming of RI. The implication of those collective beliefs for mainstreaming and RI in general are addressed in the discussion section. --- RI periods We found five RI periods in our data, spanning the period between 1982 and 2010: the «civil rights» years , the «green niche» years , the «professionalization» years , the «SRI» years and the "ESG" years . The first press article identified dated from 1982, which starts the first RI period. This corroborates Boxenbaum and Gond's observation that "RI terminology first appeared in the New York Times in the late 1980s" . The last press articles, dating from the end of 2010, do not signal the end of the fifth period, as no discontinuity in the RI discourse in the press was identified despite the 2008 financial crisis. From our data we can say that the financial crisis did not provoke any change in the normative foundation of finance. However the impact of the crisis may be long term rather than short term. Therefore it would be interesting to further monitor this last RI period to determine when it gives place to a new RI period, with new characteristics in terms of discourse and salient representations of RI. ---Insert table 3 here ---Civil rights years. Media coverage during this period is overwhelmingly turned to South Africa, apartheid, black worker wages, and targeted campaigns against companies. The main RI strategy discussed is the divestment practice. In the data, finance is linked to ethics during this period. We label it the "civil rights" period because of the focus on social issues. Green niche years. The second period corresponds to the end of the apartheid coverage in the financial press. This period is characterized by low media coverage of RI. Still, there are a few references to niche financial initiatives, mostly environmentally oriented such as "green funds". The concern for financial return linked to these ethical investments comes up for the first time in the discourse. Professionalization years. The third period corresponds to the early professionalization years. Pension funds start to get a lot of attention in the financial press, along with the issue of their social responsibility. This is a transition period during which the word "responsibility" becomes the preferred terminology when discussing ethical investments. --- SRI years. In this fourth phase , the term "ethical" is abandoned, and the term "SRI" is introduced. The professionalization is increasing and "fund managers" and "fund management" become some of the most frequent words of the sample. With the boom in the coverage of pension funds, come the first discussions on materiality and regulation linked to RI. ESG years. The fifth period begins in 2005, when the focus of RI shifts to climate change. It is also characterized by the combination of RI and corporate governance, which were so far treated separately. We label this last period the ESG period because media coverage is characterized by a search for neutrality in its wording, away from any ethical shade. It is also during this period that the term ESG appeared. Although our sample ends in 2010, there is no discontinuity so far that would allow closing the bracket. We therefore consider this period to be ongoing. --- Collective beliefs Our findings in terms of collective beliefs regarding RI consist of three secondorder constructs: defining or redefining what RI is, justifying the practice or non-practice of RI and clarifying how to approach RI, which we also refer to respectively as "what is RI", "why do RI" and "how to do RI". These collective beliefs are representations that emerged as focal points for individuals who, faced with uncertainty regarding the nature of RI, attempt to determine what the market will act upon. The collective beliefs for RI mainstreaming evolved over time around each of the three constructs. The evolution of each second order construct from period 1 to 5, illustrated in table 4, shows that the number of different collective beliefs around RI increases over time, meaning that constituents increasingly share common beliefs around this activity. ---Insert table 4 here --- The "civil rights" years are dominated by the belief that RI is about ethics, and that business and ethics are separate concerns. However there is a belief that RI will grow in the future. The «green niche» years emphasize the belief that RI does not lead to better performance than regular investment strategies, but that demand may grow. The «professionalization» years highlight the complexity of RI, complexity which reappears in the «SRI» years . However this fourth period also shows an increasing number of collective beliefs around RI including the long term perspective of RI, the need for a more sophisticated approach to practice RI, the issue of materiality and the lack of good information to evaluate companies on ESG factors. The «ESG» years bring in the notion of collaboration among actors and the importance of networks. The discourse in this period highlights the importance of engaging with companies. In this last period we witness an evolution of the collective belief justifying the practice of RI: demand for RI is collectively believed to be growing, whereas in the previous periods demand for RI was seen as potentially growing in the future. Although RI is not yet mainstream, it has gained recognition among mainstream investors. The results also show that questions of definition and understanding disappear from the debate around RI, which has shifted from "what is RI" to "how to do RI", highlighting the professionalization of the field, as illustrated in figure 3 and in the next two subsections. ---Insert figure 3 here ---Redefining what is RI Much of the RI discourse describes, defines or redefines responsible investment. This is particularly true in the early years of RI. We captured this discourse in the "Redefining what is RI" second order construct. Our data show that discussions around the ethics of RI play an important role in the first three periods. However, references to ethics are often located at the personal rather than collective level. In 2001, Sparkes raised the question: «whose ethics?» is RI referring to, because personal ethics may lead to many contradictions and tensions and hinder the process of mainstreaming. But the reference to personal ethics tends to disappear as from the professionalization years. Indeed, to mainstream RI, a higher level of abstraction may be necessary, which Donaldson and Dunfee refer to as hypernorms. Hypernorms are transcultural values that include fundamental concepts of rights and social good common to most major religions or countries. With the professionalization of RI comes a need to define the conditions and methods of the institutional investors' work . During this period, there is an attempt to obtain a certain normative control by standardizing professional norms. As a result, the predominant preoccupation relayed by the press in recent years has to do with information availability and standardization. The collective belief informs us that there is a lack of good data. And when information is available, it is often perceived as low quality and very dispersed, which does not allow benchmarking. "First comes greater standardisation of SRI performance measurement. The Social Investment Forum concluded: "With different definitions of SRI, market factors, cultural concerns and methodologies for collecting data, it is difficult to make controlled comparisons on a global scale." Source: 20070305FT The collective beliefs defining RI are characterized by ambiguity and differentiation over time: the data show a difficulty in being coherent. How can RI have values and bring financial return? Uncertainty around what is RI seems to be reduced during the green years: the belief during that period is that RI is green funds. The challenge of defining RI is not resolved for long, and resurges with the professionalization of the field. Funds cannot be "all things to all people" as one fund manager states . This has practical implications when faced with investment decisions in grey areas and suggests a fragmentation of the RI market. During the green years, RI has a narrow definition, making it easier to circumscribe. After the green years, RI has a broader definition. The discussion on defining RI still goes on but on a different, more granular level. The RI press coverage of the last period highlights the shortterm focus of financial markets, for many reasons, including quarterly reporting and remuneration structure linked to short-term objectives. In contrast, RI is believed to have a long-term focus. Investors are left to interpret this salience in different ways. If RI is long-term, it does not fit the short-term focus of financial markets. But it also means RI can offer the long-term vision missing in financial markets. Finally, it means that any effect of RI would be visible in the long-term. To conclude, the collective beliefs for defining what RI is tell us about investments driven by personal values versus hypernorms, a heterogeneous market, and long-term horizon. In addition, RI is perceived as a boundary object, with few shared representations regarding what RI is, beyond the idea that it has varying definitions, interpretations, and practical applications. ---Insert table 5 here --- --- Justifying the practice of RI An important proportion of the discourse throughout all periods attempts to legitimize RI by building business cases and disseminating examples of RI success. Discussions on the link between RI and financial return, risk management arguments, professional legitimacy considerations, as well as the market demand for RI are themes that peak during the green niche years, then fluctuate without ever being the main issue. Interestingly, the discourse justifying the practice of RI and the discourse defining RI evolve in opposition. The question of financial return is key in all periods, and the data show that RI funds have not yet convinced in terms of their performance. Articles relay the academic and practitioner studies trying to prove the link of financial return, and a new consensus seems to form around the idea that it is complex to prove performance. Our data for the ESG years affirm a positive link in 27% of cases and claims no financial underperformance for RI in another 27% of cases. The other 46% of the discourse states that there is a link but it is unclear, or notes that financial performance is key but that nothing more can be said about it. The collective belief is that the link between RI and financial return is inconclusive, unclear and complex, reflecting a very similar debate in academic literature. Demand for RI is the second most important justification discourse for practicing RI. As for financial return, it is also a market driven discourse. In the first four periods, the collective belief is that there will be a demand for RI, and that it will grow. This demand is mostly believed to come from the base and from unconventional institutional investors, such as faith based investors or NGOs. In the ESG years, the collective belief changes to there is a demand for SRI and it is growing. ---Insert table 6 here --- --- Clarifying how to practice RI A third type of collective belief clarifies how to practice RI. This discourse is practiceoriented, addressing and discussing methods, regulations, RI initiatives, available resources and skills, accountability and materiality challenges. It steadily increases over time, and becomes the main topic in the last RI period. With the professionalization of RI comes a need to operationalize RI. Before the SRI years, our European data make very few mentions of professional associations or RI training, and there is no mention of RI organizations with enough visibility to organize and regroup mainstream investors . In contrast, in the last period the focus of many articles is on networks, with statements such as "I invite all institutional investors to consider becoming signatories to the PRI and join a global network of peers working to address these priorities" or "an increasing number of networks are sharing information, developing knowledge centres and finding new ways to communicate" (source: 20071102FT). The collective belief at this stage is that RI requires collaborative engagement : active investors working together as described in the following quote. "As active owners, investors should work with investee companies to ensure comprehensive and systematic disclosure of the information they need in order to make responsible investment decisions. Ensuring the disclosure of information on ESG and other issues will enhance investors' understanding of their underlying investments and avoid a repeat of recent mistakes." Source: 20090302FT.2 Another collective belief of "how to do RI" is that RI requires a more sophisticated approach. And until this approach is available , the coordination process based on this collective belief may lead mainstream funds to cautiously not engage in RI. Another result of the belief may be to push institutional investors to develop the tools and models currently missing for RI mainstreaming. ---Insert table 7 here --- --- DISCUSSION AND CONCLUSION This paper focuses on responsible investment in financial markets, informed by an alternative theory of market participants' decision making. Our analysis of the collective beliefs for responsible investment and of the evolution of these beliefs over time allows us to develop two categories of collective beliefs which influence the capacity of RI to become mainstream: justifying RI and practicing RI. We now elaborate on how our findings contribute to RI literature and extend existing accounts of collective beliefs. --- On the RI periods Our findings show that RI is regularly reformulated in new terms, translated to fit the collective beliefs of the time. Each of the five periods that we identified is characterized by its own terminology for RI, particularly the latter two in which the terms SRI and ESG were coined. Many publications give empirical evidence of RI translation in space . Gond and Boxenbaum in particular followed the steps of RI translation to fit geographical contexts, through glocalization. They distinguished the translation in meaning brought by glocalization from an "interpretative translation, which solely involves symbolic, rhetorical or discursive changes" . In this paper we illustrate a case of translation over time rather than through space, and more specifically a translation to fit collective beliefs. In fact, we add to their assertion that translation goes beyond discursive changes if it reflects the content of the new collective belief. --- On the collective beliefs of RI Our data illustrated that beliefs are not fixed but evolve over time: from "RI underperforms" to "RI performance is inconclusive"; from "RI demand will grow in the future" to "RI demand has now grown". They thereby contribute to a changing landscape of RI. Changes in the collective beliefs are difficult because of the behavioral and institutional resistance, but possible with a significant amount of coordination work. Convention theorists put forward a taxonomy of four types of changes in conventions: general collapse, external invasion, translation and collective agreement . The last three are of particular interest for considering the adaptation of conventions that investment professionals and their agents adhere to, since the industry is by nature conservative and mindful of fiduciary obligations to beneficiaries, making a slow process of change more likely than an abrupt abandonment or collapse of existing practices. The importance of legitimacy within the conventions framework also makes collective agreement, or collaboration, preferable to going alone . Our data show that these processes are taking place. Translation --when the new convention integrates certain properties of the old one and is being reformulated in its proper terms--occurred at each change of period, with the adoption of new discourse and new terminology. A large increase in articles discussing RI has occurred since the SRI years in 2001, illustrating external invasion by institutional investors --that is, the slow increase in the number of individuals adopting the new convention until it reaches a critical mass where all will convert to the new convention. Both processes hint to mainstreaming. However, the content of the collective beliefs shows little to no sign of collective agreement . Instead, the collective beliefs recognize the complexity of RI and the need for new tools and approaches. These collective beliefs tell us RI is still under construction. Our findings raise a question for future research: how do suboptimal solutions persist? There are indeed many cases in finance where sub-optimal solutions are long lasting, such as the efficient market hypothesis. This study follows the trajectory of multiple collective beliefs to finally identify two distinct categories of collective beliefs: "justifying RI" and "practicing RI". Our data reveal that in the case of RI, collective beliefs "justifying RI" come first. It is not until these are largely resolved that collective beliefs about "practicing RI" dominate the discourse. These two types of collective beliefs emerging from our data reinforce an ongoing conversation about how institutions influence our thoughts and behavior. In addition, our findings illustrate how justification comes before action. This result may seem to contradict previous suggestions that the adoption of procedures comes logically prior to justification . Although our data do not allow to explain this observation empirically, we would like to suggest that, in the case of RI mainstreaming, justification can come before action because both take place in the pre-existing context of financial markets. As a result, the objects, people and units of measurement according to which RI will be justified already existeven if they are subsequently reinvented. In sum, the interaction between justifying and practicing plays a role in legitimizing RI and further analysis should tell us more about the implications of justification before practice. --- On the implication of collective beliefs for RI mainstreaming From our findings, it appears that investors can no longer reasonably really ignore the topic of RI. Between 1982 and2010, 3,462 financial press articles discussing RI were published in the media sources we sampled. However is it enough to claim that RI has moved from a niche activity to a mainstream practice? Some have argued that RI has become or is becoming mainstream while others are claiming that RI remains a niche . From the collective beliefs identified in the press, it seems difficult and premature to conclude that coordination has taken place among mainstream investors in terms of RI integration in mainstream finance. The nature of the collective beliefs remains confused and portrays a diffused perception of RI, but above all it highlights the numerous impediments for practicing RI. Those beliefs may well lead investors and analysts not to consider ESG factors. This cannot be described as an intentional strategythis is not a case of dominant convention based on the belief that «no one uses ESG factors»but rather as a result of the lack of clarity about RI. However, collective beliefs around RI are still evolving. As long as there is no collective belief encouraging the adoption of RI, sustainability remains a niche topic in financial markets. Although we cannot conclude on the existence of a collective belief leading to the integration of RI, we can say that the process of mainstreaming of RI is under way. Our findings support that as the discourse shifted from ethics to market logics , RI matured and attained greater professionalization, but also became less critical of mainstream finance. The collective beliefs in the fifth, "ESG years", period show no sign of RI altering the financial order, but point rather to RI being modified by conventional finance as it is slowly co-opted by the financial community. Daudigeos and Valiorgue suggest similar results for the evolution of corporate responsibility strategies, which were absorbed by the corporate system as they matured. The financial market often seems impervious to critique and to change, as illustrated by the RI periods and collective beliefs which did not change after 2008. This special issue states that, "in spite of increasing concerns about environmental and closely related social and governance issues, there has not been a significant global shift towards greater sustainability". If RI has the ambition to change mainstream finance's supply and demand for sustainability, the collective beliefs of RI have important implications. These beliefs indeed influenced the value and desirability of RI, and they do not endorse it so far. --- RI in a theoretical context of market inefficiency Because we consider that value is created at the collective level, through coordination processes and collective beliefs, our study emphasizes the meso-level, which tends to be understudied in the field of RI. In this way, convention theory is very different from typical economic theory: we are not studying micro-level individual preferences, like economic theory that presents equilibriums resulting from personal utility functions. Similarly, the focus of behavioral finance stays on the micro-level -individual's irrational beliefswhen it proposes to reconsider the rationality of actors and the efficient market hypothesis . But that framework creates new challenges, in terms of how to assess value. If there is no such thing as intrinsic value then value cannot be calculated by discounted cashflows. Furthermore, while these critiques are founded, they do not facilitate the case of RI mainstreaming because they do not fit in the financial language, which focuses on commensuration and mathematical models. If actors are irrational there is no commensuration, there is no benchmarking possible. This dilemma highlights the need for alternative models to neo-classical theories that can be translated into a market logic . With the convention theory lens, we also consider the RI market to be uncertain and confusing, thus not an efficient market. But we do not consider market players as irrational. Rather they take rational decisions based on their anticipations of collective beliefs. We built on Orléan's questioning regarding financial market's efficiency and regarding the fallacy of intrinsic value to better understand the collective beliefs around which actors coordinate. Our findings help us develop our proposition that the meso-level is the missing piece of a theoretical puzzle. Much effort has gone into demonstrating the value of responsible investment by measuring the link between RI and financial return. Those attempting to make the business case for RI, however, end up with inconclusive or unsatisfactory results. We suggest that these studies struggle to demonstrate the value of sustainability, because they approach value as an intrinsic notion instead of considering value as resulting from coordination processes. It is not that sustainability has little or no value; rather, its value is influenced by collective beliefs.

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

wide range of very different contexts. Some of these contexts are typically studied in computer science, such as wireless, ad hoc networks, transportation, vehicular networks, satellites, military, and robotic networks , while others belong to totally different disciplines. This is the case, for example, of the nervous system, livestock trade, epidemiological networks, and multiple forms of social networks . Clearly, while being different in many ways, these domains display common features; a time-varying graph is a model that formalizes highly dynamic networks encompassing the above contexts into a unique framework and emphasizes their temporal nature [13]. Knowledge mobilization refers to the use of knowledge toward the achievement of goals [14]. Scientists, for example, use published papers to produce new knowledge in further publications to reach professional goals. In contrast, patient groups can use scientific knowledge to help foster change in patient practices, and corporations can use scientific knowledge to reach financial goals. Recently, researchers have started to analyze knowledge mobilization networks using a social network analysis approach . In particular, [19] proposed a novel approach where a heterogeneous network composed of a main class of actors subdivided into three subtypes associated according to one relation, knowledge mobilization . Data covered a 7-year period with static networks for each year. The mobilization network was analyzed using classical SNA measures to produce understanding for KM using insights from network structure and actor roles [19]. The KM SNA studies mentioned above, however, lack a fundamental component: in fact, their analysis is based on a static representation of KM networks, incapable of sufficiently accounting for the time of appearance and disappearance of relations between actors beyond static longitudinal analysis. Indeed, incorporating the temporal component into analysis is a challenging task, but it is undoubtedly a critical one, because time is an essential feature of these networks. Temporal analysis of dynamic graphs is in fact an important and extensively studied area of research , but there is still much to be discovered. In particular, most temporal studies simply consider network dynamics in successive static snapshots, thus capturing only a partial temporal component by observing how static parameters evolve in time while the network changes. Moreover, very little work has been dedicated to empirically evaluating the usefulness of metrics in time . In this paper, we represent KMN by TVGs and we propose to analyze them in a truly temporal setting. We design a deterministic algorithm to compute a form of temporal betweenness in time-varying graphs that measures centrality of nodes in terms of how often they lie within temporal paths with the earliest arrival. We then provide, for the first time on a real data set, an empirical indication for the effectiveness of foremost betweenness. In particular, we focus on data extracted from [19], here referred to as Knowledge-Net. We first consider static snapshots of Knowledge-Net corresponding to the 7 years of its existence, and by studying the classical centrality measures in those time intervals, we provide rudimentary indications of the networks' temporal behavior. To gain a finer temporal understanding, we then concentrate on temporal betweenness following a totally different approach. Instead of simply observing the static network over consecutive time intervals, we focus on the TVG that represent Knowledge-Net and we compute foremost betweenness, explicitly and globally taking time into account. We compare the temporal results that we obtain with classical static betweenness measures to gain insights into the impact that time has on the network structure and actor roles. We notice that, while many actors maintain the same role in static and dynamic analysis, some display striking differences. In particular, we observe the emergence of important actors that remained invisible in static analysis, and we advance explanations for these. Results show that the form of temporal betweenness we apply is effective at highlighting the role of nodes whose importance has a temporal nature . A limitation of our algorithm is its applicability to small networks. In fact, any deterministic solution to the computation of foremost betweenness is inevitably very costly and, when faced with large networks, it is feasible to apply it only on small components. This research opens the road to the design of approximate variations of the algorithm so to make it applicable to larger scenarios, as well as to the study of other temporal measures designed for TVGs. --- Time-varying graphs --- Definition Time-varying graphs are graphs whose structure varies over time. Following [13], a time-varying graph is defined as a quintuple G = , where V is a finite set of nodes and E ⊆ V × V is a finite set edges. The graph is considered within a finite time span T ⊆ T, called lifetime of the system. ρ: E × T → {0, 1} is the edge presence function, which indicates whether a given edge is available at a given time; ζ : E × T → T is the latency function, which indicates the time it takes to cross a given edge if starting at a given date. The model may, of course, be extended by defining the vertex presence function , and vertex latency function . The footprint of G is a static graph composed by the union of all nodes and edges ever appearing during the lifetime T. --- Journeys A journey route R in a TVG G is a walk in G defined as a sequence of edges {e 1 , e 2 , . . . , e k } . A journey J , then, is a temporal walk in G comprising the sequence of ordered pairs {, , . . . , } if and only if ρ = 1 and t i+1 ≥ t i + ζ for all i < k . Every journey has a departure and an arrival that refer to journey's starting time t 1 and its finish time t k + ζ, respectively. Journeys are divided into three classes based on their variations based on the temporal and topological distance [30]. Journeys that have the earliest arrival times are called foremost journeys, journeys with the smallest topological distance are referred to as the shortest journeys, while the journey that takes the smallest amount of time is called the fastest. Moreover, we call foremost increasing journeys the ones whose route {e 1 , e 2 , . . . , e k } is such that birth-date ≤ birth-date. --- Temporal betweenness Betweenness is a classic measure of centrality extensively investigated in the context of social network analysis. The betweenness of a node v ∈ V in a static graph G = is defined as follows: B = u� =w� =v∈V |P| |P| , where |P| is the number of shortest paths from u to w in G, and |P| is the number of those passing through v. Even if static betweenness is "atemporal, " we denote here by B T the static betweenness of a node v in a system whose lifetime is T . Typi- cally, vertices with high betweenness centrality direct a greater flow and, thus, have a high load placed on them, which is considered as an indicator for their importance as potential gatekeepers in the network. While betweenness in static graphs is based on the notion of the shortest path, its temporal version can be extended into three different measures to consider the shortest, foremost, and fastest journeys for a given lifetime T [25]. In this paper, we consider foremost betweenness. Nodes with a high foremost betweenness values do not simply act as gatekeepers of flow, like their static counterparts. In fact, they direct the flow that conveys a message in an earliest transmission fashion. In other words, if the message transmission takes the path from foremost between nodes, such nodes provide a means to transmit the message in a more timely manner to all other nodes in the graph compared to the nodes that have lower foremost centrality. Thus, intuitively, they provide some form of "acceleration" in the flow of information. Foremost betweenness TB T F for node v with lifetime T is here defined as follows: where |F T | is the number of foremost journey routes between u and w during time frame T and |F T | is the number of the ones passing through v in the same time frame; n is the total number of nodes, and n is the number of nodes in the connected component to which v belongs. The factor n n is an adjustment coefficient to take into account possible network disconnections. In fact, it makes the betweenness of a node depend on the actual size of the connected component to which the node belongs, thus avoiding anomalous situations where a node in a very small component could be otherwise perceived as globally central. This would be the case, for example, of the center v of a small component in the shape of a star, where v would have maximum global betweenness while its central role is applied only to a very small portion of the overall network. --- Computing foremost betweenness The computation of betweenness centrality in static graphs can be done quite efficiently. Several approaches exist in the literature proposing either polynomial deterministic solutions or approximate ones for a variety of different graphs. Computing shortest-path betweenness in TVG can also be done in polynomial time, for example by adapting the algorithms described in [26,30]. The situation is rather different in the case of foremost betweenness, for which no algorithm has been proposed so far. In fact, it is easy to see that there exist TVGs where counting all foremost journeys or journey routes between two vertices is #P-complete, which means that no polynomial-time algorithm is known. Consider, for example, TVGs where edges always exist and latency is zero. In such a case, any journey between any pair of nodes is a foremost journey. Counting all of them is then equivalent to counting all paths between TB T F = n n u� =w� =v∈V |F T | |F T | , them, which is a #P-complete problem . In general, it is then unavoidable to have worst-case exponential algorithms to compute foremost betweenness in an arbitrary TVG. In this section, we first focus on foremost betweenness based on journey routes in the general setting . We then focus on foremost betweenness for special TVGs with zero latency and instant edges , which correspond to the characteristics of the knowledge mobilization network that we analyze in "Knowledge-Net". Note that each solution has the same worst-case time complexity, linear in the number of nodes in all the journey routes in the TVG, which can clearly be exponential. The advantages of the algorithm designed for the special temporal condition of instant edges and zero latency are mainly practical. In fact, the worst-case complexities are the same, but the execution time is better for our particular dataset. --- A general algorithm In this section, we describe an algorithm for counting all journey routes from a given node to all the other nodes in the TVG, passing through any possible intermediate node. This module is at the basis of the computation of foremost betweenness. We start by introducing some notations and functions used in the algorithm. Given an edge , let function arriv return the arrival time to y, leaving x at time t. Given a time-stamped journey π, with an abuse of notation, let us indicate by arriv the arrival time at the last node of π. The foremost arrival time in G to any node v from a given source s can be computed using the Algorithm from [30]. Let foremost denote such a time. We are now ready to describe the algorithm. The input of Algorithm CountFormem-ostJRoutes is a pair , where G = is a TVG and s is a starting node. The algorithm returns a matrix Count s [x, y], for all x, y ∈ V containing the number of fore- most journeys from s to y passing through x . The counting algorithm is simple and it is based on multiple Depth-First Search traversals. It consists of visiting every journey route of G starting from s, incrementing the appropriate counters every time a newly encountered journey is foremost. We remind that a node can reappear more than once in a journey route, with various occurrences corresponding to different times. This means that we need to store the time when a node is visited in the journey route so that, if it is visited again, we can determine whether the subsequent visit corresponds to a later time and thus the node has to be considered again. Note that this is the main difference with respect to a DFS in a static graph, where instead every node is visited exactly once. To perform the traversal managing multiple visits , we use two stacks: Path and S, where Path contains the nodes corresponding to the journey currently under visit and S contains the edges to be visited. In both Path and S, we store also time-stamps, to register the time of the first visit of nodes in Path and the time for the future visits of edges in S. If a node happens to be revisited at a later time, in fact, it is treated as a new node. The traversal starts as a typical DFS, pushing the incident edges of the source s onto stack S with their arrival times in these journeys . The nodes corresponding to the current journey under visit are kept in the second stack Path , initially containing only the source. When considering the next candidate edge to visit , we may be continuing the current journey or we may have backtracked to some previous nodes . In this last case, the content of Path is updated to reflect the backtracking . After visiting a node y , the DFS continues pushing on S the edges incident to y that are feasible with the current journey under visit (lines [17][18][19]. The if clause at line 20 checks whether the discovered journey is foremost and updates the corresponding counters. In other words, as soon as a journey π = [, , . . . , ] is encoun- tered in the traversal, Count[x i , x k ], i ≤ k is updated only if π is a foremost journey, and, regardless of it being foremost, the traversal continues pushing on the stack the edges incident to x k that are temporally feasible with π. Whenever backtracking is performed, however, the already visited nodes on the backtracking path are popped from Path in such a way that they can be revisited as part of different journey routes, not explored yet. --- Observations on complexity The running time of Algorithm CountFormemostJRoutes is linear in the number of nodes belonging to different foremost journeys, because it traverses each one of them. However, depending on the structure of the TVG, such a number could be exponential, thus an overall exponential worst-case complexity. More precisely, let µ s be the number of foremost journeys from a source node s to all the other nodes in G, n be the number of nodes belonging to those jour- neys, and n the number of nodes of G. Moreover, let µ and n be, respectively, the overall number of foremost journeys in G and the overall number of nodes in those journeys. The algorithm to count all foremost journeys from s to all the other nodes traverses every foremost journey from the source to any other node, and it performs an update for every visited node in each foremost journey that it encounters. Thus, its time complexity is O). To compute foremost betweenness, the algorithm has to be repeated for every possible source, thus traversing every possible foremost journey in G for a total time complexity of O). Since n could be exponential in n, we have a worst-case exponential complexity in the size of the network. Note that the high cost is inevitable for any deterministic algorithm to compute foremost betweenness. --- Algorithm for KnowledgeNet Algorithm 1 is applicable to a general TVG. We now consider a very special type of TVG with specific temporal restrictions that correspond to the type of network that we analyze in this paper. One such peculiarity is given by instant edges . Another characteristic is zero latency . Finally, in this setting, we base betweenness computation on increasing journey routes. We then describe a variation of the general algorithm specifically designed for those conditions , and we compute foremost betweenness applying the foremost betweenness formula restricted to foremost increasing journeys. --- Algorithm CountAllZeroLatency input : A TVG G i , starting node s ∈ V , Given a TVG G = , since we assume the presence of instant edges, we can divide time in consecutive intervals I 1 , I 2 , . . . , I k corresponding to k snapshots G 1 , G 2 , . . . G k ) , in such a way that ∈ E i implies that � ∈ E j for j � = i. Further- more, we know by ζ = 0 that an edge can be traversed in zero time. The key idea that can be applied to this very special structure is based on the observation that, given a foremost route π x,y from x to y with edges in time intervals I j , provided that j > i and j appears immediately after i, and given any journey route π ′ s,x from s to x with edges only in I i , the concatenation of π ′ s,x and π x,y is a foremost route from s to y, passing through x. This observation leads to the design of an algorithm that starts by counting the foremost routes belonging to the last snapshot G k only, and proceeds backwards using the information already computed. More precisely, when considering snapshot G i from a source s, the goal is to count all foremost routes involving only edges in ∪ j≥i E j , and when doing so, all the foremost routes involving only edges strictly in the "future" have been already calculated for any pair of nodes. The already computed information is used when processing snapshot G i in a dynamic programming fashion. As for Algorithm 1, the input of Algorithm 2 is a pair: a snapshot G i and a starting node s. The algorithm returns an array, Count s [u, v], where Count s [u, v] for all u, v ∈ V contains the number of foremost journeys from s to u passing through v counted so far . The actual counting algorithm on snapshot G i is a modified version of Algorithm 1, still based on Depth-First Search traversal. Lines 2-11 are exactly the same as in Algorithm 1, except that here we do not need to keep track of the arrival time for each edge, as we run Algorithm 2 in a single snapshot and the latency for edges is zero. In line 13, we examine whether the target of the current edge y has already been visited or not. If it has not been visited already, it either falls in the current snapshot, or it flows into the next snapshot. In the case where y stays in the current snapshot , we push its adjacent nodes into the stack S and determine whether the route ending at y is foremost. If a foremost route is discovered at y, we update Count s [z, v] by incrementing its value for all z ∈ Path . If instead it is not a foremost route in the current interval (lines [23][24][25], meaning that y is a node that existed in the "future, " a special update is performed using the data already calculated for the "future snapshots. " More precisely, when a journey route from s to x is a prefix of a journey route x y at a later time snapshot, we per- form a procedure called SpecialCount . The special count procedure involves aggregating the values of Count s [v, x] with Count x [v ′ , y], for all nodes occurring in the journey routes between s and x and between x and y . Algorithm 3 simply calculates the product of the number of foremost journeys between two routes s x, and x y, if they do not share any vertex . If instead they share some vertex v, the calculation is slightly more complicated: let a be the number of foremost journeys from s to y where v is visited at least once on the route between x and y; let b be the number of foremost journeys from s to y where v is visited at least once on the route between s and x; and let c be the sum of a and b. c represents the number of all foremost journeys from s to y that pass through v. However, c counts the journey route passing through v multiple times if v happened to exist in both Count s [v, x] and Count x [v, y], and we need to remove such multiple counting of journeys, which is done along with the update to Count s [v, y] in line 13. Count s [v, y]+ = Count s [v, x] × Count x [y, y] 6 end 7 else if v / ∈ s x and v ∈ x y then 8 Count s [v, y]+ = Count s [x, x] × Count x [v, y] 9 end 10 else if v ∈ s x and v ∈ x y then 11 a = Count s [x, x] × Count x [v, y] 12 b = Count s [v, x] × Count x [y, y] 13 Count s [v, y]+ = a + b -Count s [v, x] × Count x [v, y] --- Observations on Complexity The worst-case time complexity of Algorithm 2, CountAllZeroLatency, is the same as the one of the general algorithm, CountFormemostJRoutes. In our network, however, it performed better than Algorithm 1. We try to explain below the reasons for this. Algorithm 2 has to be executed in anti-chronological order of the different snapshots, starting from the last one, since it uses the previously calculated results in the computation of the new results. This approach is amenable to concurrent computations. In fact, since the graph is divided into independent snapshots, the number of all journeys can be computed separately for each snapshot, and the result of the calculation can be aggregated at the end. This has the advantage of eliminating all the special updates from the first part of the algorithm , and deferring them to the second part . Thus, instead of performing the special count at each level, we can postpone it to the last step of the algorithm, and loop once through all the collected counts with hard-coded intervals in the loop. While not being advantageous in worst-case scenarios, this strategy results in a more efficient solution from a practical point of view. Still, the algorithm is very costly, even in such a small network and it did run in almost a month when implemented in C++ with a machine with 40 cores and 1TB RAM. --- Knowledge-Net Knowledge-Net is a heterogeneous network where nodes represent human and nonhuman actors and edges represent knowledge mobilization between two actors. The network was collected for a period of 7 years [19]. Once an entity or a connection is created, it remains in the system for the entire period of the analysis. Table 1 provides a description of the Knowledge-Net dataset. The dataset consists of 366 vertices and 750 edges in 2011. The numbers of entities and connections vary over time starting from only 10 vertices and 14 edges in 2005 and accumulating to the final network year in 2011. Knowledge-Net is mainly composed of non-human actors, 272 in total , in relation with 94 human actors . Human actors include principle investigators , highly qualified personnel , and collaborators . It is through mobilization actors that individual, organizational actors and mobilization actors associate and mobilize knowledge to reach goals. For example, scientists mobilize knowledge through articles where not all contributing authors might be in relation with all other authors, yet all relate with the publication [19]. These non-human mobilization actors make up the bulk of the network including conference venues, presentations , articles, journals, laboratories, research projects, websites, and theses. According to an interpretation of the the Actor-Network Theory [37], the nature/ type/ characteristics of the mobilizer nodes have no interference with their role as a mobilizer. Following this interpretation, we consider that knowledge mobilization is beyond the role and nature of the nodes and we treat KnowledgeNet as a homogeneous network of knowledge mobilizers. All nodes of this network have the same function as knowledge mobilizer despite the fact that they might be quite different from each other from the view point of nature, type, and/or characteristics. Classical statistical parameters have been calculated for Knowledge-Net, representing it as a static graph where the time of appearance of nodes and edges did not hold any particular meaning. In doing so, several interesting observations were made regarding the centrality of certain nodes as knowledge mobilizers and the presence of communities [19]. In particular, all actor types increased in number over the 7 years indicating a rise in new mobilization relations over time. Although non-human individual actor absolute numbers remained small , these actors were critical to making visible tacit knowledge mobilization from around the world . Finally, embedded in human individual actor counts were individuals that the laboratory acknowledged in peer-reviewed papers, thus making further tacit and explicit knowledge mobilization visible. When representing Knowledge-Net as a TVG, we notice that the latency ζ is always zero, as an edge represents a relationship and its creation does not involve any delay; moreover, edges and nodes exist from their creation to the end of the system lifetime. Let birth-date denote the year when edge e is created. An example of a small portion of Knowledge-Net represented as a TVG is shown in Fig. 1. --- Table 1 Knowledge-Net data set with characteristics of actors and their roles at different times --- Start We also notice that, due to zero latency, edges spanning only one interval, and to the fact that edges never disappear once created, any shortest journey route in G is equiva- lent to a shortest path on the static graph corresponding to its footprint; moreover, the notion of fastest journey does not have much meaning in this context, because on any route corresponding to a journey, there would be a fastest one. On the other hand, the notion of foremost journey, and in particular of foremost increasing journey, is extremely relevant as it describes timely mobilization flow, i.e., flow that arrives at a node as early as possible. Note that in this setting the computation of foremost betweenness can be performed using Algorithm 2 introduced in the previous section. --- Study of KnowledgeNet --- Analysis on consecutive snapshots To provide more clear statistics on the Knowledge-Net dataset and a ground for better understanding of temporal metrics, we first calculated classical statistical measures on the seven static graphs, corresponding to the 7 years of study. The average for each value for the graphs is calculated to represent a benchmark on how the rank for each node is compared to others. The statistical data presented in Table 2 provide valuable information about the graph. The steady decrease in the centrality values confirms that the network growth is not symmetric, so the centrality values have long tails. According to Hanneman and Riddle [38], we should expect a high value of betweenness in Fig. 1 A small portion of Knowledge-Net represented as a TVG dense graphs due to the fact that it is highly possible that a path crosses every node. Meanwhile, when the betweenness values are normalized, they become low if all of the betweenness values are close to each other. Thus, the high value of betweenness , and the low value of its normalized counterpart in Knowledge-Net, indicates that the graph is either dense or is coupled in a way that there is a large number of shortest paths between any two arbitrary vertices. The graph is not dense as it is confirmed by the highest density metric of six. Therefore, the high number of shortest paths in the graphs caused the betweenness for most vertices to be similar and quite low when compared to the ones of nodes with the highest betweenness. Low average path length is a sign that the network presents small-world characteristics and the knowledge mobilization to the whole network is expected to be conducted only in a few hops. Meanwhile, the decreasing graph density along with the increasing average degree represents the slow growth in the number of edges compared to the number of nodes. Escalation in the number of communities with an increase in graph modularity metrics shows that the knowledge mobilization actors tend to form communities as time progresses. As the normalized average betweenness decreases steadily, it might be concluded that a few vertices in each community play the role of mediators and create the link between communities. Apart from these general observations, a static analysis of consecutive snapshots does not provide temporal understanding. For example, it does not reflect which entities engage in knowledge mobilization in a timely fashion, e.g., by facilitating fast mobilization, or slowing mobilization flow. To tackle some of these questions, we represent Knowledge-Net as a TVG and we propose to study it by employing a form of temporal betweenness that makes use of time in an explicit manner. --- Foremost betweenness of Knowledge-Net In this section, we focus on Knowledge-Net, and we study TB T F for all v. Nodes are ranked according to their betweenness values and their ranks are compared with the ones obtained calculating their static betweenness B T in the same time frame. Given the different meaning of those two measures, we expect to see the emergence of different behaviors, and, in particular, we hope to be able to detect nodes with important temporal roles that were left undetected in the static analysis. --- Foremost Betweenness during the lifetime of the system Table 3 shows the temporally ranked actors accompanied by their static ranks, and the high-ranked static actors with their temporal ranks, both with lifetime T = [2005][2006][2007][2008][2009][2010][2011]. In our naming convention, an actor named Xi is of type X, birth-date yy, and it is indexed by i; types are abbreviated as follows: H , L , A , C , J , P , C , I , and O . Note that only the nodes whose betweenness has a significant value are considered, in fact betweenness values tend to lose their importance, especially when the differences in the values of two consecutive ranks are very small [34]. Interestingly, the four highest ranked nodes are the same under both measures; in particular, the highest ranked node ) corresponds to the main laboratory where the data are collected and it is clearly the most important actor in the network whether considered in a temporal or in a static way. On the other hand, the table reveals several differences worth exploring. From a first look, we see that, while the vertices highest ranked statically appear also among the highest ranked temporal ones, there are some nodes with insignificant static betweenness, whose temporal betweenness is extremely high. This is the case, for example, of nodes S1 and J1. The case of node S1 To provide some interpretation for this behavior, we observe vertex S1 in more detail. This vertex corresponds to a poster presentation at a conference in 2010. We explore two insights. First, although S1 has a relatively low degree, it has a great variety of temporal connections. Only three out of ten incident edges of S1 are connected to actors that are born on and after 2010, and the rest of the neighbors appear in different times, accounting for at least one neighbor appearing each year for which the data are collected. This helps the node to operate as a temporal bridge between different time instances and to perhaps act as a knowledge mobilization accelerator. Second, S1 is close to the center of the only static community present in [2010-2011] and it is connected to the two most important vertices in the network. The existence of a single dense community, and the proximity to two most productive vertices can explain its negligible static centrality value: while still connecting various vertices S1 is not the shortest connector, and its betweenness value is thus low. However, a closer temporal look reveals that it plays an important role as an interaction bridge between all the actors that appear in 2010 and later, and the ones that appear earlier than 2010. This role remained invisible in static analysis and only emerges when we pay attention to the time of appearance of vertices and edges. On the basis of these observations, we can interpret S1's high temporal betweenness value as providing a fast bridge from vertices created earlier and those appearing later in time. This might indicate reasons for further study of the importance of poster presentations that can blend tacit and explicit knowledge mobilization in human-poster presentation-human relations during conferences, and continue into future mobilization with new non-human actors as was the case for S1 . The case of node J1 J1, the Journal of Neurochemistry, behaves similarly to S1 with its high temporal and low static rank. As opposed to S1 , this node is introduced very early in the network ; however, it is only active in 2006 and 2007. It has only three neighbors, A1, A3, and C1, all highly ranked vertices statically , A3), or temporally ). Since its neighboring vertices are directly connected to each other or in close proximity of two hops, J1 fails to act as a static short bridge among graph entities. However, its early introduction and proximity to the most prominent knowledge mobilizers helps it become an important temporal player in the network. This is because temporal journeys overlook geodesic distances and are instead concerned with temporal distances for vertices. These observations might explain the high temporal rank of J1 in the knowledge mobilization network. --- A finer look at foremost betweenness A key question is whether the birth-date of a node is an important factor influencing its temporal betweenness. To gain insights, we conducted a finer temporal analysis by considering TB T F for all possible birth-dates, i.e., for T = [x, 2011] ∀x ∈ {2005, 2006, 2007, 2008, 2009, 2010, 2011}. This allowed us to observe how tempo- ral betweenness varies depending on the considered birth-date. Before concentrating on selected vertices , and analyzing them in more detail, we briefly describe a temporal community detection mechanism that we employ in analysis. Detection of temporal communities According to Tantipathananandh et al. [27], accurately detecting communities in TVGs is an NP-hard and APX-hard task. Tantipathananandh et al. [27] used a heuristic to approximate the community detection for a more efficient algorithm. However, when the number of nodes in a dense graph exceeds double digits, the algorithm becomes computationally unfeasible to run. To the best of our knowledge, the only other work that attacked the community detection problem in TVGs is [39], where the problem is tackled by transforming the TVG into a series of static snapshot graphs with no repeated nodes in snapshots, and by incrementally detecting and adding to communities. While the complexity of the algorithm is not provided, it immediately proves inapplicable to our problem as it works only on series of snapshots with no repetition and includes aging factor in calculations. Thus, we take an approach similar to the one proposed in [27], by only focusing on approximating the communities for the purpose of this research. To do that, we first transform our TVG into a static weighted directed graph , which gives a rough indication of the foremost journeys of the actual TVG. We then use the journey graph as input to an existing community detection algorithm, designed for weighted graphs [40]. More precisely, given a TVG G = , we construct the journey graph of G, J = as follows: the nodes of J correspond to the original nodes of G and ∈ E ′ if there exists at least a foremost journey between x and y in G. The weight associated to edge ∈ E ′ is equal to the number of foremost journeys between x and y in G. An example of this construction is shown in Fig. 2. Note that Knowledge-Net, over time, creates only one connected component, but the community analysis of the Knowledge-Net graph results in 14 communities. The largest Fig. 2 Transformation of a TVG into a journey graph community consists of almost 39% of nodes and is centered around L1. Given the large number of the nodes belonging to communities and the low number of detected communities, it is clear that some of the central nodes share communities with each other. The case of node P1 This is a research project led by the principle investigator at L1. The project was launched in 2006 and its official institutional and funded elements wrapped up in 2011. Data in Table 3 support that P1 has similar temporal and static ranks with regard to its betweenness in lifetime [2005][2006][2007][2008][2009][2010][2011]. One could conclude that the temporal element does not provide additional information on its importance and that the edges that are incident to P-1 convey the same temporal and static flow. However, there is still an unanswered question on whether or not edges act similarly if we start observing the system at different times. Will a vertex keep its importance throughout the system's lifetime? The result of such analysis is provided in Fig. 3, where TB T F ) is calculated for each birth-date , with all intervals ending in 2011. While both equally important during the entire lifetime [2005][2006][2007][2008][2009][2010][2011] of the study, this project seems to assume a rather more relevant temporal role when observing the system in a lifetime starting in year 2007 , when its static between- ness is instead negligible. This seems to indicate that the temporal flow of edges incident to P1 appearing from 2007 on is more significant than the flow of the edges that appeared previously. With further analysis of P1's neighborhood in [2007][2008][2009][2010][2011], we can formulate technical explanations for this behavior. First, its direct neighbors also have better temporal betweenness than static betweenness. Moreover, its neighbors belong to various communities, both temporally and statically. However, looking at the graph statically, we see several additional shortest paths that do not pass through P1 . In contrast, looking at the graph temporally P1 acts as a mediator and accelerator between communities. More specifically, we observe that the connections P1 creates in 2006 contribute to the merge of different communities that appear only in 2007 and later. When observing within Fig. 3 Comparison between different values for vertex P1. Ranks of the vertex in the last interval are not provided as both betweenness values are zero interval [2006][2007][2008][2009][2010][2011], we then see that P1 is quite central from a static point of view, because the appearance of time of edges does not matter, but, when observing it in lifetime [2007][2008][2009][2010][2011], node P1 loses this role and becomes statically peripheral because the newer connections relay information in an efficient temporal manner. In other words, it seems that P1 has an important role for knowledge acceleration in the period [2007][2008][2009][2010][2011], a role that was hidden in the static analysis and that does not emerge even from an analysis of consecutive static snapshots. For research funders, revealing a research project's potentially invisible mobilization capacity is relevant. Research projects can thus be understood beyond mobilization outputs and more in terms of networked temporal bridges to broader impact. The case of node A3 Comparison between different values for vertex A3 are shown in Fig. 4, where ranks of the vertex in the last interval are not provided as both betweenness values are zero. The conditions for this node, a paper published in 2007, illustrate a different temporal phenomenon. Node A3 has several incident edges in 2007 ) when both betweenness measures are high. Peering deeper into the temporal communities formed around A3 is revealing: up to 2007, this vertex is two steps from vertices that connect two diffrent communities in the static graph. The situation radically changes, however, with the arrival of edges in 2008 that modify the structure of those communities, and push A3 to the periphery. The shift is dramatic from a temporal perspective because A3 loses its accelerator role where its temporal betweenness becomes negligible, while statically there is only a slight decrease in betweenness. The reason for a dampened decrease in static betweenness is that this vertex is close to the center of the static community, connecting peripheral vertices to the most central nodes of the network and H1). It is mainly proximity to these important vertices that sustains A3's static centrality. Such temporal insights lend further support to understanding mobilization through a network lens coupled with sensitivity to time. A temporal shift to the periphery for an actor translates into decreased potential for sustained mobilization. --- Invisible rapids and brooks On the basis of our observations, we define two concepts to differentiate the static and temporal flow of vertices in knowledge mobilization networks. We call rapids the nodes with high foremost betweenness, meaning that they can potentially mobilize knowledge in a timelier manner, and brooks the ones with insignificant foremost betweenness. Moreover, we call invisible rapids those vertices whose temporal betweenness rank is considerably more significant than their static rank , and invisible brooks the ones whose static betweenness is considerably higher than their temporal betweenness, meaning that these vertices can potentially be effective knowledge mobilizers, yet they are not acting as effectively as others due to slow or non-timely relations. Invisible rapids and brooks can be present in different lifetimes as their temporal role might be restricted to some time intervals only; for example, as we have seen in the previous section, S1 4 and5 indicate the major invisible rapids and brooks observed in Knowledge-Net. The presence of a poster presentation, a research project, two journals, and a conference publication among the invisible rapids supports that different types of mobilization actors can impact timely mobilization while not being as effective at creating short paths among entities for knowledge mobilization. In other words, they can play a role of accelerating knowledge mobilization, but to a concentrated group of actors. As for invisible brooks, we observe a journal ), three papers , C4, and C5) that cite publications by the main laboratory in the study ), a publication ) mobilizing knowledge from members of L1, and a research assistant who worked on several research projects as an HQP. In comparison with invisible rapids, there is a wider variety in the type of mobilization actors that act as brooks which does not readily lend itself to generalization. --- Table 4 Major invisible rapids Interestingly, we see the presence of journals among invisible rapids and brooks. From our analysis, it seems that journals can hold strikingly opposite roles: on the one hand, they can contribute considerably to more timely mobilization of knowledge while not being very strong bridges between communities, while on the other hand they can play critical roles in bridging network communities, but at a slow pace. A brook, the journal Biochemica et Biophysica Acta-Molecular Cell Research ), for example, helped mobilize knowledge in two papers for L1 , and is a journal in which a paper citing a L1 publication was also published. Given expected variability in potential mobilization for a journal, further research is needed to establish their roles in mobilization, whether these mobilization actors exist at both ends of the spectrum, or they have a neutral role in mobilization of knowledge. In contrast, the presence of a research project as an invisible rapid might indicate meaningful observations that should be studied further. First, because when public funders invest in research projects as a mobilization actor, an implicit if not explicit measure of success is timely mobilization with potential impact inside and outside of academia [19]. Ranking as a rapid is one measure that could therefore help funding agencies monitor and detect temporal change in mobilization networks. Second, a research project as rapid might be meaningful because by its very nature a research project can help accelerate mobilization for the full range of mobilization actors, including other research projects. As such, it is not surprising that they can become temporal conduits to knowledge mobilization in all of its forms. --- Conclusions In this paper, we proposed the use of a temporal betweenness measure to analyze a knowledge mobilization network that had been already studied using classical "static" parameters. Our goal was to see the impact on the perceived static central nodes when employing a measure that explicitly takes time into account. We observed interesting differences. In particular, we witnessed the emergence of invisible rapids: nodes whose static centrality was considered negligible, but whose temporal centrality appears relevant. Our interpretation is that nodes with high temporal betweenness contribute to accelerate mobilization flow in the network and, as such, they can remain undetected when the analysis is performed statically. We conclude that foremost betweenness is a crucial tool to understand the temporal role of the actors in a dynamic network, and that the combination of static and temporal betweenness is complementary to provide insights into their importance and centrality. The algorithm proposed in this paper to compute foremost betweenness constitutes a deterministic solution and its running time can be exponential in the worst case, which makes it applicable only on very small-scale networks. Since counting all foremost journeys in a graph is a #P-complete problem, such a high cost is inevitable for any deterministic solution. An open interesting direction is the design of approximate solutions, feasible for large networks. Temporal network analysis as performed here is especially pertinent for KM research that must take time into account to understand academic research impact beyond the narrow short-term context of academia. Measures of temporal betweenness, as studied in this paper, can provide researchers and funders with critical tools to more confidently investigate the role of specific mobilization actors for short-and long-term impact within and beyond academia. The same type of analysis could clearly be beneficial when applied to any other temporal context. In conclusion, we focused here on a form of temporal betweenness designed to detect accelerators. This is only a first step toward understanding temporal dimensions of social networks; other measures are already under investigation. --- Authors' contributions PF has proposed the problem. AAR and PF have discussed and designed together the two algorithms for the computation of foremost betweenness. AAR has implemented the algorithms. JG has provided the knowledge mobilization network data, which she had previously collected for a different study. AAR has conducted the analysis of foremost betweenness for these data. All three co-authors have discussed the results; in particular, JG has provided interpretation in the context of knowledge mobilization. All authors read and approved the final manusript. --- Authors' information Paola Flocchini is Professor at the School of Electrical Engineering and Computer Science. Her work and background are in distributed computing and algorithms. Amir Afrasiabi Rad has recently completed his Ph.D. on temporal analysis of social networks under Prof. Flocchini's supervision. Joanne Gaudet is co-president of an Ottawa-based company. The data collection she performed is from the time when she was a Ph.D. student at the University of Ottawa. Author details 1 School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, Ontario, Canada. 2 Alpen Path Solutions Inc., Ottawa, Ontario, Canada. --- Competing interests The authors declare that they have no competing interests. ---

Highly dynamic networks are networks where connectivity changes in time and connection patterns display possibly complex dynamics. Such networks are more and more pervasive in everyday life and the study of their properties is the object of extensive investigation in a

Introduction The primary objective of this paper is to explore the evolution and interrelationships of ideas pertinent to social ethics within the scope of philosophical discourse in Slovakia throughout the 20th century, a period that was marked by significant political and social transformations. Two of the most influential schools of thought that have affected the issue of social ethics in our region are Christian socialism and Christian realism. Both have significantly impacted the region's social ethics. The Lutheran church has traditionally been at the forefront of Christian socialism in Slovakia, and therefore, this paper will concentrate on representatives of this church when examining both Christian socialism and Christian realism. Understanding the historical development and dynamics of these theological and philosophical traditions is important for grasping the complex interplay between religion, ethics, and politics in Slovakia. As a secondary objective, this paper will also consider the potential influence which generational affiliations might have had on the development of these ideas. The methodology selected to achieve the aims of this study is grounded in the 'theory of generations' or the 'problem of generations'.1 --- The theory of generations The theory of generations is predominantly a sociological concept that highlights the significance of social generations, as distinct from blood relations,2 as a key framework for understanding the structures and developments of social and intellectual movements. Intriguingly, despite contributions from almost all scientific disciplines to explore this concept, comprehensive study of the theory has remained limited. Nevertheless, the practical significance of the theory becomes apparent when one endeavours to gain a more nuanced understanding of social changes and historical developments . Although the theory has faced critiques, particularly regarding its complexities and a predominantly Western perspective, it remains vital for our study. Specifically, this paper seeks to unearth insights to ascertain if the dynamics of intergenerational change3 significantly influenced social ethics4 in Slovakia during the 20th century. While Mannheim's theory of generations is closely associated with sociology, it has also exerted an indirect influence on philosophy, particularly in instigating debates and reflections on various subjects including history, temporality, identity, and the social construction of meaning. Even though the theory of generations is not central to philosophy, it can be applied differently depending on the specific context. Notably, the theory's influence is observable across diverse branches of philosophy, encompassing the philosophy of science, epistemology, metaphysics, and ethics. In moral philosophy, the theory of generations presents a unique perspective for investigating moral values and principles. It delves into how moral frameworks are established, the process of making moral judgments, and the guiding principles for ethical decision-making . By analysing the relationship between the theory of generations and social ethics, the study examines how generational consciousness and social values inform political and ideological perspectives, inclusive of religious and socio-economic beliefs. In this regard, the paper will explore how belonging to a particular generation may have influenced the embrace or rejection of ideas among selected authors. --- Social issues in reflection of Lutheran thinkers -Christian socialism Christian socialism is an ideological synthesis of socio-political, economic principles, and Christian ethical values, with a focus on social justice, poverty alleviation, inequality reduction, and the pursuit of the common good. 5 Emerging in the 19th century, Christian socialism responded to the challenges of industrialization and the excesses of capitalism. It aimed to provide a Christian alternative to secular socialism, which was often critiqued for neglecting moral and spiritual dimensions. Inspirational figures for Christian socialism include Francis of Assisi and Thomas More, who despite being an opponent of the Protestant Reformation, contributed significantly through his work "Utopia," which envisioned an ideal society founded on Christian principles. Among more contemporary figures, Charles Kingsley, an Anglican priest, played a significant role in advocating social reform and worker rights in the 19th century . In the Slovak context, authors drew inspiration from international figures such as Leonhard Ragaz, Paul Tillich, and Jacques Maritain. 6 The concept of Christian socialism evolved and reached an apex in the 20th century across various Christian denominations, especially within the Anglican, Lutheran, and Roman Catholic churches. In modern-day Slovakia, Christian socialism emerged as a significant political and social movement during the interwar period, responding to deepening socio-economic inequalities. Christian socialists aimed to enhance the working and impoverished classes' conditions, often burdened by high unemployment, low wages, and poor working conditions . Christian socialism strived to address these issues through social reform, human rights protection, and promoting social solidarity. It's noteworthy that the movement's prominence in the political and social life of Slovakia was particularly significant during the first half of the 20th century. However, within the Lutheran church during this period, social issues were often considered peripheral, with a primary focus on spiritual and moral human improvement . The relegation of social issues to the periphery within the church could be attributed to the conservative nature of religious institutions that were more focused on maintaining spiritual purity than engaging in socio-political matters. While Christian socialism was not an official doctrine of any denomination, its supporters can be found across various Christian churches. The acceptance of Christian socialism was often based on personal convictions, and within the Lutheran community, there were divergent views on the extent to which Christian principles should be integrated into social and political activism. Furthermore, reflecting on the intergenerational dynamics, it's crucial to consider how different generations within the Lutheran community in Slovakia perceived or engaged with Christian socialism . Were the values and principles of Christian socialism transmitted across generations? How did the social and political changes of the 20th century affect the reception and adaptation of Christian socialism among newer generations? Christian socialism was a pivotal movement in shaping social ethics among Lutheran thinkers in Slovakia during the 20th century. Its emphasis on integrating Christian ethical values into social and political reforms was significant, although received with varying degrees of enthusiasm within the Lutheran church. In the following sections, we will delve into how the movement evolved further and how other schools of thought, such as Christian Realism, emerged as significant influences on social ethics in Slovakia. --- Development of social ethics issues in Slovak region The interplay between generational consciousness and social ethics can be interpreted through multiple lenses. Within a historical framework, generational cohorts, such as Lutheran authors in this case, are shaped by the social, economic, and political conditions prevalent during their time. Specific historical events like economic downturns, social movements, or political upheavals, have the capacity to mold generational attitudes and values. Within the context under consideration, the critical historical events encompass those related to World War II and the emergence of communism. Additionally, generations are nurtured within distinct cultural and religious settings that influence their beliefs and values. Ideas of social ethics may find greater acceptance among individuals raised within religious communities that prioritize social justice, compassion, and communal accountability. Moreover, generational attitudes can intersect with other facets of identity, such as class, race, and gender, which in turn shape individuals' viewpoints and experiences regarding social ethics. In the first half of the 20th century, elements of Christian socialism are chiefly discernible among authors affiliated with the Protestant denomination, specifically Lutherans. On the contrary, Catholic philosophical authors 7 engaged in the discourse on social ethics, particularly in relation to Christian socialism, were considerably influenced by Catholic social doctrine that was not particularly receptive to socialism. 8 The historical context in present-day Slovakia is distinct due to its association with Catholicism during World War II. In 1939, the Slovak state was established following the dissolution of Czechoslovakia, with Catholic priest Jozef Tiso as its leader. This led to the emergence of an uncommon variant of what could be termed Christian socialism, which evolved into Christian National Socialism, the official doctrine of the Slovak state beginning in 1940. 9 However, determining the extent to which these ideas are interrelated is beyond the scope of this paper. During this time, Lutheran authors were largely marginalized, and as Gluchman observed, their focus was predominantly on practical, day-to-day issues . Nevertheless, authors from different denominations shared common criticisms of modernity, especially regarding its economic, social, political, and spiritual aspects. They condemned egoism, excessive individualism, and collectivism, along with economic liberalism, capitalism, communism, and fascism. This analysis will concentrate on two generations of Lutheran authors who actively participated in discussions concerning social ethics. Ľudovít Žigmund Seberini , a Lutheran author, was among the first to openly criticize Marxism from a Christian standpoint with his book "Od Marxa po Lenina. Boj moderného pohanstva s kresťanstvom" [From Marx to Lenin. Fight of contemporary paganism against Christianity]. However, the first systematic exploration of socialism emerged from Samuel Štefan Osuský ). 10 His seminal work, "Vývin socializmu po stránke sociálnej teórie" [The development of socialism in terms of social theory], published in the journal Cirkevné listy [Christian Letters], delves into the socio-economic and philosophical theories of the socialist movement. Osuský's work analyses various ideologies within the socialist movement, ranging from Marxism to anarchism and Christian socialism. He makes a clear distinction between socialism and Bolshevism, equating the latter with fascism. Osuský endeavoured to devise solutions to social and economic problems that the socialist movement 7 Predominantly Neo-Thomists who relied on the teachings of Thomas Aquinas, social and other encyclicals, and especially the Social Code published by the International Unity for Social Studies in 1927. They dedicated much attention to issues of natural, economic, and social equality among humans and discussed private property, social progress, and the role of the individual. Neo-Thomism's influence waned during the 1960s and 1970s. 8 It should be noted that Catholic doctrine evolved over time, with positions in the 20th century differing from earlier stances that condemned socialism, such as Pope Pius IX's Quanta Cura and the attached Syllabus of Errors . 9 Christian National Socialism was not Christian Socialism in the true sense of the term. However, both sought a compromise between socialism and capitalism based on Christian values. Christian National Socialism was more of a political and social movement that attempted to combine Christian principles and values with social reform and national self-determination. The concept emphasized the necessity for social justice, protection of workers' rights, and poverty alleviation. 10 The issue of social problems had been addressed much earlier, at the end of the 19th century by the Lutheran priest J. A. Fábry . sought to address, and he considered socialism to be the most influential component of the social movement in the 20th century. Osuský critically evaluated Ragaz's concept of religious socialism and highlighted its shortcomings. He perceived it as an unrealistic utopia detached from reality. According to him, neither religious socialism nor Christian ideology possesses the fundamental foundations necessary for tangible social change. His criticisms were centred on the goals and social consequences of religious socialism. Nonetheless, Osuský acknowledged the value in Ragaz's ideas as they shed light on numerous shortcomings and future challenges, albeit through a critical lens . Emil Boleslav Lukáč, another author from the pre-war generation, explored social progress which he believed was contingent on Christian values. He asserted that social progress is unattainable without Christianity as Christian values such as brotherhood, solidarity, and love for one's neighbour are essential for bridging the gap between the rich and poor. In his work "Kam ho položili" [Where They Put Him], Lukáč rejected Ragaz's notion of religiously oriented socialism. He believed that socialism and Christianity could not be integrated into a singular functional social system. However, he succeeded in reconciling Lutheran dualism through the synthesis of Christianity and humanism . Ľudovít Čečetka, another representative of Christian Socialism in Slovakia, demanded to take a Christian-critical stance on social problems and argued that socialism and Christianity meet on the common platform of a humanitarian effort to improve the human condition. Therefore, Christianity, according to him, should come closer with Christianity. The fact that in modern society there has been a mutual alienation has brought negatives to both of them . Čecetka, like his contemporaries, believed that Christian principles should guide social reform and that the church has a responsibility in addressing social problems. Rudolf Koštial ) is regarded as the pioneer of Christian socialism as the principal school of thought concerning social ethics in the region. In his work, Koštial emphasizes that the social question is a central issue of modern times, and that Christianity is obligated to address it. His main work, "Kresťanstvo a sociálna otázka" [Christianity and Social Question], outlines a Christian social program that closely aligns with the social teachings of the Catholic Church. According to Koštial, Christian socialism can only emerge through the revival of Christianity. He believed that a synthesis of Christianity and socialism is plausible. Interestingly, both Catholics and Lutherans criticized Koštial's approach. Lutherans accused him of failing to differentiate between Lutheran and Catholic teachings, and for deviating from the Lutheran doctrine of dualism . Catholics, on the other hand, criticized him for his perceived attacks on Catholicism . Ján Hano and Július Madarás are among the final representatives of the pre-war generation of Christian socialism. Hano saw Christian socialism as a middle ground between Christianity and socialism, whereas Madarás adopted a more overt stance against secular socialism . Hano contended that capitalists, in their relentless pursuit of wealth, held nothing sacred and disregarded religious values. He accepted socialism as a tool for reducing social disparities but rejected the Marxist ideology associated with it. Hano emphasized that the goal of Christian socialism is a spiritual revolution that embraces both Christianity and socialism . The pre-war generation was characterized by critical evaluations of Marxist socialism, Marxism, and Bolshevism, and believed that Christian socialism could either solve or contribute to solving social ethics issues. In contrast, the post-war generation faced a broader and more urgent array of social issues. The authors of the post-war generation had to contend with political and ideological pressures from Marxist socialism and sought alternative strategies to address social ethics issues. They found their answer in Christian realism. --- Post-war generation Christian realists distinguished themselves from Christian socialists in several notable ways, primarily through their refusal to assert the exclusivity of Christianity as the solution to societal issues. Instead, they contended that the role of religion was to foster an environment in which adults could be nurtured into becoming actively engaged individuals who were capable of viewing the world and its problems through a realistic lens, without overemphasizing the role of Christianity. Notable representatives of Christian realism in Slovakia included Ján Michalko, Karol Gábriš, and Július Filo . Gluchman posits that a salient distinction between Christian realists and Christian socialists lies in the former's ability to more cogently articulate their theoretical assumptions, thereby establishing a more solid foundation for addressing real-world problems. Contrary to the Christian Socialists who primarily adapted the Swiss model of religious socialism to fit the Slovak context, Christian realists devised innovative approaches that aligned with the political, social, and economic realities that emerged post-1948. This approach evolved into what became known as face-to-life theology,11 which is rooted in the teachings of the Lutheran scholar, Paul Tillich . Paul Tillich ), a prominent German-American theologian and philosopher, is widely recognized for his contributions to existentialist theology and Christian existentialism. He was particularly influenced by the Social Gospel movement early in his career, which sought to employ Christian principles in addressing social problems, and which aligned with Christian socialist values, particularly in advocating for economic and social justice. Tillich was actively engaged in religious socialist movements in Germany, which aimed at reconciling Christian teachings with socialist principles.12 His theology was also deeply imbued with existentialist notions, particularly the concept of existential anxiety. He viewed economic and social issues as sources of existential angst and believed that addressing these issues was crucial to tackling broader human existential concerns . Slovak proponents of face-to-life theology were particularly drawn to the pragmatic elements of Tillich's theology. In a period when Christianity in the region was undergoing transformation post-1948, Tillich's teachings offered viable alternatives. With regard to Marxism, Tillich adopted a balanced perspective, acknowledging the merits of Marx's critique of capitalism and the promotion of economic justice, but expressing reservations regarding Marxist materialism and its view of religion. Notably, unlike Ragaz and Tillich, representatives of face-to-life theology refrained from overt criticism of political socialism. It is imperative to highlight that while the pre-war generation was closely aligned with Ragaz's Christian socialism, the post-war generation exhibited a pragmatic inclination towards Tillich. This shift can be partially attributed to confessional differences-Ragaz was Calvinist while Tillich was Lutheran-but primarily to Tillich's more conciliatory stance towards Marxism and socialism. Face-to-life theology, which can be regarded as a Slovak variant of Christian realism, emerged within a highly specific historical and cultural milieu, and addressed particular challenges. Ján Michalko defined the principal tenets of this theology in his 1953 article "Teológia tvárou k životu" [Face to Life Theology]. He characterized humans as frail, fallible, and sinful, yet also as proactive, accountable, and dedicated individuals striving to infuse the world and human relationships with the virtues of love and justice . In the postwar era, societal issues evolved and differed from those of the pre-war era. The pre-war generation grappled with poverty, pauperism, and unemployment, which were ostensibly eradicated post-1948. The post-war generation, however, was preoccupied with promoting active Christian participation in worldly affairs. Christian involvement, rather than socialism, was deemed essential in providing social service to others. Both Gabriš and Michalko advocated the notion that assistance is perennially required by someone, somewhere. Consequently, the binding thread of this generation was activism, a renunciation of passivity, and a celebration of love . Despite minor divergences-for instance, Michalko emphasized global solidarity, Gabrish underscored collective cooperation, and Nandransky highlighted commitment-these differences did not detract them from their shared objectives. Another shared characteristic was their uncritical acceptance of socialism and deliberate oversight of its shortcomings . As Michalko asserted, Christianity's role in society is manifested through spiritualism, realism, activism, optimism, universalism, and humanism; social aid, which was previously within Christianity's purview during the pre-war era, became either a state or socialist domain, or was deemed redundant. --- Comparison and conclusion Both Christian Socialism and Christian Realism were grounded in Christian ethics and values. They both emphasized the role of Christian teachings in shaping moral and ethical behaviour in society. Both movements having in focus social ethics issues showed concern for social justice and sought to address societal problems. They emphasized the role of the individual and community in working towards a more just and equitable society. Both Christian Socialism and Christian Realism were critical of unchecked capitalism. They questioned the societal impact of economic systems that did not incorporate ethical considerations. Both movements engaged with socialist principles to varying degrees. Christian Socialism sought to synthesize Christian values with socialist principles, while Christian Realism did not dismiss socialism and recognized the need for economic justice but did not wholly embrace it. Christian Socialism tended to advocate for a synthesis between Christian values and socialism, believing that socialist principles could be in line with Christian ethics. On the other hand, Christian Realism was more pragmatic and did not assert that Christianity held the exclusive answer to social problems. It acknowledged the role of different social systems, including socialism, without endorsing any one system exclusively. The pre-war generation focused on issues like poverty, unemployment, and critiques of Marxist socialism. The post-war generation, influenced by Christian Realism, was more concerned with the active participation of Christians in society, social engagement, and personal responsibility. The theology of face-to-life, which was an outcome of Christian Realism in Slovak region, was concerned with issues that were allowed to it. This theology was the result of the social and political constraints that existed in the region after 1948. In this context, while Christian Socialists viewed Christianity as an integral framework for solving societal problems, Christian Realists viewed the role of religion as assisting in nurturing engaged individuals who could address societal issues without overemphasizing Christianity. Christian Socialists were generally critical of Marxist socialism. In contrast, Christian Realism, especially as influenced by Paul Tillich, had a more nuanced view of Marxism. It appreciated Marx's critique of capitalism and the idea of economic justice but held reservations about Marxist materialism and views on religion. Christian Socialism in the Slovak region was influenced by a range of thinkers including Catholic and Lutheran theologians. Christian Realism, particularly in its face-to-life theology form, was significantly influenced by Paul Tillich, a Lutheran theologian. Christian Realism emphasized activism and rejected passivity. It encouraged individuals to actively participate in society to bring about love and justice. Christian Socialism focused more on the synthesis of Christian values with socialist principles as a solution to social issues. In summary, Christian Socialism and Christian Realism share a foundation in Christian ethics and concern for social justice, but diverge in their attitudes toward socialism, the role of Christianity in society, and their focus on societal issues. The transition from Christian Socialism to Christian Realism in the Slovak region reflects the adaptations and responses to the changing social, political, and economic contexts of the times. Based on the provided study it can be inferred that the dynamics of intergenerational change played a substantial role in the issue of social ethics in Slovakia during the 20th century. In the early part of the 20th century, the pre-war generation in Slovakia was influenced by Christian Socialism, which emphasized addressing issues such as poverty, unemployment, and critiquing Marxist socialism. This generation leaned towards synthesizing Christian values with socialist principles to tackle social issues, believing that socialism could be aligned with Christian ethics. They viewed Christianity as an essential framework for addressing and solving societal problems. However, post-World War II, the social and political landscapes changed dramatically, and this led to a shift in perspectives on social ethics among the subsequent generations. The post-war generation in Slovakia adopted a different approach. Face-to-life theology which came from Christian Realism was a reaction to the social and political constraints of that time and evolved in response to the need for a more pragmatic approach towards social ethics. Christian Realism did not claim the exclusivity of Christianity in solving social problems. Instead, it recognized the importance of preparing and nurturing engaged individuals who could address societal issues realistically, without overemphasizing the role of Christianity. This approach was influenced by Paul Tillich's theology, which was more benevolent towards Marxism compared to the Christian Socialist perspective. The post-war generation focused on activism, rejecting passivity, and emphasized love and justice as key values. This shift from Christian Socialism to Christian Realism reflects the dynamics of intergenerational change in Slovakia during the 20th century. The changes in political, social, and economic realities after World War II shaped the ways in which the post-war generation addressed social ethics issues. The emphasis shifted from integrating Christian values with socialism to a more realistic and pragmatic approach that recognized the role of different social systems in addressing societal problems. In conclusion, the dynamics of intergenerational change had a substantial impact on social ethics in Slovakia during the 20th century. This was manifested in the shift from a Christian Socialism framework in the early part of the century to a Christian Realism approach in the latter part, as society adapted to the changing social and political contexts.

This paper explores the dynamics of intergenerational change and its impact on social ethics in Slovakia during the 20th century, with a particular focus on authors of Lutheran background. The methodology selected to achieve the aims of this study is grounded in the 'theory of generations.' The purpose of this analysis is to examine how shifts in political, social, and economic realities influenced the ethical frameworks guiding societal engagements in Slovakia. Through a historical analysis of theological movements and their socio-political contexts, the study reveals a significant shift from Christian Socialism, which was prevalent in the pre-war era among Lutheran thinkers, to Christian Realism, which emerged among the postwar generation. Christian Socialism sought to synthesize Christian values with socialist principles, while Christian Realism adopted a more pragmatic approach that acknowledged the role of various social systems without asserting the exclusivity of Christianity. The findings of this study highlight the flexibility and responsiveness of social ethics to changing social realities, and the role of intergenerational dynamics in shaping these ethical frameworks.

Introduction In 2016, the Chinese government issued a health blueprint which emphasized that general practitioner service is an imperative way to achieve "Healthy China 2030" [1]. In China, GPs are primarily in charge of grassroots and community health centers that provide residents with a full range of medical care to achieve the goal of a hierarchical medical system. Especially in the current post-COVID-19 era, GPs work as the bridge that connects patients and illnesses, reducing the risk of infection and releasing the workload for hospitals. Since the emergence of the COVID-19 in China, primary healthcare centers have played an unprecedented and critical role in curbing its spread [2]. As such, the needs of GPs have rapidly increased, not only for the COVID-19 but also for future unpredictable pandemic crisis. The government plans to train 35,000 GPs in various ways, particularly focusing on improving community-based health services [3]. Not just in China, there is a shortage of GPs across the world [4]. For example, in the United Kingdom, approximately 30% of all GPs plan to quit direct patient care within five years [5]. In the United States, it is expected to have a shortfall of 7,800 to 32,000 GPs by 2025 [6]. However, young medical students and doctors have been found to have a lower interest in choosing GP careers [7][8][9]. Thus, it is crucial to find out the factors that would impact medical students -as a GP reserve force -to choose a GP career. Due to the imbalance in economic development and the diversity in Chinese geography, a large proportion of China's population is at the grassroot-level, primarily using CBHS [10]. CBHS is mainly delivered by the grassrootlevel health care institutions including community health service centers, urban/suburban health centers, and rural clinics [11]. Compared with higher level hospitals, these grassroot-level health institutions have accessed to fewer medical resources but played as the "gatekeepers" of Chinese residents' health in the first line [12]. However, types and/or professional capacities of medical personnels in CBHS vary largely; there are GPs, doctors with undergraduate medical degrees, assistant physicians, and even unqualified medical personnels in some rural and/ or remote areas, which makes governments urgently promote GP trainings to establish a reliable CBHS workforce for the vast and varied grassroot-level population [13]. Given the current landscape and importance of CBHS, it is equally important to understand the determinants of medical students' commitments to CBHS. Many have considered China's medical education system to be one of the most complex in the world [14]. Medical students in China can obtain varying levels of medical degree, allowing them to have distinct medical education pathways and then careers [15]. As the increasing demand for GPs, GP-oriented college majors/ specialties have been established since 2010s. Unlike clinical medicine majors that prepare for varying health professionals [16], GP-oriented majors aim to train more future GPs. However, either GP-oriented students or clinical medicine students could have varying career choices due to the perplexing medical education system. These medical students could choose to a health professional career either stick with their college majors or transfer to a specialty after several years training and practices in hospitals. Also, many of them are not willing to be CBHS workers in the grassroot-level health institutions particularly those in rural and/or remote areas [17]. Such contexts present a compelling and critical opportunity to understand what factors may affect medical students' GP career choices and their commitments to CBHS. Previous research has revealed a variety of factors as possible influences on the career choices of medical students, including individual, familial, and institutional levels [18][19][20][21]. For example, Heiligers [18] found that there were sex differences in medical students' motivation and career choices. Bittaye et al. [20] included medical students' familial characteristics such as parental occupation and educational level when examining their choices of specialty. With respect to characteristics of medical schools, Campos-Outcalt and Senf [21] found that they were associated with students' specialty choices. Focusing on GP career choices, Deutsch et al. [22] found that practice-orientated GP courses offered during undergraduate medical education positively affected students' GP career choices. In addition, a significant correlation between GP teaching and their future GP careers choices has been found [23]. Building upon this existing literature, we included a richer set of individual, familial, and institutional levels to explore the factors that potentially relate to medical students' GP career choices. Notably, research has found that medical students' attitudes to GP careers may change due to the work environment [24]. It should be noted that there are distinguished differences in the working environment between hospitals and community or rural health services providers in China. GPs based in the community are restricted from prescribing medicines compared to GPs in hospitals, which leads to a lower level of trust between doctors and patients [25]. Moreover, the earnings of those working in hospitals are much higher than those of those working in CBHS. Research has found that medical students' career choices are associated with their educational debt [26,27], therefore, it is possible that students are less willing to commit to CBHS. This study attempts to understand the factors that influence medical students' GP career choices and commitments to CBHS. Specifically, we mainly address the following two research questions: What individual, familial, and institutional factors will affect Chinese medical students' GP career choices? What individual, familial, and institutional factors will affect Chinese medical students' commitments to CBHS? To our knowledge, the current literature fails to simultaneously address both questions. Using a sample of undergraduate medical students across the country, we conducted logistic regression models to examine the factors that affect students' GP career choices and commitments to CBHS. It contributes to the global literature on medical students' pathways to GP careers and CBHS. In practice, these findings are extremely important for China's medical education stakeholders, which could help to foster more GP professionals and lead them to CBHS, thereby responding to the substantial needs for health services in GP, especially at the CBHS level. --- Materials and methods --- Data and sample This study leveraged the self-administered data: College General Practice Education to conduct the quantitative analysis. The CGPE suited the present study particularly well because it comprehensively collected medical students' GP-related information such as GP career choices, CBHS commitments, GP knowledge, and GP internship. Additionally, it collected abundant students' individual information such as sex, ethnicity, urbanicity, familial information such as household annual income and parental occupations and educations, institutional information such as medical school types, regions, and GP curriculums and teachers. The CGPE survey was designed by the research team which was supported by the National Social Science Foundation of China. It was distributed to medical students across the country via WJX.CN -a widely used online survey platform in China. The survey period was from June 16 to October 19, 2022. There were two steps in the data collection process. First, we nationally recruited 20 faculties from medical schools using a convenience sampling strategy. Second, these faculties distributed the survey link to their medical students. All the participants were voluntary and were informed that their data would be used confidentially and could not be identified through the research. Validity check was performed before conducting the analysis, yielding a final analytic sample of 3438. Among them, 53.37% were women students, which closely aligned with the female proportion of the whole Chinese university students population from 113 studies [28]. With regards to ethnicity, 89.09% were Han -the socially dominant group in China, and 10.91% were other minorities, such as Zang, Hui, and Man [29]. The corresponding Han proportion was 92.0% in Gao et al. [28], which was also reasonably similar. Therefore, the final analytic sample demonstrated a reasonably unbiased representative of a wider population in terms of sex and ethnicity. --- Measures --- GP Career choices and CBHS commitments GP career choices and CBHS commitments were the two dichotomous outcome variables in the present study. Students were asked whether they would choose GP career after graduation, if yes, it was coded as "1", otherwise, "0". Similarly, if a student committed to CBHS, it was coded as "1", otherwise, "0". --- Individual characteristics Except sex and ethnicity, other individual characteristics included urbanicity, college majors, GP familiarity, and GP internship. Urbanicity measured the original places that a student was from, urban, suburban/town, or rural. College majors indicated whether a medical student was majoring in GP-orientated major or clinical medicine major such as internal medicine, pediatrics, neurology, and chirurgery. GP familiarity had two componentsfamiliarity of GP policy and GP knowledge. Familiarity of GP policy measured how familiar a student was with GP-related policies using a five-point Likert scale, with 5 = extremely familiar and 1 = extremely unfamiliar. To comprehensively measure familiarity of GP knowledge, a series of survey items were consisted of -"are you familiar with differences between general practice and other specialties?", "are you familiar with general practice or basic health services?", "are you familiar with the roles of general practitioners in the health service system?", and "are you familiar with work tasks of general practitioners?". Reliability test and factor analysis were conducted to assess whether these survey items well measured this generated latent variable [30]. The results showed a high reliability, with Cronbach's alpha = 0.94 -and high factor loadings ranged from 0.82 to 0.91, meaning 67.24-82.81% of the variances in these items were explained by familiarity of GP knowledge. Internship has been found to be associated with students' GP choices [31]. This study also included students' CBHS internships experience as a dichotomous variable , and their beliefs on current internships using a five-point Likert scale with 5 = extremely useful and 1 = extremely useless. --- Familial characteristics Familial characteristics were sorted into four categories. First, since one child policy has substantial effects on the Chinese society [32], we measured students' sibling status as "1" if a student was from a one-child family, otherwise, "0". Second, household annual income was categorized into low-, middle-, and high-income . Third, father's occupation and education level were included. For simplification, occupations were divided into two categories: government/public institution employees and other jobs; education level were divided into three categories: less than four-year degree, fouryear degree, and four-year degree beyond. Fourth, mother's occupation and education level were coded similarly with these of fathers. --- Institutional characteristics Seehusen et al. [33] conducted a systematic review and found that institutional characteristics were likely to affect medical students' career choices. In this study, we included students' medical school types, either operated in a comprehensive university or as an independent school . In addition, based on China's geographical division, regions of medical schools were sorted into Southwest, North, Eastern, Central, Northwest, and Southern China. Furthermore, GP educational resources such as whether institutions offered GP curriculums and teachers -were likely to affect students' GP career choices, thereby being included as dichotomous variables . --- Analytic plans The analysis of this study was conducted by Stata 16. First, descriptive statistics of outcome variables and independent variables were provided. To further understand how students' GP career choices and CBHS commitments distributed by sex and ethnicity, we conducted Pearson's chi-squared test. Second, given the outcome variables were binary variables, logistic regressions were performed to examine whether or not the potential student-, family-, and institution-level factors affect medical students' GP career choices and CHBS commitments. Stepwise logistic regression models were built, starting from adding individual characteristics, and then familial and institutional characteristics. The final logistic regression models could be expressed as follows: logit P gp 1 -P gp i = β 0 + β 1 SC i + β 2 F C i + β 3 IC i , (1 ) logit P cbhs 1 -P cbhs i = β 0 + β 1 SC i + β 2 F C i + β 3 IC i , (2 ) where gp = whether student chose a GP career after graduation in Eq. ; cbhs = whether student committed to CBHS after graduation in Eq. ; SC i = a vector of stu- dent i's individual characteristics; F C i = a vector of stu- dent i's familial characteristics; SC i = a vector of student i's institutional characteristics. Noteworthily, we additionally assessed whether these individual, familial, and institutional characteristics were too highly correlated with each other. Specifically, we checked the variance inflation factors for multicollinearity before conducting logistic regressions. The results showed that the largest VIF for institutions located in Central China was 2.40, suggesting that multicollinearity may not need to be taken into account. Last, post-estimation was conducted to further assess whether there were significant marginal effects. In other words, since medical students' GP choices and CBHS commitments were likely to be associated with their sex, we were particularly interested in whether these relationships varied by regions, as China's economic growth has been unbalanced among regions. --- Results --- Descriptive statistics Table 1 reports the distribution of medical students' GP career choices and CBHS commitments by sex and ethnicity. Chi-square tests were included to assess the significant differences in the distribution of GP career/ Non-GP career choices and CBHS/non-CBHS commitments among students. First, with respect to GP career choices, we observed significant differences by sex and insignificant differences by ethnicity . Specifically, we found that 67.74% of women would like to choose GP careers as compared to 74.55% of men. Then, we turned to focus on the disparities in CBHS vs. non-CBHS commitments, which we found a different pattern -varied significantly by sex and ethnicity . We found that 63.87% of women would like to commit to CBHS, as compared 2 reports the descriptive statistics for independent variables. For dichotomous variables, percentage results were reported in the main text instead of mean results in Table 2. Concerning students' individual level characteristics, 26% of students were from urban areas, while 22% were from suburban/town and 52% were from rural areas. The majority of students were majoring in clinical medicine. With respect to the familiarity of GP, the mean of students' familiarity of GP policy and knowledge was 2.91 and 3.31, respectively, around "moderately familiar". For GP internship, 33% of students had a CBHS internship and the mean of beliefs on current internship was at 2.45, which ranged between "neutral" and "useless". With respect to familial characteristics, 32% of students were from only-child family, and 55% were from low-income families. Governments/public institution employees comprised about 1/5 of father's and mother's occupations, with 20% and 19%, respectively. Notably, majority of fathers and mothers had a less than four-year college degree, with 77% and 83%, respectively. Turning to institutional characteristics, 54% of students were studying at a medical school located in a comprehensive university, while 46% were studying at an independent medical school. Students from medical schools located in Eastern China consisted of the largest share, with a 30%. In addition, 83% and 77% of medical schools offered specific GP curriculums and teachers, respectively. --- Predicting medical students' GP career choices Table 3 reports the results from a series of logistic regression models in predicting medical students' GP career choices. For easier interpretation, odds ratios were reported. Model 1 shows that women were 25% less likely than men to choose a GP career, holding other variables constant . In addition, students from GP-orientated majors were 4.53 times more likely to choose a GP career than those from clinical medicine majors. Interestingly, students who were more familiar with GP policy and knowledge were less likely to choose GP careers. Notably, students who had CBHS internships were 1.97 times more likely to choose GP careers. Model 2 introduces familial characteristics, as additional predictors that may affect students' GP career choices. However, no additional significant relationships were observed after controlling these familial characteristics. Model 3 further introduces institutional characteristics. The results showed that sex, GP-orientated major, and the familiarity of GP policy and knowledge remained significantly in predicting students' GP career choices. Additionally, we observed that students from medical schools in comprehensive universities were 1.44 times more likely to choose GP careers. Notably, comparing to Southwest region, students from Eastern region were more likely to choose GP careers , while those from Central region had a lower probability . --- Predicting medical students' CBHS commitments Table 4 reports the results of factors in predicting medical students' CBHS commitments. Model 4 shows that women were 1.24 times more likely than men to commit to CBHS. In addition, students from urban areas were 36% less likely to commit to CBHS than those from rural areas. GP-orientated major was still a strong predictor for CBHS commitments . Similar to the analytic process in predicting GP careers above, Model 5 shows that students' beliefs on current internship and mothers with four-year college degree and above -significantly and negatively predicted their CBHS commitments, after controlling familial characteristics. Including all individual, familial, and institutional characteristics, Model 6 displays that Han students were 28% less likely than other minority students to commit to CBHS. We also found that students who had CBHS internships were 1.2 times more likely to commit to CBHS than those who had not. In addition, students from high-income families were 31% less likely to commit to CBHS than those from low-income families. Regional differences were also observed. Comparing to Southwest region, North, Central, and Northwest regions were 72%, 53%, and 57% less likely to make CBHS commitments, respectively. --- Marginal effects of sex on GP careers and CBHS commitments by regions From above logistic regression results, we observed significant relationships between sex and regions and GP career choices, and between sex and regions and CBHS commitments. To further explore whether the relationships between sex and GP career choices and CBHS commitments were varied by regions, we added interaction terms -sex by regions in our final preferred Model 3 and Model 6, which included all individual, familial, and institutional characteristics. Figure 1 displays that the lines of women and men were not visually parallel, after adding the interaction terms, indicating that these relationships did vary by regions. Note. Total N = 3,438. SD = standard deviation. For consistency, means were reported for all dichotomous and continuous variables Table 5 reports the results of marginal effects of sex on GP careers and CBHS commitments by regions, while Fig. 2 visually displays these marginal effects. Results showed that in Southwest region, women were 9% less likely than men to choose GP careers, while this value was 11% in Central region. With respect to CBHS commitments, women were 11% and 24% more likely than their counterparts to commit to CBHS in Northwest and Southern regions, respectively. No significant differences between sex groups were found in other regions. --- Discussion This study investigates the factors that influence medical students' GP career choices as well as commitments to CBHS. Our results show that individual factors like students' sex, GP-orientated majors, and institutional factors like regions significantly predicted both GP career choices and commitments to CBHS. Familial characteristics like annual income and mother's educational level only significantly predicted commitments to CBHS. Specifically, our results show that women students were less likely to choose GP careers, which is inconsistent with Lefevre et al. [34], which found that 76.3% of women students would choose GP careers in France. It is possible that the inconsistency is due to different GPs' workloads in different countries. China has one of the largest populations of patients in the world, which requires GPs to work out-of-hours regularly. Past studies like French et al. [35] have found that in Scotland, men GPs worked more hours and were more likely to work after hours than their women peers. Thus, Chinese women students may not choose GP careers in order to avoid the extreme workload of GPs. In addition, we find that women students were more willing to commit to CBHS than their male peers. This is consistent with Chuenkongkaew et al. [36], which surveyed medical students from five Asian countries and found that female medical students were more likely to work in rural areas than their male peers. This is probably because those who commit to CBHS require high empathy, and women in general are more empathetic due to maternal instinct [37]. Another reason may be that in China, CBHS usually does not need to work after hours, however, primary hospitals do. Unsurprisingly, our results reveal that students who majored in GP-oriented majors were more likely to choose GP careers and commit to CBHS than those who majored in clinical medicine. Many clinical medicine students believe that GP careers have lower salaries and are not valued by professionals [38]. Therefore, they prefer to stick with their major. However, students in GP-oriented majors may put family first and pay more attention to the flexibility of working hours [39], thereby choosing GP careers to have a better work-life balance. With respect to committing to CBHS, our findings are supported by Gill et al. [40], which demonstrated that students in GP-oriented majors had a preference for working in a rural community and placed more emphasis on continuity of care. Interestingly, our results show that students who are knowledgeable about GP-related policy and knowledge do not choose GPs, which may suggest that GP-related policy and knowledge mastered by medical students did not meet their expectations. The existing research has found that curriculum, internships, and the power of role models play an imperative role in the choice of GP careers [22,41]. However, current medical education in China is plagued by issues such as one-sided teaching of the general practice curriculum, short-term rural-oriented internships, or teachers' failure to act as positive role models, which will cause students' unsatisfaction with GP careers. Furthermore, GP-related policies have unattractive salary provisions [42], which are much lower than those of specialties. These help explain why students who are more knowledgeable about GP, on the contrary, are less likely to choose GP careers. With respect to familial characteristics, we find that students from high-income families were less likely to commit to CBHS. According to Nicholson et al. [19,36], medical students' pre-med choices were also influenced by their families. Thus, it is possible that high-income families tend to invest more in their children's education and expect them to have more decent jobs. However, GP careers have a low social status in China which may decrease high-income students' interests. This new finding may have implications on establishing a GP-welcoming social environment to improve GPs' social status. In addition, we also found that students whose mothers had more than a four-year college degree had less probability of committing to CBHS. This is consistent with Liu et al. [36], which surveyed 232 medical students and found that medical students' attitude toward remaining in rural areas was strongly associated with mothers' education level at postsecondary or above. With respect to institutional characteristics, we find there are regional differences in GP career choices and commitments to CBHS. Specifically, students in the Eastern region were more likely than Southwest region students to pursue a GP career. Eastern region in China is generally regarded as the economically high-developed region [43]. Research has found that general practitioners with low salaries have lower job satisfaction [44]. It is reasonable that GPs in economically high-developed regions have higher salaries than economically less-developed regions like Southwest, and have higher job satisfaction, which in turn, influences medical students' GP careers. To our knowledge, limited existing literature has explored regional differences in medical students' intentions to GP careers, meanwhile the present study contributes to this research gap. In addition, students in the Southwest region--where has a higher rural density--were more willing to commit to CBHS. This is consistent with Zhang et al. [45], which was conducted among 2,714 medical students from three medical schools, suggesting that students from rural areas were more likely to choose to work in communities than those from urban areas. Similarly in Australia, after surveying the employment locations of medical graduates at 12 universities, McGirr et al. [46] demonstrated that medical students with rural backgrounds were 3.1 times more likely to practice medicine in rural areas than medical students with metropolitan backgrounds. Moreover, we found that sex plays an important role in regional differences in GP career choices and commitments to CBHS. Women in Chinese traditional culture are often seen as "homemakers" whose primary role is to take care of the family. These prejudices have lasted over time, especially in the areas of gender inequity, where most people still believe that it is a woman's job to take care of the family [47], thereby avoiding jobs with high workloads and choosing to work close to home [48]. This could help to partially explain our finding that women students in Southwest region are less likely to choose GP careers, and women students in Southern region were more likely to commit to CBHS, as both regions were known for gender inequality. However, the reasons of insignificant results of Southwest women students' GP career choices and Southern women students' commitments to CBHS, remain unclear, and could not be explored by the present study. Despite comprehensively design the present study, three limitations should be noted. First, although the study included individual, familial, and institutional characteristics, other students' psychological variables could also be generated. Studies like Lent et al. [49] conducted structural equation models to better understand how psychological variables predicted students' career choices. In addition, specific majors could be collected. Thus, future research could redesign the study and add new insights to the influential factors that affect medical students' GP career choices and commitments to CBHS, by specific majors. Second, the present study could not conclude a causal relationship. Causal inferences are important to evidence-based administrations [50]. Future research could conduct quasi-experiment designs such as propensity score matching to reduce selection biases. However, it is still valuable for stakeholders to understand what individual, familial, and institutional characteristics impact students' GP career choices and commitments to CBHS. Third, the measure of GP career choices was not the declared GP careers, instead, we used medical students' willingness to choose GP careers. However, this could still inform stakeholders with practical implications on the cultivation of GPs. Future research could survey GPs rather than medical students to more explicitly identify these potential factors. --- Conclusions Understanding the factors that affect medical students' GP career choices is an important alternative to respond to the increasing needs for GPs in China. Additionally, understanding the factors that affect students' commitments to CBHS can contribute to establish a reliable CBHS workforce that can provide first line protection for the numerous and diverse grassroot-level population. This study sought to explore these influential factors. The results indicated that individual factors like students' sex, GP-orientated majors, and institutional factors like regions significantly predicted both GP career choices and commitments to CBHS. Familial characteristics like annual income and mother's educational level only significantly predicted commitments to CBHS. It should be noted that GP-orientated majors have been established since 2010s to cultivate more GPs, and our results suggested that GP-orientated majors do function well in such a role. In addition, by understanding these influential factors, medical education stakeholders can make informed decisions on attracting more medical students to GP-orientated majors, which in turn cultivates more GP professionals to meet the nation's demand for GPs. Furthermore, by understanding the factors that influence medical students' commitment to CBHS, policymakers could make beneficial policies to increase medical students' motivations to the grassroot-level health institutions, and devote to CBHS as gatekeepers for a large population of residents' health. --- Data Availability The datasets used and analyzed during the present study are not publicly available due to the protection of participants privacy, but are available from the corresponding author upon reasonable request. --- --- --- Competing interests The authors declare no competing interests. ---

Background There is a substantially increasing need for general practitioners (GPs) for future unpredictable pandemic crises, especially at the community-based health services (CBHS) level to protect the vast and varied grassroot-level population in China. Thus, it is crucial to understand the factors that affect Chinese medical students' GP career choices and commitments to CBHS.Leveraging the self-administered data collected across the country, this study conducted logistic regressions with 3,438 medical students. First, descriptive statistics of outcome variables and independent variables were provided. Then, stepwise logistic regression models were built, starting from adding individual characteristics, and then familial and institutional characteristics. Last, post-estimation was conducted to further assess whether there were significant marginal effects.Results showed that women students were 24% less likely to choose GP careers but were 1.25 times more likely to commit to CBHS than their men peers, holding other individual, familial, and institutional characteristics constant. In addition, students who major in GP-orientated were more likely to choose GP careers and commit to CBHS, respectively, than those who major in clinical medicine. Furthermore, familial characteristics like annual income and mother's educational level only significantly predicted commitments to CBHS. Notably, sex-related differences in GP career choices and commitments to CBHS -by different regions -were observed. Conclusions Understanding the factors that affect medical students' GP career choices sheds light on how medical education stakeholders can make informed decisions on attracting more medical students to GP-orientated majors, which in turn cultivates more GP professionals to meet the nation's demand for GPs. In addition, by understanding the factors that influence medical students' commitment to CBHS, policymakers could make beneficial policies to increase medical students' motivations to the grassroot-level health institutions, and devote to CBHS as gatekeepers for a large population of residents' health.

Introduction High levels of drug consumption and mental health problems are reported among the Roma population in Spain [1]. Research indicates that these two related problems are connected to social determinants of health, such as living conditions, employment, education and housing, among others [2,3]. Considering that EU official surveys show the Roma to be consistently found in unacceptable levels of deprivation, marginalization, and discrimination [4], these problems tend to persist within Roma communities, thus having not been well studied yet. While there is a relative wealth of data showing higher incidence of chronic diseases, infant mortality, less healthy lifestyles [5][6][7], there is scarce research shedding light on strategies, factors or transformative dynamics signaling how to improve the Roma health status. In this context, the increasing importance of the Iglesia Evangélica Filadelfia and, more particularly its Pentecostal denomination, cannot be ignored. Based only in Spain, the IEF is part of the Christian movement known as Pentecostalism. The origin of Pentecostalism among the Roma dates back to the 1950s. In the north of France there is an evangelization movement initially headed by Clément Le Cossec that led to the conversion to Pentecostalism of some Spanish Roma families living in that area. On their return to Spain, these families shared their conversion experience with other Roma families, and this is how Pentecostalism expanded among Spanish Roma. This evangelization movement led to the founding of the Iglesia Evangélica Filadelfia in Spain in 1969. In 2013, a total of 1100 congregations, between 150,000 and 200,000 members and between 3500 and 4000 pastors were reported throughout Spain [8]). For more than three decades now, this church has successfully penetrated the most remote Roma communities, incorporating many of the Roma cultural features into its organization and functioning, and gathering the largest number of Roma participants [9][10][11]. Previous researchers have indicated the role of the IEF in preventing certain habits and promoting healthy ones. These studies have focused on the effect of concrete prohibitions against consuming alcohol or drugs on the health of their members [12][13][14][15][16]. The Spanish Ministry of Health and Consumer Affairs has recognized the health benefits provided by the IEF, particularly its role in rehabilitating and preventing drug consumption [17]. Also, Estruch, Gómez, Griera and Iglesias [18] highlighted the work of the IEF in the prevention of addictions through the organization of campaigns to raise awareness and the promotion of detoxification in collaboration with Evangelical organizations that have rehabilitation centers [18]. The present article represents a step beyond previous studies, as it is aimed at disentangling whether there are factors that promote the rehabilitation and protect the Roma members of this church from drug consumption and mental health problems. In what follows, a literature review about Roma and drug consumption and mental health is presented. After, the methodological design is provided followed by the results section. The article ends with the discussion of the results and some concluding remarks. This research is framed under the work conducted by the authors under the IMPACT-EV project [19]. --- Greater Risk of Drug Consumption among the Roma During the 80s, drug consumption and more particularly heroine hit the Spanish population, the Roma community being not an exception to it. This boom directly affected the core of many Roma families in Spain, not only as consumers but sometimes being taken advantage as traffickers by those who were making a business of it, leaving important damages behind. Due to this vulnerability, at that moment and at the present, research identifies a greater risk among the Roma of consuming drugs [1,2]. This greater exposure is identified in different countries. The Roma population is excluded and lives in marginalized neighborhoods, in unhealthy conditions where there are low levels of education, high unemployment rates and a high percentage of alcohol consumption and drug injection converge [1,2]. This is the case of Dzsumbuj, a neighborhood located in Budapest with a large Roma population. In 2004, a study was carried out in this neighborhood. While no cases of human immunodeficiency virus were detected, the study drew attention to the high levels of HIV risk behavior in the study population, as indicated by high levels of drug injecting, the high prevalence of Hepatitis A Virus, Hepatitis B Virus and Hepatitis C Virus infections, and the alarmingly high levels of hepatitis co-infections. The results identified that the risky behaviors among Roma or mostly Roma populations may put them at higher risk of contracting HIV and other blood-borne and sexually transmitted infections than the behaviors among mostly non-minority populations [2]. Similarly, in a study carried out in a juvenile detention center, it was identified in its sample that the highest percentage of acquired immune deficiency syndrome virus infections was concentrated among Roma drug addicts. It also warned of a high risk for said group to contract diseases such as hepatitis or tuberculosis. The consumption of alcohol, tobacco or injected drug increases this risk. If high poverty rates and unhealthy living conditions affecting the Roma population are combined, the risk for Roma people skyrockets. In fact, a tuberculosis outbreak was detected among the Roma in the Camp de la Bota , a marginalized neighborhood in Barcelona in 1985, which gave rise to a contact study. The tuberculosis incidence among the Roma people was 115 times higher than the tuberculosis incidence rate in Barcelona in the same year [1]. In addition, impoverished conditions and low levels of schooling, specifically among Romani women, are reported to have a negative impact on their ability to navigate the health system, leading to the proliferation of harmful practices and habits. In fact, studies indicate that Roma women with lower education tend to consume more drugs than those who have completed higher levels of education [7,20]. While relatively clear information is found about the incidence of drug consumption among Roma communities, very little knowledge has been generated on the specific processes of rehabilitation. The present article is meant to expand on this gap. --- Mental Health Problem Incidence among the Roma Research has already shown that experiences of discrimination and social exclusion are detrimental to mental health. While a reactive system responds to social exclusion by activating defense mechanisms, social support stimulates reward mechanisms. Defense mechanisms activated in response to episodes of exclusion induce intense stress responses with detrimental effects on mental health. However, supportive networks can reverse defensive reactions, such as startling, or inhibiting the negative effects of depression [3]. Despite these clear-cut connections, little research has been dedicated to exploring mental health problems among the Roma. Along similar lines, although recent studies [3,21] have proven the negative impact of the consequences of the financial crisis, and more particularly, house evictions on the mental health among the Spanish people, the impact of exclusion and poverty on the mental health of the Roma community has not yet been contemplated. Studies looking at mental health problems among the general population show the highest risk of depression or anxiety among women [3], and an even a higher incidence is reported among Roma women compared to non-Roma women [7]. For unhealthy habits, Roma women stand out for the highest percentage of alcohol consumption and the lowest rate of physical activity. However, there is evidence that they smoke less for cultural reasons [7]. If there is little information about the incidence of mental health problems within Roma communities, no research effort before has focused on identifying which particular processes can contribute to combat it. It is precisely this gap that the present research aims at fulfilling. --- Materials and Methods --- Communicative Methodology Design A communicative qualitative study design was selected to conduct the present research. The communicative methodology [22] allows for the establishment of an egalitarian dialogue between the researchers and the study participants, leading to a process of dialogic creation of knowledge. While the former are responsible for providing scientific knowledge, the investigated people contribute their ways of life [23], their experiences, their life trajectories and their ways of understanding the world. This communicative paradigm opens the doors of research to groups such as the Roma people, who have often been excluded from the processes of the creation of scientific knowledge [24], and it has been identified by the European Commission as being especially suitable for working with disenfranchised communities [22,25]. The authors followed a qualitative strategy to collect data, as it was necessary to capture the perceptions, experiences and opinions of those members of the church who have been exposed to cases of detoxification and mental health problems. The fieldwork has been developed during 2017. --- --- Pseudonym Position Profile Description --- Rebecca Female Pastor A 43-year-old married Roma woman who is a mother of 5 children and a grandmother. Rebecca works as a street vendor and lives with her husband and three of her children. She was baptized when she was 12 years old and has been a female pastor for approximately 12 years. Rebecca has closely experienced the lives of her sister and sister-in-law, two women with depression problems who participate in the IEF. --- Jonathan Pastor Jonathan is a 60-year-old Roma man who is married, a father of 4 children, and a grandfather. He is an old-age pensioner who lives with his wife and one of his daughters. Jonathan has been participating in the IEF for 37 years and served as a pastor for 32 years. He experienced a process of detoxification himself. Pedro None Pedro, a 47-year-old Roma man is married to Carmen, 49. Together, they have two children, and they are grandparents. Both are ex-convicts, and they have been Carmen None participating in the IEF for about eight years. When they started, Carmen had already left prison and Pedro was still on parole. After 20 years of consuming drugs, Pedro decided to abandon drugs completely. --- Isaac Pastor Isaac is a 25-year-old young Roma man who has been a pastor in the IEF for a year. He is married with three children ages 6, 4 and 2 years old. He has attended church since he was a child because he was raised in a Christian family, and his father is a pastor, too. Isaac is a street vendor, and has now entered college. Isaac says that he is the youngest of four brothers and that, although he has never been a drug addict, he has had contact with drugs since a very early age. Table 1. Cont. --- Pseudonym Position Profile Description --- Samara Female Pastor Samara is a 35-year-old Roma woman. She is married with three children ages 16, 11 and 5 years old. She lives with her husband and her three children. She is a street vendor and would like to be a psychologist. She left school at the age of 11 and is currently participating in a university access course for people over 25. She has closely experienced the case of her mother, a woman with depression problems who has sporadically participated in the church. --- Rocío None Rocío is a 23-year-old Roma woman who lives with her partner, is pregnant and has a daughter from a previous relationship. She participated in the IEF as a girl with her family. Currently she is out of work, and she has completed basic studies. She has overcome depression. --- Isabel Female Pastor Isabel is a 44-year-old Roma woman; she is married, a mother of five children and a grandmother. She lives with her husband and three younger children. She has been a female pastor for approximately twelve years. She works in the cleaning sector. She has been participating in the IEF for 20 years. --- Ana None Ana is a 45-year-old married Roma woman and the mother of three children, ages 11, 10 and 6. She has always been dedicated to street vending, but right now she does not work. She has finished basic studies. She began to participate in the IEF at the age of fifteen. She says that she started because she has always had nervous problems, and her participation in the Pentecostal denomination helped her mental health, including during serious problems such as having her daughter diagnosed with leukemia. --- Data Collection Communicative everyday life stories were used. The CELS consists of a collaborative dialogue between the people investigated and the researchers to reflect together on specific situations relating to the concept being studied [28]. --- Analysis According to the Communicative Methodology, the analysis of the information is based on identifying the transformative and exclusionary dimensions of the target reality. While the former are all those elements that tend to promote rehabilitation and to recover from mental health problems, the latter are all those elements that contribute to maintain to mental health problems and drug abuse. The present article presents all those elements identified as the protective factors that allow detoxification and positive mental health. --- Results In Table 2, a summary of all the identified protective factors are presented. Although for the purpose of clarity, we have separated those more connected to drug consumption from those more related to mental health, our data indicates that they are all strongly connected. --- Detoxification The resources that the IEF provides for combating addictions stood out especially during the 1970s and 1980s when the consumption of heroin was booming, so the impact on detoxification was very noticeable. Although there has been a decrease in the consumption of heroin and other drugs in the Roma community, rehabilitation processes are still in need. In our analysis, a set of anti-drug discourses have been identified in the three churches that are recognized as playing a key role by our study participants. The first one consists of perceiving one's own body as a temple of the Holy Spirit, so it becomes essential not to contaminate the body with harmful substances such as drugs. The consumption of drugs is, therefore, forbidden, since it represents an attack on the integrity of the body. This is one of the most recurrent arguments already highlighted in the literature that explains why drugs are prohibited [12] within this denomination. The incorporation of these norms in the day-to-day of the members of the IEF contributes to overcoming addictive behaviors, and thus, to initiating detoxification processes. The same rationale also prevents church members from committing suicide, as will be argued in the next section. Our subjects' testimonies shared how their conversion process implied the internalization of a second important anti-drug discourse, that is, how drug trafficking is considered to be incompatible with being part of the IEF, because it is an illegal activity. For this reason, among many other respondents, Pedro not only gave up his addictions but also stopped drug trafficking. Pedro and Carmen trafficked drugs, which in their case meant that both had been sentenced to prison. This rehabilitation goes beyond health and promotes a radical transformation in their new lives. Thus, a third anti-drug discourse is widespread, which is how social recognition is connected to respect due to adherence to Pentecostal norms. Therefore, someone who does not respect the norms will not have a positive image within the IEF and the Roma community in general. This is the case of those who engage in drug consumption or trafficking, and therefore, the effect of the social control is to persuade people to stop such practices. In these cases, penalization measures are envisaged, such as not receiving communion, the sacrament in which those publicly baptized participate during Sunday services. Likewise, they will not be able to hold positions of responsibility such as that of pastor, since they are contaminating "their temple." Jonathan is a 60-year-old Roma man, married, father of 4 children and a grandfather. He has been participating in the IEF for 37 years, 32 years of which have been as a pastor. He is part of the group of Roma men who experienced the conversion process during the 1980s when he was a drug addict, and after detoxifying, he became a pastor. "God made you free, he has given you the freedom to choose, he does not force you, but he calls you. The Iglesia Evangélica Filadelfia is a delegation of God on earth to call you, and through the Iglesia Evangélica Filadelfia God calls you to put your life in order and to get out of the existing mud that brings the sin that there is into the world and that poverty often makes you embrace, because poverty sometimes makes you embrace sin in order to survive. The Church that announces and proclaims a message of cleansing is the voice of God, so that believing it, you can come out from the bad and unhealthy habits." It should be noted that a similar process is also recognized beyond the church, reaching the Roma community in general. This connects with the idiosyncrasy of the ethnic minority, which, regardless of economic status or level of education, gives greater recognition to people who respect Roma values. Our study participants explained how since their conversion they have accomplished a personal revaluation incorporating the Pentecostal norms, among which were the rejection of drug consumption, the condemnation of violence and the abandonment of illegal practices. As part of these anti-drug discourses, and recognizing groups at risk such as the youth, the three Churches organize a wide range of leisure activities such as football matches, prayers for the children, Bible studies or dinners aiming to keep the youth engaged and active there, moving away from harmful contexts and risky circles. The anti-drug discourse is embedded in many activities, including those oriented toward the children. Samara is a 35-year-old Roma woman, who is a female pastor, and emphasizes the importance of educating children about rejecting drugs. In her narrative, she refers to the case of her 11-year-old son: "My son is being raised in the Gospel which is a healthy environment, in which drinking or getting high with drugs is not often seen, nor going out at night and returning at the end of the morning, because in my house it would be inconceivable that my son would do that-never say never-but likewise my child and the children around, too, know that God demands that they stay clean, because our body is a temple of the Holy Spirit, so we have to keep it clean. It is something that God demands of us, and our children grow in this environment." --- Supporting Environment For many of our study participants, their initial involvement in the church represented entering a supportive environment that allowed them to make crucial decisions in their lives. One example of this was Jonathan, who clearly identifies his participation in the IEF as a key element in overcoming his drug addiction. Jonathan identifies in his conversion process a personal transformation that has allowed him to change the course of his life. Most of his previous friends died due to drug addictions or are in prison for illegal activities. He is convinced that had he not joined the IEF, his fate would have been that of his friends. Jonathan described his first contact with the IEF and how he found the needed conviction to giving up drug consumption: "I realized that the road I was taking was not the right one, and when I arrived at the Iglesia Evangélica Filadelfia, I met people who had gone through my own personal situation and some of them had come out of the anguishing situation in which I found myself, and they were for me a sort of a mirror where I looked at myself. I also found support from people who felt sorry for me and who God put in my way to help me out, to know how to advise me, to have patience and with the help of God and with the love and affection of the brothers and sisters I was able to get out of that dark cave I was in." Jonathan's words highlights the support network that is represented by the community of church members. Both cooperation and complicity are key factors for the consolidation and success of the rehabilitation processes. The affection and support found in the community help to reverse the isolation, loneliness and social refusal they have previously faced. In this regard, Jonathan notes that the support of the brothers and sisters in the church helped him out of "the dark cave" where he was. Isaac is a 25-year-old young Roma man who has been a pastor in the IEF for a year. He says that even though he has never been a drug addict, he has had contact with drugs since very early ages. In fact, he admitted to having smoked cannabis when he was 12, and when he was 13, he snorted cocaine 3 or 4 times. Therefore, there was a high degree of risk that Isaac would end up becoming a drug addict as many of his peers did. One of his brothers died because of drugs. However, he notes that his active participation in the church since he was 15 acted as the impetus that allowed him to get away from drugs: "At 15 years old, I was not a rascal but I was a sort of a dissolute lad, and I was aware that this would not end well A Roma pastor entered the church and simply bet on me. To be baptized in the Iglesia Evangélica Filadelfia, you are asked for some requirements that show you have converted to the Gospel, that you see a change in your life. The pastor came and baptized a few young people from the church, except me. At that time, I smoked, I wore earrings, I was a scoundrel... He had been working with me for quite awhile and he had very strong feelings for me, which always I observed, and talked to me about God and that I would regret it and that I could not continue like this. This man spent 15 months as a pastor in the church, and at least every time he saw me he said, 'Man, you have to convert, you have to start following God, your house is a Christian house and you will end up very badly.' I did not pay attention to him, but his words remained in my subconscious, and I thought he was right. In addition, I remember that once he explained to me an anecdote that he imagined that I was walking along the edge of a wall, and he told me you are walking, but at any moment you fall and hurt yourself, stop walking around here and start walking on the mainland. All those words remained with me in my subconscious." For Isaac, the pastor's trust and support were crucial for him to make the decision to be baptized and to change the course of his life path, giving drugs and tobacco up in order to actively participate in the church. Isaac notes that for him it was very important that his pastor believed in him and that he could change. It is important to consider the impact of this support for this young Roma, who comes from depressed contexts with high levels of drug consumption among his peers. The church represented for this youth an alternative social space where new relations, new habits and role models could be found. --- Becoming Role Models The church is full of people's stories of personal transformation, and thus, these persons become role models for the newcomers. They serve as examples for the rest of the members of the church and promote a life pattern away from drugs. Thus, those who are in the process of transformation are highly motivated to become role models and to prove to the community their effort to move away from difficult situations. Next, we find Isaac's father Pablo's story. Pablo was a man with a serious cocaine addiction. He also began to participate in the IEF during the 1980s and represents another example of the many detoxifications that occurred at that time. Isaac notes that his father was an orphan at a very early age, and it hurt him a lot because the father figure is very important in the Roma community. When he lost his behavior model, Pablo began to take cocaine very often. With a serious addiction, he joined the IEF and began his detoxification process, which he successfully completed in six months. Once detoxified, Pablo began his training process to become a pastor, and with that, his story became an example to follow within the church. Isaac speaks about his father's rehabilitation process: "My father at 31 was a young rascal and a crazy lad with no one around to stop him; he started ingesting cocaine every day and he got addicted, he was hooked for many years, and this had repercussions for gender violence on top of many other problems at home I have not lived it but they explained it to me that my father was nervous My father went to my mother's town because he wanted to have different friends because he believed that would change him. When my grandmother arrived here, she was a very good Christian and always spoke to him about God. They preached the Gospel to her at church. It was not until my father decided to give God a chance, what in fact occurred was that God was giving it to him. In addition, it completely changed my father. In only 6 months, my father was already raised from candidate to pastor." --- New Social Relations The last protective factor identified is connected to the improvement of relationships within the family as well as beyond. This is the case with Pedro, a 47-year-old Roma man who has been participating in the IEF for approximately 8 years. Pedro started very young in the world of drugs and had been taking drugs for 20 years when he first started participating in the church. His partner, Carmen, describes the impact on Pedro's health: "My life changed completely. When I observed him, he was also sick, and I observed that with the worship his illness was gradually disappearing. Until there arrived a moment when the doctor said that Pedro was cured, and that it appeared to be a miracle. He had Crohn's disease and because of that, he gave up medication. From then onwards, there has been a tremendous change at home. I cannot believe it, the way he was and how is he now." Parallel to his detoxification, Pedro explains his recovery from Crohn's disease as he was back to medication. Thus, due to his new life, he also recognizes improvements of his emotional relationship with Carmen. Both refer to more prudent behavior, a rejection of violence, a better use of economic resources and an improvement in trust in the couple: "A complete change, I had been hooked on drugs for 20 years, whereas today I do not even smoke. Then, in my life, I thought that I would never see myself like this, because I said, I will be 80 and still be like this, however I do not even smoke, so I have given up all my addictions. I do not leave my house, I go out in the morning for a while to the park, for a while with the birds, then to my worship and from here to my house and I do not go out anymore For me, the biggest thing is that I was able to give up all drugs, tobacco, methadone, I have left everything. I think that today, I still do not believe it, you think that you have been hooked for 20 years, and here since I have been in the cult for 6 or 7 years, it has cleaned me up completely. Not a pill to sleep, not even diazepam, that normally that everybody takes it currently, not even that I take to sleep." Improvements in his relationships have been reported beyond his marriage. Pedro explains how he made new friendships, and thus, he stopped seeing those with whom he used to consume drugs and going to these places. Pedro notes that he moved from being surrounded by drug addicts to getting associated with people who participated in the church and who have a clear rejection of drugs. This new socialization process is also a factor of social pressure that reinforces the change of lifestyle and, consequently, the processes of detoxification. However, the fact of integrating into the congregation, assuming and internalizing Pentecostal values and creating meaning that represented their participation in the church have been sufficient to achieve detoxification. --- Mental Health --- Promoting a Wellbeing/Happiness Discourse Many of the values and habits spread by the three analyzed churches are aligned with healthy mental wellbeing. Our participants highlighted how their participation represented a reinforcement of their self-esteem. Thus, Pentecostal values emphasize the importance of kindness or bravery, downplaying insecurities. Religious beliefs directly collide with considering committing suicide at any point, and support and companionship is displayed to ensure that people do not consider it as an escape route from their situation. Rebecca, a 43-year-old Roma woman, shared the case of her sister-in-law, a woman with depression, who at the time she received the news that her youngest daughter had leukemia, decided to commit suicide. However, she explains that the constant support in the church was crucial for her: "Well, they care a lot for her, 'Calm down' they say, and they give her advice, 'that the Lord is with you, you have to get out of this, look at the miracle that God has done to your daughter' and they are with her, and they advise her. How do you think that this whole process would have been with your daughter's illness if she had not been in church? She said it with these words, 'I would have gone to the 13th floor and jump out.'" Some of the core values of the church act as a protective factor against suicides. For instance, there is the belief that there is life after death, and therefore, members of the church who live according to Christian ethics aspire to be welcomed into the Kingdom of Heaven after they die. However, for those who commit suicide, the expectation is different, as they would be denied entry into the Kingdom of Heaven. This conviction plays a crucial role in suicide attempts among church members, especially in the face of the loss of a loved one. When Rebecca 'sister in law suffered the death of her daughter, the conviction that those who commit suicide cannot reunite with their loved ones in the Kingdom of Heaven led her to desist from her suicidal plans. When a case at risk for depression in the church is identified, this support network sets in motion a number of measures. Samara narrates her mother's case. She was at the time immersed in a deep depression that led her to consider taking her life. Her participation in the church and in spiritual activities has led to clear improvements in her mental health. She has even abandoned the medication that kept her in bed to regain her normal rhythm of life. During the religious service and in the interactions between the members of the church, there is an emphasis on the improvement and empowerment of oneself. They cope with the fears and complexes by counting on a "God Almighty" and the members of the church who intervene as a guarantee of success. This speech has a significant impact on the expectations and the day-to-day lives of the members of the IEF. As already mentioned, pastors, because of the spiritual authority they hold, are recognized as crucial points of support in difficult times. Jonathan explains that his brother-in-law came to him because he felt that this would be his last night. It is usual for prayers to be held together as a show of support and to give more strength to the petition. Jonathan describes how, through the prayers that he offered with his brother-in-law, he managed to overcome the feeling of panic that had invaded him. Providing support and companionship when someone is in trouble are resources identified to favor wellbeing and overcoming depression. Samara described the situation of a young girl with agoraphobia. Since she is the pastor's partner, Samara plays a crucial role in the care of women. She gets involved in solving problems that go beyond participation in the church. Likewise, the activity of the pastors contributes to healthiness within the congregation. "A girl had a phobia about going out, and I noticed that when I talked to her, it wasn't me, I noticed that I was speaking on behalf of the Holy Spirit, and every time I spoke to her, the girl felt better. She called me on the phone and explained what was happening to her at that moment, so I started talking to her about God, and she was right on the street and said 'Pastor, thank goodness you have spoken to me, how much goodness you have given me' and she carried on with her shopping, with her life." This companionship and advice can contribute to overcoming panic situations associated with mental problems such as anxiety or depression. --- A Sense of Belonging Religious services occur every day. Bearing in mind that most Roma women are unemployed, their daily attendance at church is the activity that breaks with the routine of domestic chores and family responsibilities. Thus, our respondents recognize that it forces them to take care of their image, to relate to other people beyond the family, and thus, that it has a direct positive impact on their self-esteem and personal well-being. Samara describes how the support network they find in the church helps them to better face personal and family problems and to cope with stressful situations connected to poverty or social exclusion with hope and more positive attitudes. Samara describes what advantages for the mental health of Roma women she finds in participating every day in the church: "Making up yourself every afternoon, leaving your environment, the monotony, I do not know . . . the change of our daily environment, going out every evening to congregate with people, to talk to one another." The positive interactions add up and reinforce the self-esteem of the participants. Samara described how participating in the church has helped her appreciate some of her virtues. At the same time, it has allowed her to develop skills such as public speaking. "I've been a very self-conscious girl. People told me I was beautiful, and I looked ugly, but really, I felt it, I was ashamed to speak in a loud voice, I did not have the security or the freedom to speak in front of a group of people and . . . I --- Discussion Evidence has been found of how the participation of Roma in the three studied IEF churches has contributed to the rehabilitation throughout the success of numerous detoxification processes and to the prevention of drug consumption. The IEF spread a set of protective factors that have been identified as key elements for the detoxification process, namely anti-drug discourse, a supportive environment, new social relations, and role models. Our data shows how these three churches become supporting environments to initiate radical personal transformations, as well as to generate new healthier relationships, and make our respondents feel like role models who are proud of their personal achievements and who are motivated to help others to become so. All these elements make these churches the most effective spaces for detoxification that have existed throughout these years for the Roma in Spain. This is also consistent with previous research that had already noted this role [10,17,18]. In this paper, we move beyond the previous research to present a systematic analysis of particular dynamics and discourses that have become protective factors. The social activities organized by the churches have been shown to be beneficial in preventing young people from being in contexts where they are at risk of drug use. Similarly, Adamczyk and Felson [29] illustrated how the participation of adolescents in activities organized by religious institutions, such as sports, have a positive impact on their health. They found feelings of wellbeing, a better physical condition, a reduction in the consumption of alcohol and drugs, as well as a delay in the age of initiation to sex. Our results are thus consistent with these studies, contributing to the wealth of research that shows how religious participation tends to encourage healthy habits [29] and therefore prevent risky behaviors. The churches analyzed here, besides being social spaces where preventive interactions occur, all contribute to spreading an anti-drug discourse that keeps their members away from addictions and drug use [9,10]. Thus, the IEF tends to project an ideal of a Roma individual who lives according to Roma culture and Christian values and who refuses drugs and any illegal activity. This ideal type has penetrated beyond the church and has also been embraced by other members of the Roma community. This success has triggered the proliferation of detoxification since the beginning of the penetration of the IEF into the Roma community in Spain in the early 1980s. This role has been maintained and strengthened throughout the years. These churches are also safe spaces in which mental health problems are approached, a finding that is also aligned with previous research [30]. Nevertheless, the specific benefits of participation in the IEF for the mental health of the Roma minority have not been explored yet. The present paper fills this gap, and we inquired about the benefits of participation in the IEF regarding the counteraction of the negative effects of exclusion on the mental health of the Roma people. Social engagement is vital for the survival of many species. When people are socially excluded or have a greater sensitivity to rejection, four fundamental needs are shown to be affected: belonging, self-esteem, control and meaningful existence, which are required for human survival and effective social functioning [31]. Here, we argue that the IEF enhances their members' feeling of belonging, in this case, in the Pentecostal community. According to research, the feeling of belonging improves the conception of one's own social identity [32,33]. This is a particularly interesting aspect in minority groups such as the Roma community, who experience a strong burden of stereotypes, discrimination and rejection. All these advantages reverse the negative effects of exclusion such as isolation or loneliness that can lead to depression or anxiety [32,33]. Furthermore, the importance of reinforcing self-esteem and self-confidence has been outlined as crucial for our respondents' lives. They all agree that since their participation in the church, they felt more kind and brave, and thus, they all recognize improvements in their communication-related skills. It has been noted how the development of communication skills improves interactions and mental health while encouraging empowerment and freedom. In the same vein, a longitudinal analysis in Taiwan recognized that participation in religious movements or secular activities organized by religious institutions promoted an improvement in mental health and reduced symptoms of depression [32]. Recent empirical works on suicide [34,35] show that greater social support and attachments are related to fewer depression symptoms and other mental health issues among very diverse populations. Improvement in the feeling of belonging and promotion of interactions and social cohesion among those who profess religious beliefs suggests a lower percentage of depression, suicide attempts, and feelings of anguish and a higher level of wellbeing [36][37][38][39]. Thus, research indicates that people active on a religious level present a better state of health than those who are inactive [33]. Our research adds consistency to this argument by showing that active participation in religious movements is related to the promotion of a healthy lifestyle, with the consolidation of support networks and the development of strategies to cope with stress in the best way possible. All these advantages are clearly related to the improvement of mental health. In addition, this participation increases social cohesion within the community, overcoming loneliness and contributing to wellbeing. Particularly, we noticed preventive factors for depression and suicide in the analyzed IEF. The daily participation in the church implies leaving home and breaking with the routine of many Roma women who mostly devote themselves only to domestic chores. It is well recognized that an improvement in taking care of one's image benefits self-esteem and wellbeing. Additionally, participation in the church enhances interactions and promotes networks of solidarity and support. Besides all these factors, our research is also consistent with recent studies that highlight the crucial role of churches in reinforcing positive attitudes towards standard medical treatments [40]. In the particular case analyzed here, our testimonies highlight how their participation in the church helped them to follow medical assistance; however, they also recognize how a minority of members would not fully agree with these statements. This is an unexplored reality that should be further pursued in future studies. --- Conclusions There is still a long way to go to fully flesh out the crucial role played by the IEF within the Roma community in Spain. This article is an attempt to shed light on its particular role in rehabilitation and preventing drug consumption and mental health-related problems. We note the role of the three analyzed churches as transformative platforms that have resulted in an improvement of their members' lives. The successful acceptance of the Pentecostal denomination by the ethnic minority also justifies that the contributions of the IEF have transcended beyond their members to be adopted also by the Roma community in general. Therefore, the healthy lifestyle promoted by the Pentecostal denomination has also had an impact on the prevention of drug use more broadly in the Roma community. Similar effects have occurred with other excluded minorities when adhering to Pentecostalism [15]. These three churches have also represented a springboard for the social participation of the Roma community. This is particularly important if we bear in mind that the ethnic minority faces numerous barriers to their participation in social life [41][42][43], a process that might positively correlate with being away from drugs and being mentally healthy. The identified factors against drug addictions and mental health problems have been identified. Banning of the use of drugs is the main norm to explain the effect of these three churches belonging to the IEF on the processes of detoxification of the ethnic minority. Beyond this norm, support networks, socialization against drugs, and positive role models are identified as protective factors. Thus, the promotion of a healthy lifestyle that overcomes exclusion also affects mental health problems. Therefore, preventative factors affecting depression, suicide attempts, poor self-esteem, and phobias have also been recognized. These factors contribute to the promotion of interactions, the consolidation of support networks, the reevaluation of oneself and the development of skills and competencies that foster the improvement of personal self-confidence. The IEF discourse breaks with the stereotype that links the Roma identity with exclusion, poverty and criminality. In contrast, it presents an ideal model of a Roma person who follows the values of Pentecostal ethics but also Roma values. The present research represents only the tip of an important iceberg still understudied. There are still many related aspects both regarding processes that improves health status of the Roma, as well as the role played by the IEF within it, which have never been systematically approached. Their future analysis would contribute to reinforce those transformative and protective factors that make possible for an increasing number of Roma individuals and families to carry on the flourishing lives they wish to have. --- Author Contributions: R.F.G., J.A.L and T.S.M. conceived and designed the research; J.A.L. performed the research; R.F.G., J.A.L. and T.S.M. analyzed the data; R.F.G. and T.S.M. contributed with thorough revisions and analysis; J.A.L. wrote the paper. ---

Background: High incidences of drug consumption and mental health problems are found among the Roma population in Spain, a reality that remains understudied. Past studies have indicated the positive role played by the Iglesia Evangélica Filadelfia (IEF) in promoting rehabilitation and prevention of these practices. Objective: In this article, authors analyze in which ways the IEF favors processes of drug rehabilitation and mental health recovery as well as the prevention of these problems among its Roma members. Methods: A communicative qualitative approach was developed. It was communicative because new knowledge was created by dialogically contrasting the existing state of the art with study participants. It was qualitative because everyday life stories were collected, gathering the experiences, perceptions and interpretations of Roma people who are actively involved in three different IEF churches based in Barcelona. Results: This article identifies these protective factors: anti-drug discourse, a supportive environment, new social relations, role model status, the promotion of interactions, the revaluation of oneself, spiritual activities and the improvement of the feeling of belonging and the creation of meaning. Conclusion: The present research contributes new evidence to the current understanding of the role played by the IEF in improving Roma health status and how the identified protective factors can contribute to rehabilitation and recovery from such problems in other contexts.

INTRODUCTION The pandemic due to the novel coronavirus 2019 , which causes the disease now known around the world as 'COVID-19', has spread throughout the globe infecting almost 17 million people and causing over 660 000 deaths V C The Author 2021. Published by Oxford University Press. All rights reserved. For permissions, please email: [email protected] . We are increasingly aware of COVID-19's differential impact on vulnerable populations, including their physical and psychosocial well-being, exacerbating preexisting systemic health and social inequities . In Latin America, countries such as Ecuador have been severely hit, with initial reports sharing dramatic images and descriptions of corpses lying at home or left on the streets because funeral services and the local government of a major city were not able to adequately respond, in early March . By 11 October, there were 691 confirmed deaths per million people, placing Ecuador in ninth place in this category worldwide . It could be argued that the 'coproduction' of services between government and community-based organizations and the 'cooperation' of citizens in following measures in response to COVID-19 have been privileged over actual participation of civil society in decision-making. In more extreme contexts, lockdown was enforced with the aid of the military ; while in countries such as the United States the response effectively became militaristic, exclusionary and anti-democratic . In Lebanon, shortly after protests, calls for military intervention in the COVID-19 response were announced . Ecuador, with a population of 17.5 million people, faces this global pandemic already weakened by a severe economic crisis, due in part to large external debt commitments, government corruption and low oil prices. While adopting the US dollar is credited with stabilizing the economy at the time, over the years it has limited the country's ability to dictate fiscal policy . More recently, the government's attempt to increase fuel prices prompted large protests that drove much of the country to a halt in October 2019. Public revolts led to violent repression of protesters by police forces with support from the military, reinforcing their traditional role of social control . The legacy of authoritarian governments has meant that civil society and social organizations in Ecuador have been pressed to develop 'coping strategies' to compensate for or counteract their limited participation in decision-making . After almost two weeks of protests, the Ecuadorian president agreed to withdraw its proposal, further weakening an already unpopular administration battling low credibility and people's support. In this context, Ecuador was one of the first Latin American countries to identify a confirmed case of . Three weeks later , the government decided to impose a national lockdown . However, COVID-19 spread rapidly in the initial hotspot and moved on relatively quickly throughout the country . As of 28 July, Ecuador has 82 279 confirmed cases and 5584 deaths associated with COVID-19 . Quito, the country's capital, has surpassed Guayaquil with the largest number of confirmed cases . Ecuador has a segmented and fragmented health system guided by a vertical biomedical approach, which has been shown to perpetuate health inequities and limit community participation . Such an system echoes global health shifts toward a biomedical, disease-focused orientation that does not consider the values of social justice and equity, and therefore limits community participation, even though others argue it is a right that should be protected . Concurrently, the scientific community has been all but absent from governmental response efforts, and the country continues to confront the pandemic without the leadership of an expert health committee. In contrast, a group supporting the risky, unproven treatment of COVID-19 with chlorine dioxide had ample time to address the national legislature, in tune with 10 bishops 'demanding' the ousting of the Ministry of Public Health for refusing to authorize this chemical for treatment . During the study period, confirmed COVID-19 cases rose from 28 on 14 March to 35 484 on 10 June, with only 4765 RT-PCR tests per million people . On 1 July, Ecuador had one of the highest number of excess death rates during the pandemic among countries that report that figure, having reached more than 2000 per million people or 88% above normal . According to Imperial College simulations, without intervention, by the beginning of February 2021, there could be a total of 175 000 deaths due to COVID-19, that is, 10 000 per million people . Although Ecuador is classified as an upper-middle country, the current pandemic has exacerbated preexisting, systemic social, economic and political issues, and exposed the state of the country's healthcare delivery system. According to the World Health Organization, Ecuador's preparedness is at level 3 , which means reaching 60% of capacity benchmarks . It should not surprise us then that, despite having considerably more resources, Guayaquil was overwhelmed by the onset of the pandemic. And, as the virus moved across the rest of the country, other localities have found themselves facing similar challenges with fewer resources. The scars left by the pandemic in the health sector include a series of corruption scandals related to the purchasing of emergency supplies at exorbitant prices, which, together with previous contracts, add to more than 20 'fraudulent' business arrangements tracked by the National Anti-Corruption Commission . To make matters worse, the Ministry of Public Health has had two different Ministers during the pandemic, and continues to struggle with provision of personal protective equipment for health personnel, testing and tracing capacity, with limited clinical and public health resources, making the outlook of the pandemic response highly uncertain. Moreover, early on in the government's response to the pandemic, the leadership role of the National Emergencies Committee was vested upon the Secretariat of Risk Management . But even the Secretary of Risk Management had to step down amid accusations that she authorized the purchasing of food kits at inflated prices, which was subsequently verified by Ecuador's Office of the Comptroller . Although on 8 April there was a call for attention to indigenous populations during the pandemic , the Pan American Health Organization waited 2 months to publish related guidelines , and sent a belated alert on 15 July , after COVID-19 had begun to ravage Ecuadorian indigenous populations in the Amazon River Basin. Finally, while there were reports of an increase in intimate partner violence during national lockdowns around the world, Ecuador saw a decrease of related emergency calls, which was implicitly explained by the government as due to a lack of community-based mechanisms of support . In this complex scenario, and given the extended first wave of COVID-19 in Ecuador, the present study critically examines whether decision-making in Ecuador's governmental response to COVID-19 has taken into consideration-or omitted, negated or distorted-community participation, a value and tenet of health promotion. The study begins by identifying the policy documents that defined the approach to decision-making, including the institutional stakeholders involved, the process of decision-making and the focus of decisions that were made. Then, it centers on a systematic textual analysis of the almost daily formal resolutions of Ecuador's Emergency Operations Committee . Accordingly, recommendations are made toward improving healthfocused or health-related decision-making in Ecuador by creating and supporting conditions for a communitybased health promotion to respond to either random events such as a health emergency or the persistent health-related needs of the population. In particular, our interest lies in the limits and possibilities for building community-based policy making, assuming that its framing evolves and is dynamic , especially during a health emergency of the nature and scale of the COVID-19 pandemic. --- METHODS --- Analytical framework Health promotion considers 'community participation' as an integral value and a right leading to an improved public health practice aimed at greater social justice. However, it has been acknowledged that lack of robust data on its impact continues to undermine support for it . Among the weaknesses of the concept, is the fact that 'community' and 'participation' have not been defined with clarity or in a standardized manner within health . Also, heightened focus on consensus has replicated power relations that minimize the importance of having a plurality of opinions, which lies at the center of democratic, empowered participation toward greater social justice . In addition, since participation should be a choice , not an obligation, it brings about the possibility of non-participation as a form of participation itself . An interpretive approach to policy analysis assumes that policy is not value free , and that 'dominant political values' influencing public health decision-making are 'hidden' from us. This implies that critically analyzing community participation in the 'live' policy making process involved in responding day by day to a developing pandemic will allow us to understand whether it has been relevant or not, and how, for the government. If participation is the result and aim of an 'iterative learning process' [ , p. 79], its consequential inclusion in decisionmaking would involve accepting opinions can be challenged and, therefore, priorities may shift and strategies change, along with power relations. --- Analytic sample We downloaded and compiled 53 consecutive resolutions between 14 March and 10 June 2020 emitted by the National EOC that was convened in response to the COVID-19 pandemic in Ecuador, following the Guidelines for the EOC . These resolutions are available in the Secretariat of Risk Management website . Crucial policy documents that were cited in the resolutions, Decrees 1017 and 1019 , were also analyzed for context, as well as the guidelines that define the characteristics of the EOC . --- Content analysis --- Qualitative analysis We used thematic analysis to, in a first round, identify words and phrases through line-by-line open coding, and, based on these, concepts were grouped into categories and subcategories through axial coding. In a second round of coding, words and phrases were counted by mention per resolution, according to date. Simultaneously, extracts from the quotes were selected and translated into English. In a third, final round, subcategories were regrouped to correlated data according to themes through selective coding. --- Quantitative analysis We tabulated the number of times each major category was mentioned in each resolution. We then added the counts and calculated the average number of mentions per resolution. Finally, we explored the correlation between mentions per category corresponding to each of the dates the resolutions were released. --- RESULTS --- Quantitative results Table 1 shows the total and average number of mentions of key sectors among the EOC resolutions we analyzed. Law enforcement had the largest number of mentions of any sector , followed by the private sector . Although COVID-19 is at its core a public health crisis, health was only mentioned 20 times . We then explored the correlation between mentions per sector for the corresponding dates of each resolution . We found that the number of private sector and law enforcement mentions were positively correlated . --- Qualitative results --- Main themes Our thematic analysis identified three major categories mentioned across the resolutions analyzed ; private sector, social sector and law enforcement. The Private Sector included mentions of construction efforts , transportation , food and agriculture, and commerce. Of these, commerce and transportation were most frequently mentioned . The Social Sector encompassed a wide range of subcategories including: social and solidarity economy, Law enforcement sector consisted of mentions of presidential emergency decrees , police and military forces, control, undisciplined population and/or regional and local authorities, and punishment, fines and calls to action. Policy and military forces were most commonly mentioned , followed by presidential decrees . --- Private sector Although almost all of the participants in the EOC meetings have been representatives of the national government , the private sector was consistently mentioned across resolutions. One of the major concerns of the EOC has been securing the local food supply chain, and charging the public sector to coordinate with private companies to limit disruptions in the provision of food products across the country: The Ministry of Production, International Commerce, Investments & Fishing will coordinate, with supermarkets and production chains, the mechanism to prioritize the purchasing of Ecuadorian products . Another EOC concern has been securing public transportation services, particularly so that essential workers are able to go to and from work: The urban transportation system must remain operational across the country. Its frequency and number of trips will be lower than usual, but must not be suspended. A total suspension of transportation increases population risk, by preventing health care workers, security service workers, or food workers to reach their workplaces, which is inadmissible during this emergency . --- Social sector As mentioned earlier, while the social sector encompassed the largest list of subcategories, they were seldom mentioned across the resolutions we analyzed. Expectedly, given the public health nature of this crisis, the Ministry of Health was the most commonly mentioned subcategory, including how its senior leadership employees needed to come back to in-person availability: In reference to the working hours of the MOH staff, it is resolved . . . starting on April 9 . . . that all employees that make up the higher hierarchy of the MOH will be reintegrated in a permanent in-person way, working daily in the activities and tasks within their competencies . . . personnel from priority care groups will be excluded from these teams . . . in the family group where there is more than one person who must carry out a face-to-face day, and there is responsibility for children or people who require permanent care, remote work is authorized . This measure was revised on 15 April, to authorize remote work regardless if workers have child/elder care responsibilities. A similar measure was latter applied to Social Security Institute workers, but without a provision to work remotely for employees with family/caregiving responsibilities . In the Introduction section, we mentioned that the Secretary of Risk Management had to step down due to evidence that she authorized the purchase of food kits at inflated prices. It is interesting to see how 'food kits' are positioned in the resolutions . In the first mention of 'food kits' , there is language signaling the urgency of fast tracking approval at the national and local levels. Similar text is used again on 31 March, adding that: . . .the Public Company National Storage Unit is empowered so that, within the provisions of the public procurement law, its regulations and SERCOP [national public contracting service] resolutions, it carries out the necessary procedures to guarantee the delivery of food kits for families that require it nationwide. One of the most visible social sector measures taken throughout the period studied has been the return of Ecuadorians stranded abroad, particularly more vulnerable groups, for which the EOC released a 'Protocol for the entry into the country during the validity of the state of exception, of children and adolescents who are outside the country without their parents or legal guardians, pregnant women, people with disabilities and the elderly' . Other social protections mentioned included domestic violence, which was first brought up on 14 March, tasking the Secretary of Human Rights with addressing this issue. However, there didn't seem to be follow-up resolutions that would provide further details regarding its implementation. In turn, deaths related to the pandemic has been a critical item of discussion in EOC resolutions. In fact, this topic was the only one that merited an appendix providing a robust legal framework justification for police and military officers in the Guayas province to serve as witnesses and signatories of a death certificate, when the presence of a physician authorized by the Ministry of Health could not be secured. --- Law enforcement Our analysis suggests that law enforcement was robustly mentioned across resolutions in the study period. In other words, maintaining social control was a salient priority for the EOC. The response to the coronavirus in Ecuador was legally bound by Decree 1017 , which declared the state of exception on 16 March 2020, and Decree 1019 , which created a 'special security zone' in the city of Guayaquil, the first major epicenter of the pandemic. Both decrees gave law enforcement officials broad powers to monitor and enforce EOC orders and recommendations. By 21 April, some power was delegated to the 221 municipal governments in Ecuador by making them decide on the level of restrictive measures according to a 'traffic light' tool , and 'implement adequate mechanisms to comply with and control the different resolutions and dispositions approved by the National EOC for the management and handling of the sanitary emergency due to COVID-19' . On 23 April, the National EOC 'reminds' mayors 'that problems derived in the fight against the virus COVID-19 will be channeled and resolved through the National Emergency Operations Committee'. Importantly, EOC resolutions paint a picture of an unruly and undisciplined population depicting 'manifest civil disobedience' that justifies strong control mechanisms to be ultimately enforced by police and military forces. In a country where the minimum wage is $400 US dollars, 'people who fail to comply with the provision will be penalized, the first time, with a fine of USD100 . . . the second time with a fine equivalent to one month of minimum wage [USD 400]; and, the third time with prison, according to the procedures established by the competent entities' . Even regional and local authorities are considered unruly: Isolated, inconsistent decisions, impossible to apply or outside the framework of their respective powers generate confusion and chaos. We are asking the population to act with discipline in compliance with national instructions, this same request is made to the authorities of decentralized governments . Several arbitrary measures taken by local governments are confirmed . . . the municipality of Guayaquil was responsible for the violent intrusion of municipal agents on the airport runway . . . this type of action does not contribute to the management of the emergency. Accordingly, we make a call to the local authorities and to all citizens, in order to follow the official channels and act calmly and in coordination . In this context, it seems justified to heavily rely on police and military forces to monitor the correct implementation of EOC resolutions: Due to non-compliance by citizens on the restriction of mobility between provinces . . . the Armed Forces are ordered to tighten the control of the flow of private persons in the provincial limits of the entire national territory . --- And even extending their responsibilities beyond their training: The police will continue to carry out random controls. . .of people placed in mandatory preventive isolation . To order the national service of legal medicine and forensic sciences, the national police and the armed forces to lend their contingent in whatever is necessary for the application of the 'Protocol for the Manipulation and Final Disposal of Cadavers with a Background and Presumption of COVID-19' . --- Stakeholder participation A salient aspect of control is that most resolutions affirmed that decisions of the National EOC had been approved 'unanimously', and participation was for the most part limited to national government representatives. Regional and local government associations representatives were part of many but not all sessions, while civil society organizations did not participate in these meetings . --- Information control Information control was a key priority of the national government, reminding local authorities, time and again, of the preeminence of their decision-making: 'Municipal EOCs must go to or require official information from the Provincial EOC' , which in turn receive information from the national EOC. The 28 April resolution points out that provincial EOCs will 'permanently provide information to municipal EOCs to define traffic light color signal' and commits the national EOC to 'publish and update all available information such as statistics, protocols for different activities and other relevant data regarding the actual and evolving status of the pandemic, so that municipal EOCs can make informed decisions'. Local governments then will have access to 'relevant data' for decision-making through the provincial EOC, which will 'report using official data uploaded through technological tools' . Control of information went to the extent of denying or masking the trajectory of the pandemic when on 27 May, the resolution affirmed that 'there is no increase in the number of contagions in municipalities that have switched color [in the traffic light system], with a decrease in contagions, and control of cases, as well as the decrease in the number of deaths'. --- DISCUSSION Our study is the first, to our knowledge, to critically examine whether a country's governmental response to COVID-19 has taken into consideration-or omitted, negated or distorted-community participation, a value and tenet of health promotion. Our systematic textual analysis of the almost daily formal resolutions of Ecuador's EOC show that the 'lifecycle' of the central government's response and its evolving, dynamic policy framing has been centered around law enforcement, private sector, and social sector priorities, in that order. Although the social sector had the largest number of subcategories, they were exceptionally mentioned across resolutions. Moreover, we found little to no evidence of the participation of civil society or social organizations. A review of participant lists registered by the EOC show that ministries and other national governmental entities constitute the largest contingent of representatives at EOC meetings. In addition, representatives from the Red Cross, the Ecuadorian Municipalities Association, the Consortium of Provincial Governments, and the Ecuadorian Conference of Catholic Bishops have participated in these meetings. Hence it seems clear that, following Rifkin et al.'s key elements of community participation-action oriented, involving choice and having a potential effect on health-we find that, at least in Ecuador's case, there were more limits than possibilities of 'live' policy making grounded in a community-based response to COVID-19 . According to the Guidelines for the EOC , 'it is the duty of authorities and institutions of the National Decentralized System of Risk Management to recognize, facilitate and promote the organization and participation of ethnic communities, civil society, community, charity, voluntary and of common good associations' . Further, 'risk management processes must be respectful of the cultural nuances of each community' . In theory then, community participation in its broadest sense, i.e. including social protection measures that, presumably, would be in the agenda of the 'ethnic communities, civil society, community, charity, voluntary and common good associations' should be taking part in decision-making of the National EOC, as envisioned by its guidelines . In practice, however, we find no evidence that any social or civil society representatives have been involved in the EOC meetings and the release of EOC resolutions. Hence, the EOC decision-making process relies heavily on the feedback and judgment of national-level authorities with, to say the least, questionable public credibility. It is no surprise then that, in the absence of community engagement, the resolutions present the public as culpable, undisciplined, unruly and consequently, in need of a strong law enforcement response. Such an approach may be due in part of the erosion of public trust of the current administration, with low approval ratings, and still reeling from the October 2019 protests that paralyzed much of the country for almost two weeks. Although the previous government seemed to favor more inclusive mechanisms of development, such as popular and solidarity economy, it also eroded civil society trust by legally suppressing dissenting voices . In this context, it is somewhat understandable that EOC lacks civil society representation. Moving forward, we recommend EOC and other national authorities to rethink their approach to the pandemic by engaging in meaningful dialogue with civil society organizations, representing vulnerable and marginalized groups, to rebuild trust and increase the likelihood that EOC recommendations are more meaningful and actionable to the realities of the Ecuadorian population, and moving away from law-enforcement-centered initiatives. A number of limitations must be acknowledged. First, our analytic sample of resolutions was limited to a specific timeframe , to allow for our thematic analysis to take place. However, given the ongoing nature of the pandemic, the EOC continues to meet and release resolutions almost every other day. After 10 June, sixteen additional resolutions have been released. Future studies should certainly incorporate these resolutions and any other that may be released until the pandemic crisis subsides. Second, due to the primarily qualitative interpretive nature of our inquiry, results may not have taken into account all aspects of such a complex crisis. Hence, generalizability of our findings is not to be implied. Third, our observations rely primarily on written resolutions and related public documents , and our expertise regarding the Ecuadorian realities of policy and decision-making. That said, we did not attend any of the EOC meetings or interview any of its participants. Future research should certainly consider adding these and other triangulation methods to incorporate additional vantage points of view. --- CONCLUSION Similarly to other countries, Ecuador implemented a vertical, militaristic response to the COVID-19 pandemic. The authoritarian legacy in Ecuador has eclipsed the potential for a participatory approach to health. In this context, mechanisms of community participation in policy making were mostly missing, despite the concept being explicitly guaranteed in guidelines for emergency response. Further, it is evident that governmental decisions to tackle the COVID-19 pandemic were not centered on the social sector, which is supposed to be a priority in a participatory, communitycentered approach. Not even the consideration that both participation and community-based efforts are essential to adequately confront the COVID-19 pandemic and its aftermath has been elevated to discussion. This means that, to the government's credit, the term 'participation' is not used for utilitarian or even symbolical purposes. Concurrently, in the absence of community participation, a health promotion approach was not embraced by the EOC in charge of the pandemic response. Moreover, the scientific community has been absent from decisionmaking, with information being distorted or interpreted subjectively. Without a culture of participation and open channels of communication, what was left for the government was an emphasis on enforcement, derived from a 'unanimous' perspective in which there was no options for challenging or disagreeing. This means that, for the future, assessing or evaluating the response presupposes that everyone-or nobody-involved in decision-making can be made accountable. Lack of checks and balances also involves the police conducting epidemiological surveillance and the Catholic Church being an interested stakeholder, lobbying for self-serving risky exceptions such as permission for gatherings in closed buildings, while no other nongovernmental organization that could have contributed toward greater social protection measures was participating in EOC meetings. Having legitimized the exclusion of community participation in Ecuador's response to the COVID-19 pandemic, the government will continue, unchallenged, to sustain decision-making and resolutions that do not consider their social implications. In particular, the limits to local governments becoming informed and making decisions without mediation by the National EOC will further impede community participation in health decisionmaking in the future. Consequently, local knowledge and experiences will be least likely to inform health policy. Addressing the gap between the intentions and realities of risk management policy in Ecuador will require that the central government must define more clearly the powers that local authorities should have during emergencies, but also in regular times, with regard to health promotion. There must also be clearly defined procedures and supporting mechanisms for community-based organizations to take part in meetings and decision-making processes, for which all public officials and participants involved should be publicly announced, together with relevant information on the outcomes, in a timely manner. Further, all centralized decision-making bodies, in the case of Ecuador, the national and local EOCs, should guarantee fully accessible information, together with established, transparent mechanisms for providing feedback and improving accountability. --- SUPPLEMENTARY MATERIAL Supplementary material is available at Health Promotion International online.

Global shifts toward a disease-oriented, vertical approach to health has involved limiting the right for communities to participate in decision-making. Ecuador's authoritarian legacy has forced civil society and social organizations to adopt 'coping strategies', while large protests recently derived into violent struggles. The country has been severely hit by the COVID-19 pandemic amid corruption scandals involving hospital and food purchases by government during the response. This study critically examines how Ecuador's government took into consideration 'community participation' as a value and tenet of health promotion. Our systematic textual analysis focuses on 53 consecutive resolutions by the National Emergency Operations Committee (EOC) leading the decision-making processes, which, explicitly requires community participation. Results show that the 'lifecycle' of the central government's evolving policy framing centered on law enforcement and the private sector, followed by the social sector. Further, there is no evidence of stakeholders from civil society or organizations taking part in decision-making. Having legitimized the exclusion of community participation in Ecuador's response to the COVID-19 pandemic, it is possible that the government will fail to consider the wider social implications of its impact. In particular, the limits to local governments becoming informed and making decisions without mediation by the National EOC will further impede community participation in health decision-making in the future. This implies that local knowledge and experiences will also not inform health policy.

Introduction Catastrophic events such as civil conflicts, epidemics, and other humanitarian emergencies have engendered significant scholarly interest in recent demographic studies . Part of this interest is driven by the need to understand how demographic shocks affect subsequent social adjustment processes. Accordingly, a number of studies now catalogue the specific impacts of wars, epidemics, and other emergencies on fertility, mortality, and migration processes . Studies on how the outcomes of children are affected by these events also point to a range of deleterious consequences of these crises for the welfare of children . In African societies, for example, the socio-demographic outcomes of children have been found to be negatively affected by the HIV/AIDS epidemic ; wars ; as well as famines . More recently, research on African children affected by these events has drawn attention to their effects on orphan status, and in turn, the effects of orphan status on schooling . Despite increased research on the negative consequences of orphan status in African societies, lingering questions remain about the specific ways in which the welfare of orphans is affected by other demographic processes. Thus, although the association between orphans status and children's living arrangements is now well established , previous research gives limited attention to the mediating impacts of family structure on the dynamics of orphans' schooling. Consequently, the logical nexus that exists between the impacts of humanitarian crisis on both orphan status and family formation processes has not been fully examined. Wars and HIV/AIDS, for example, can increase parental mortality which can affect both orphan status and whether or not orphans live in single-parent or two-parent families. These differences in family structure, especially for single-parent orphans, will generally depend on whether an orphan's surviving parent chooses to re-marry. As maintained by some scholars, a significant number of HIV/AIDS widows and widowers in some societies choose to remarry and have children in their new relationships . Since differences in family structure are known to have significant implications for the welfare of children the extent to which family structure influences the association between orphan status and schooling needs to be further clarified. This study, therefore, uses data from the 2002 Rwandan census to investigate the relationship between family structure and schooling disruption among orphans and non-orphans. In general, the study attempts to achieve three specific objectives. First, it examines the association between orphan status and discontinuities in schooling progress, with specific reference to the current non-enrollment status of children previously enrolled in school. In this process, explicit attention is given to the association between maternal, paternal, and double orphan status and schooling disruption among young children. As its second objective, the study locates the dynamics of schooling disruption within the context of the social influences of single-parent and two-parent families. In addition, the study identifies the extent to which orphans' schooling progress is associated with such factors as the sex of their household head and whether or not they are the biological children of their household head. The final objective of the study involves examining the relationship between household wealth and schooling disruption among orphans and nonorphans in different family contexts. The study, therefore, investigates whether household wealth has a differential association with schooling among paternal, maternal, and double-orphans in single and two-parent families. --- Genocide, HIV/AIDS, and Orphan Status in Rwanda Most studies seem to agree that at least 500,000 people were killed in Rwanda during the 3 month period of the genocide in 1994 . This, according to some scholars , was one of the most rapid rates of ethnic cleansing in modern history. Estimates of the impacts of the genocide also suggest that genocide-related deaths accounted for a 10% decrease in Rwanda's pregenocide population . As expected, the demographic impacts of the genocide were not limited to its impact of Rwanda's population size. Previous studies also highlight its deleterious effect on other demographic indicators , and describe disparities in these effects across specific regions of the country . Not surprisingly, Rwanda now has one of the world's highest rates of orphanhood . Nevertheless, although orphan status in contemporary Rwanda is directly associated with the extreme mortality conditions of the genocide, the impacts of other factors such as HIV/AIDS on increasing its orphan population cannot be discounted . Post-genocide research on children in Rwanda, however, gives more attention to the effects of the conflict on the outcomes of children. As a result, a number of studies now reveal the impacts of the conflict on children's health and nutritional outcomes ; their psychosocial outcomes ; while others focus on the social outcomes of children in Rwandan forced-migrant populations . Broadly speaking, orphan status is generally associated with less favorable socioeconomic indicators among children in Rwanda. Veale and Dona , for example, indicate that orphan status is a predictor of a higher likelihood of homelessness among Rwandan children in the years following the genocide. Siaens et al. maintain that Rwandan orphans are more likely to be poor and malnourished, but less likely to be enrolled in schools than non-orphans. 1De Walque also reports a similar schooling disadvantage among orphans. In Akresh and De Walque's study, educational attainment in the postgenocide years is shown to be lower among both orphans and non-orphans than in the pre-genocide years, although the relative disadvantage of orphans was consistent across time. As with the broader literature on orphans, however, the question of whether the apparent schooling disadvantage of Rwanda's orphans is consistent across family structure, as well as the broader influences of family contexts on children's outcomes, have not been systematically identified in previous studies. --- Contextual Issues A major determinant of the schooling outcomes of orphans is their degree of relatedness with their household heads. Accordingly, previous research consistently reports that orphans have better schooling outcomes when they live with relatives than with non-relatives . Attempts to account for the positive effects associated with living with relatives have generally been based on evolutionary perspectives on altruism. Thus, a number of scholars now discuss the importance of these relationships within context of Hamilton's rule that hypothesizes that altruistic behavior will be associated with degrees of relatedness between individuals . Schooling differentials that are conditional on orphan status are, therefore, seen to be a product of parents' decisions to invest more resources in the welfare of their biological children compared to their investments in non-biological children . In fact, a growing body of evidence consistent with these differential parental investments in children is now found in several studies . Within this evolutionary framework, therefore, the schooling disadvantage of orphans is considered to be a direct result of lower parental investments in their schooling compared to parental investments into the schooling of non-orphans. Despite the intuitive appeal of altruistic explanations, our understanding of the familial influences that account for the orphan schooling disadvantage is far from complete. Given what we know of the mediating impacts of family structure on children's schooling, part of the orphan disadvantage is likely to be explained by differences in family-structure. Non-orphans, for example, are by definition children with two living parents. Thus, their schooling advantage may not only be explained by greater altruistic investments from their parents, but also by their greater expected likelihood of living in two-parent relative to single-parent families. Living in single-parent families, for example, is negatively associated with parent-teacher contact and parental involvement in schools, both of which have a negative impact on children's schooling performance . Single-parent families are also associated with a lower quality of parenting and a higher exposure to psychological stress compared to two-parent families . Other effects of family structure on children's schooling are economic. For example, children in single-parent families are more likely to live in worse socioeconomic circumstances which can in turn explain their lower levels of educational attainment compared to children in two-parent families . Notwithstanding the positive association between living in two-parent families and children's schooling outcomes, the ways in which family structure affects the schooling outcomes of orphans has not been systematically addressed. As a result, there is also limited evidence regarding whether or not orphans derive the same benefits from two-parent families that have been found in previous studies. At the same time, orphans in two-parent families are likely to be the step-children or foster-children of at least one parent. As noted by Manning and Lamb , step and foster children in two-parent-families generally have worse schooling outcomes compared to their counterparts who are biological children. Analogous disadvantages have thus been found among orphans living with step parents in some African countries. Specifically, Nyamukapa and Gregson report that part of the schooling disadvantage of maternal orphans in Zimbabwe can be explained by the limited support they receive from step-mothers. Related to the effect of family structure are the effects of the number of parents deceased and the sex of the deceased parent. Among single-parent orphans, maternal orphans have been shown to have less favorable schooling outcomes than paternal orphans . Since paternal orphans have a surviving mother, while maternal orphans do not, these findings broadly imply that maternal influences on school enrollment among orphans are more important than paternal influences. Among paternal orphans, surviving mothers have also been found to take on an even greater role in the education of their children after the death of the father. In contrast, other studies suggest that the maternal effect on the schooling of paternal orphans may be undermined by the greater income loss associated with the death of a father . In families with maternal orphans, maternal deaths are likely to have a negative effect on schooling since children are likely to be used as substitutes for female household labor previously provided by their mothers. Not surprisingly, the impacts of parental deaths are even more dramatic among children with two-dead parents . Accordingly, a number of studies now show that the schooling outcomes of double-orphans are less favorable relative to those of their other orphan and non-orphan counterparts . An equally important determinant of children's schooling outcomes is household wealth. Conceptually, however, the direction of the wealth effect on orphans' schooling appears to be unclear. Some studies associate the impacts of wealth to the loss of parental income that comes with parental deaths . According to this perspective, the loss of income associated with parental deaths may be associated with children's increasing tendency to work outside the home to meet current consumption needs . On the contrary, other studies suggest that as a result of economic hardships, the likelihood of fostering an orphan is in fact positively associated with household wealth. This relationship is clearly intuitive. At the same time, it is still not clear whether living in wealthier households will decrease disparities in parental investments in children. In African countries, some scholars also argue that the economic impacts on the outcomes of orphans are moderated by the strength of kinship networks and the financial contributions of extended family members . In contrast, other studies also argue that kinship resources have very limited impacts when orphanhood rates are high . Children's schooling outcomes are also negatively affected by the number of children in the early childhood ages living within their families . Explanations for this effect are generally economic and are reflected in resource dilution theories that maintain that larger sibship sizes decrease the amount of resources available for human capital investment . Non-human capital investment explanations have also been given in other studies that suggest that school-age children will mainly work to care for their younger siblings as the number of younger siblings in their household increases. --- Hypotheses Based on the conceptual relationships found in previous studies this study examines two main hypotheses on the impacts of family contexts on orphan disparities in schooling disruption. First, the schooling disadvantage of orphans is hypothesized to be conditional on differences in family structure. Thus, differences in altruistic investments between single-parent orphans and non-orphans are expected to be conceptually smaller in single-parent than in twoparent families. This is because non-orphans are expected to receive more altruistic investments in two-parent families than in single-parent families . Consequently, disparities between single-parent orphans and non-orphans are hypothesized to be smaller in the single than in twoparent families. Similarly, although double-orphans have no living parent they are also expected to have a smaller schooling disadvantage relative to non-orphans in single-parent than in two-parent families. Secondly, since single-parent orphans are intuitively more likely to have step-parents or foster-parents in two-parent than in single-parent families, the study also hypothesizes that the disadvantage of singleparent orphans in two-parent families will be driven by whether or not they are the biological children of their household heads. --- Data and Methods The data used in the empirical analysis are taken from a 10% sample of the 2002 Rwandan census. These data are currently available in the Integrated Public Use Microdata-International database of the Minnesota Population Center . One clear advantage associated with these data is that they contain a larger number of observations compared to other surveys conducted in Rwanda in the years following the genocide. In addition, they contain information on a variety of demographic, educational, and household characteristics. Information on current school attendance is based on indicators identifying respondents age 6 and above who indicated that they were students during inquiries about their primary economic activity. Furthermore, for all respondents aged 25 or younger, the 2002 census provides information on whether or not their parents were alive. This information is used to distinguish between three types of orphans-double orphans, who have two dead parents; maternal orphans, who have a living father but a dead mother; and paternal orphans, who are defined as children with a dead father but a living mother. Following previous studies, non-orphans are defined as children who have two living parents. Children with missing information on the mortality status of their parents, i.e. about 3.5% of the sample, are excluded from the analysis. Since the study focuses on young children most likely to have been orphaned during the genocide in 1994, the final sample is restricted to children between the ages of 8 and 13. Members of this cohort were generally between ages the ages of 0 and 5 during the 1994 genocide. This restriction is important because including children born after 1994 in the assessment of the impacts of the genocide is likely to bias the presumed genocide effect downward. 2The Rwandan 2002 census data also provides information on several householdlevel indicators. For example, information identifying household heads and their spouses can be used to generate indicators of their respective demographic and socioeconomic indicators that can then be linked to child-specific information within each household using unique household identification numbers. In terms of wealth, the census provides information on the availability of various assets and amenities within each household. It is therefore possible to identify households that own automobiles, TVs, or radios, those that have access to electricity, the internet, and pipe-borne water supply, and households that don't. These data are generally important for two reasons. First, they are useful for assessing the comparative prevalence of specific household amenities in orphan and non-orphan households. Secondly, they are also useful for generating proxy indicators of household wealth that are essential to the analysis of how wealth affects schooling disparities by orphan status. Summary indices are increasingly being used to as indicators of wealth in countries in which conventional indicators, such as data on incomes or expenditures, are either unavailable or unreliable . Following previous studies, the study uses Principal Components Analysis to develop a living standards index that captures variations in household wealth. This data reduction strategy transforms data available in multiple dimensions into a one a dimensional indicator in a process that involves no significant loss of information. Analytically, the dynamics of schooling progress focuses on schooling disruption, or the likelihood of not currently attending school, among children who at the time of the census had already made some progress in their schooling. The final sample is therefore restricted to children who have completed at least 1 year of schooling. Admittedly, this focus presents only one perspective on the schooling outcomes of orphans since the impact of orphan status can manifest itself in other schooling outcomes e.g. enrollment. As mentioned earlier, previous studies have demonstrated that orphans in Rwanda are less likely to be enrolled in school than non-orphans. 3 Of interest in this analysis, however, is the question of how orphan status affects schooling progress among children already enrolled in school. In other words, beyond the impact of parental death on enrollment status, the study examines the possible ways in which parental death may be associated with disparities in children's schooling progress. The study's measure of schooling disruption is also a proxy indicator of current dropout status since non-enrollment is examined from the perspective of children for whom there is evidence of prior schooling progress . A similar approach is used in other studies using cross-sectional data to examine the likelihood of schooling dropout . Logistic regression analyses with household-level random effects are then used to examine how orphan status and family contexts affect schooling disruption. The general estimation equation used in the regression analysis is described in Eq. 1. The dependent variable, D ij , captures the logit of the probability that child i in household j is not currently attending school, given that child i completed at least 1 year of schooling prior to the census. This dependent variable, D, is considered to be a function of orphan status , a vector of child-level attributes , e.g. age and sex, householdlevel characteristics , e.g. Single-parent or two-parent and household size, and household-level random effects, l j . D ij ¼ a 0 þ a 1 þ b a O ij þ b b C ij þ b c H ij þ l jð1Þ Descriptive Results Summary characteristics of orphans and non-orphans in the sample are presented in Table 1. Young orphans in Rwanda are more likely to have lost only their fathers, , or both parents , compared to having lost only mothers. The actual numbers of orphans, by orphan-type, in single-parent and two-parent households are presented in Appendix 1. Regardless of orphan-type, however, Table 1 shows that orphans are more likely to live in disadvantaged household contexts compared to non-orphans. For example, non-orphans are the least likely to live in single-parent and female-headed households. In terms of household wealth, measured in national quintiles from the poorest-Quintile 1, to the wealthiest-Quintile 5 , non-orphans are also less likely than all orphans to live in the households within the lowest national wealth quintiles. Notwithstanding the comparative wealth disadvantage of orphans, however, other important variations in household socioeconomic conditions exist that are conditional on orphan-type. Paternal orphans, for example, are more likely to live in single-parent households than other orphans. This is generally consistent with reported decreases in the remarriage rates among Rwandan widows in the years following the genocide . At the same time, the lowest percentage of orphans living in female-headed or single-parent households is observed among maternal orphans, i.e. orphans who fathers are still alive. Apart from the socioeconomic disadvantages implied by the higher percentage of single-parent and female-headed households among paternal orphans, Table 1 suggests that these orphans are the most likely to live in households with the lowest levels of wealth. Thus, about half of all paternal orphans lived in households in the two lowest wealth quintiles. Among orphans, paternal orphans are more likely to live with their own biological parent , although both maternal and paternal orphans are still less likely than non-orphans to live with biological parents. Not surprisingly, non-parental living arrangements are more important for double orphans compared to other children. Double orphans are, therefore, the most likely to live with grand parents, siblings, non-relatives, or other relatives. Disparities in the prevalence of schooling disruption by sex, orphan status, and family structure are presented in Table 2. Several striking patterns are illustrated in these results. First, despite extensive research confirming the buffering effects of two-parent families on children's schooling, Table 2 suggests that these impacts are not consistent across orphan status. Specifically, the prevalence of schooling disruption among paternal orphans and female double orphans is significantly higher in two-parent families than in single-parent families. For paternal and double orphans, therefore, more positive schooling outcomes are associated with living in single-parent families and not in two-parent families. These associations are antithetical to those observed among non-orphans. As Table 2 demonstrates, nonorphans generally have outcomes consistent with the expectation that children will have more favorable schooling outcomes in two-parent than in single-parent families. In general, the lower prevalence of schooling disruption among orphans in single-parent families seems counter-intuitive. Yet there is a differential relationship between schooling and family structure, that appears to reflect the fact that orphans in two-parent families are more likely to be involved in non-biological relationships than children in single-parent families. In fact, further analysis of the data revealed that while only 25% of all orphans in twoparent households were the biological children of their household heads, this percentage was much higher, i.e. 69%, among all orphans living in single-parent households. 4 These differences are important because, as mentioned earlier, nonbiological relationships are known to have a negative association with schooling outcomes especially among orphans. With regard to schooling disparities among orphans, Table 2 shows that paternal orphans in all households have a lower prevalence of non-enrollment compared to other orphans. This apparent paternal-orphan advantage is driven by the outcomes of paternal orphans in singleparent families. Given the fact that Table 1 suggests that household socioeconomic disadvantages are higher among paternal orphans, their lower comparative levels of non-enrollment are somewhat surprising. At the same time, maternal orphans have the highest relative levels of schooling disruption among children in single-parent households but not among children in two-parent households. In two-parent households, schooling disparities among paternal, maternal, and double orphans in Table 2 are less distinct. --- Multiple Regression Results A more comprehensive analysis of how family contexts mediate orphan status disparities in schooling disruptions is presented in Table 3. According to these results, child-level demographic factors, i.e. age and sex, do not explain the non- enrollment disadvantage of orphans compared to non-orphans . Similarly, accounting for child-level demographic controls does not eliminate the higher relative disadvantage of maternal and double orphans previously observed in Table 2. Chi-square tests , however, reveal that there is no significant difference between the estimated likelihood of schooling disruption of maternal and double orphans. Model 1, therefore, confirms that it is among young Rwandan children without living mothers and those with two dead parents that the highest risks of schooling disruption are found. Although Rwandan orphans live in households with greater socioeconomic disadvantages relative to non-orphans, socioeconomic disadvantage does not account for their higher likelihood of schooling disruption. Thus, when household-level covariates, such as household size and wealth are controlled , the relative disadvantage of maternal orphans is only slightly reduced. Nevertheless, the relative risk of schooling disruption among both maternal and double orphans remains. Significantly, paternal orphans continue to have more favorable outcomes relative to other orphans, even after other factors are controlled . This finding is generally consistent with the notion that mothers take on an expanded role in their children's schooling after the death of their husbands. Furthermore, given the relatively higher socioeconomic disadvantage of paternal orphans this finding also reflects the fact that among orphans having a living mother is likely to offset the negative association between schooling and household poverty status. Model 3 examines whether household wealth has a differential association with schooling that is conditional on orphan status. It therefore includes three interaction terms that capture the relationship between household wealth and likelihood of schooling disruption for each orphan-type, relative to non-orphans. Unlike maternal orphans, paternal and double orphans experience significantly higher risks of schooling discontinuity relative to non-orphans as household wealth increases. Orphans with a living father, i.e. maternal orphans, do not experience significantly different risks of schooling disruption than non-orphans as household wealth increases. The significant disparities between non-orphans and other orphans in wealthier households are consistent with the expectation that wealthier parents invest more resources on the schooling outcomes of biological children relative to those of non-biological children. Notably, as Model 3 suggests, significant differentials in schooling with increasing household wealth are less likely to occur if orphans have a living father. --- Family Structure and Schooling Disparities By stratifying the sample based on whether or not children lived in single or twoparent families, Table 3 also investigates how contextual influences associated with differences in household structure mediate the relationship between orphan status on schooling. A striking feature of these results is that family structure has a stronger association with the schooling of paternal orphans relative to other children. Significantly, family context is a facilitator and impediment to the schooling of paternal orphans in single and two-parent households respectively. In single-parent households , for example, paternal orphans are the least likely to drop out of school. This paternal orphan advantage is not significant after differences in household-level attributes such as the characteristics of household heads, the number of young children living in households, and household wealth are controlled . Nevertheless, the direction of the coefficient estimate for paternal orphans in Model 5 continues to suggest that they have better outcomes than other children in single-parent household contexts. In contrast, maternal orphans in single-parent households are more likely than other children to experience schooling disruption . This maternal-orphan disadvantage persists even after other household-level attributes are controlled . In Model 7, the contextual influences of two-parent households are shown to have a stronger negative association with schooling among paternal orphans than on other children. Chi-square tests, however, indicate that the differences between paternal and double-orphans in two-parent households are not significant. The general disadvantage of paternal orphans in two-parent families is, however, robust and is therefore not explained by differences in other household-level attributes . The results also indicate that part of the paternal orphan disadvantage in two-parent households may be driven by the differential impacts of household wealth on paternal-orphans relative to non-orphans. Accordingly, interaction terms for household wealth and orphan status reveal that it is especially in two-parent households , and not single-parent households , that the disparity between paternal orphans and non-orphans becomes more accentuated as household wealth increases. Significantly, however, the contrasting associations between single-parent and two-parent households and schooling among paternal orphans suggest that the importance of maternal role expansion after the death of a father is generally contextual. In short, while having a living mother is generally beneficial, the presumed significance of mothers for schooling outcomes is mediated by family structure. --- Sex of Household Heads and Schooling Disparities Crucial to the understanding of how orphans fare in single and two-parent households is the question of whether the sex of their household heads affects schooling disparities in both household types. Since single-parent orphans still have one living parent, it is important to examine, for example, whether their outcomes will differ from those of non-orphans in single-parent households if the sex of all single-parents is held constant. Similarly, there is also the question related to how the schooling outcomes of paternal and maternal orphans in two-parent households are affected by whether or not these households are headed by the living parent or a proxy parent of these orphans. Understanding the nature of these differences is important because of the need to clarify whether association conditional on differences in family structure are driven by altruistic relationships. In Table 4, Panel A, estimates are presented for the likelihood of schooling disruption, by orphan status, in single-parent and two-parent households stratified by the sex of household heads. These models only include additional controls for child age and sex. Panel B of Table 4 presents similar estimates of the association between orphan status on schooling. However, its estimated coefficients are derived from models that also control for age, sex, and whether or not orphans and non-orphans are the children of their household heads. Among children in single-parent household headed by females , paternal orphans, i.e. orphans with a living mother, have better outcomes compared to other orphans and non-orphans. This paternal-orphan advantage is driven by the fact that in such households, paternal orphans are more likely to be the children of their household heads. Accordingly, when we control for whether or not orphans and non-orphans are the children of their household heads , the significance of the paternal orphan schooling advantage is eliminated. For both maternal and paternal orphans, however, schooling outcomes are comparatively worse when the sex of their household head is the same as the sex of their dead parent. Thus, among paternal orphans, living in male-headed two-parent households, for example, is associated with comparatively larger schooling disadvantages , but not in those headed by females . Accounting for whether or not children have a biological relationship with their household heads completely eliminates the paternal orphan disadvantage in male-headed two-parent households. In short, two-parent households are likely to be worse for paternal orphans relative to other children because the former are more likely to lack biological relationships with the male head of such households. While the correspondence between the sex of a single-parent household head and that of a living parent is associated with greater schooling advantages for paternal orphans relative to all children this relationship does not appear to hold for maternal orphans. In other words, living in male-headed single-parent households does give maternal orphans a schooling advantage over other children . At the same time, maternal orphans experience a higher likelihood of schooling disruption in single-parent households headed by females . With regard to double-orphans, the coefficients in panel A suggest their schooling outcomes are less favorable than those of non-orphans across family contexts, regardless of the sex of their household heads. Unsurprisingly, however, this disadvantage is generally explained by the lack of biological relationships between double-orphans and their household heads . --- Other Determinants of Schooling among Orphans Restricting the analysis only to the sample of orphans, Table 5 further examines the extent to which factors such as sibship size, specific types of relationships between orphans and their household heads, and other factors affect disparities in schooling disruption. Furthermore, Table 5 also investigates whether the differential relationships between the sex of household heads on orphans schooling observed in the previous section are robust to the possible influence of household socioeconomic indicators. Orphans are generally more likely to drop out of school as the number of children below age 7 living in their household increases. This finding has two important implications. The first is that because children below age 7 are below the official school entry age in Rwanda , it provides confirmation for the presumed schooling impediment associated with family childcare responsibilities . Wealth dilution explanations are unlikely to account for this relationship since household wealth is concurrently controlled. Secondly, within households, children below age seven are likely to include new children born to the parents of maternal or paternal orphans who remarried after the genocide.5 Thus, the negative association between schooling and living with such children may also reflect the negative impacts of competition for other parental resources among orphans and these new younger children. Model 1, however, clarifies that the presumed negative relationship between children below age seven and orphans' schooling is mediated by orphan-type. Accordingly, interaction terms examining the significance of larger young-sibship sizes among orphans indicate that maternal, and to some extent paternal orphans, are less likely than double orphans to be negatively affected by increases in the number of younger children within households. Although the patterns of association are similar for paternal and maternal orphans in single-parent households , the association is only significant among the former. There are no significantly different results on the relationship between schooling and sibship size, by orphan type, in two-parent households. Interaction terms are also included in Model 2 to test the extent to which the suggested impacts of having female household heads among paternal and maternal orphans are robust to the influences of household size, wealth, and the number of children below age 7 living within households. In general, the robustness of the differential patterns of association among female household heads is confirmed. Furthermore, Models 5 and 7 confirm that the positive association between schooling and having a female household head among paternal orphans is consistent across family structure. However, they also show that the respective relationship between schooling and having female household heads among maternal orphans is not significant after other household-level factors are controlled. Significantly, Table 5 also shows a strong association between degrees of relatedness between orphans and their household heads and schooling disruption. Also underscored in these findings is the important role of the extended family in caring for the needs of orphans in Africa that has been reported in other studies . Accordingly, Table 5 shows that orphans do have better schooling outcomes when they have an extended familial relationship their household head than when they do not . Additionally, the disadvantage of non-related orphans is also mediated by family structure since they have even less favorable outcomes compared to other orphans in two-parent than in single-parent contexts. Unlike other previous studies, there is no significant sex difference in the likelihood of schooling disruption among orphans. Unsurprisingly, however, schooling discontinuities decrease as household wealth increases . This wealth effect is more important among orphans in single-parent rather than in two-parent households. --- Discussion and Conclusions A decade and a half after the genocide in Rwanda, lingering questions remain about the effects of this crisis on subsequent population processes. This study, therefore, contributes towards expanding our understanding of post-genocide Rwanda, by locating the schooling outcomes of orphans and non-orphans within the context of their association with familial contexts. Given the exceptionally high rates of orphanhood in Rwanda and several other African countries, the limited focus on whether the relationship between schooling and family structure mediates orphan schooling disparities is somewhat surprising. A broader conceptual contribution of this study is that it situates the possible influences of single and two-parent families within the larger context of degrees of relatedness between children and their household heads. Despite orphans' higher overall risk of dropping out of school, the study finds that the disadvantage of orphans associated with differences in family structure, and for single parent-orphans, the sex of their surviving parent. Family structure is associated with schooling discontinuities among orphans in ways that deviate from its known relationship with children's schooling outcomes. Accordingly, while the percentage of non-orphans experiencing schooling discontinuities is lower in twoparent families but higher in single-parent families, the reverse is true among orphans in such families. On the basis of these differentials, the intuitive expectation that disparities in altruistic investments are smaller in single-parent families represents a plausible explanation of these findings. Although two-parent family contexts have a stronger negative association with schooling disruption among all orphans, paternal orphans are less likely to experience such disruptions than both their non-orphan and other orphan counterparts in single-parent families. While the expanded maternal role that follows the death of a father likely explains the advantage of paternal orphans over other orphans in single-parent households, the reasons for the schooling advantage of paternal orphans relative to non-orphans in such households are not quite clear . One plausible hypothesis is that the higher risk of conflict between the residential parent of children in single-parent families and their non-residential parent , negatively impacts stress-related factors which can, in turn, affect non-orphan's schooling outcomes. In spite of the paternal-orphan advantage in single-parent households, all other orphans still continue to be more likely to experience schooling disruptions than non-orphans in single-parent families. More importantly, because paternal orphans live in the least favorable socioeconomic conditions but are still the least likely to drop out of school in singleparent families this study argues that among orphans, having a living mother possibly offsets the negative impacts of household poverty on school enrollment outcomes. The disparate outcomes of children in both family types are consistent with previous research on the schooling outcomes of children living with biological mothers, but not with biological fathers, in single and two-parent family contexts. Nord and West , for example, argue that while living with a biological mother in a single-parent family has a positive impact on schooling, having a biological mother and a non-biological father in two-parent families negatively affects children's schooling. Reinforcing this argument is the study's finding that paternal orphans in Rwanda have better schooling outcomes in single-parent families because their household heads in such families are likely to be their own biological mothers. In two-parent families, the findings also suggest that their disadvantage is associated with the fact that although they have lost their biological fathers, their household heads in two-parent households are more likely to be males. Part of the paternal-orphan disadvantage in two-parent families may therefore be explained by the fact that non-biological fathers are less likely to provide for children's schooling expenditures and help with their homework, which can lead to negative schooling outcomes . Similarly, since biological mothers in two-parent families are less involved in the education of their biological children from previous relationships who also live with her , the paternal-orphans disadvantage in such contexts may be exacerbated by maternal role reductions among the mothers in blended two-parent families. Notwithstanding the heuristic significance of these findings the extent to which they are replicable in other contexts will require additional scholarly attention. In particular, unprecedented mortality during the Rwandan genocide generated rates of orphanhood that are currently unmatched even by countries with a high prevalence of HIV/AIDS. Consequently, the question of whether the scale of orphanhood affects familial responses to orphans should be further examined. Furthermore, familial responses to orphans may also differ in Rwanda than in high HIV/AIDS mortality countries if orphans in these countries are viewed differently than their counterparts in Rwanda. Significantly, however, many conceptual relationships found in the analysis are consistent with previous studies on issues such as the significance of altruism and the role of blended families on the welfare of children . In this regard, these results provide a useful illustration of how these factors operate in a uniquely Rwandan context to affect the welfare of orphans. At the same time, these findings also have other implications for policy and future research. Although two-parent families have traditionally been known to provide positive contextual influences on the outcomes of children, better strategies need to be developed to ensure that orphans experience comparable schooling outcomes as their non-orphan counterparts in such families. Effective support systems are also needed to build upon the slight advantage of orphans in singleparent families, since despite their better disruption outcomes in single-parent families, the schooling outcomes of maternal and double-orphans are still worse than those of non-orphans in single-parent families . Finally, future research needs to further examine whether differences in family contexts affect other welfare indicators among orphans. Comparisons of orphans in pre and postgenocide societies will also illuminate our understanding of how these contexts operate before and after civil conflicts. As these issues are illuminated in future research, the impacts of family processes on the outcomes of orphans will be better understood and enhance our comprehension of the role of family contexts on the welfare of orphans. --- Appendix 1 See Table 6.

This study examines the relationship between orphan status and schooling disruption in post-genocide Rwanda. The results indicate that while nonorphans have more favorable schooling outcomes in two-parent than in singleparent families, the reverse is true among Rwandan orphans. In single-mother households, paternal orphans, i.e. orphans with only a living mother, have better outcomes than their orphan and non-orphan counterparts. In contrast, paternal orphans have worse outcomes than other children in two-parent households, especially in households headed by males. Maternal orphans are more likely to experience schooling disruptions than non-orphans regardless of family structure. The maternal-orphan disadvantage is nevertheless greater in female-headed than in male-headed households. As expected, non-related orphans are more disadvantaged than orphans related to their household heads. However, non-related orphans have a greater disadvantage in two-parent than in single-parent households. The results also suggest that within households, the provision of childcare to children below schooling age is an impediment to orphan's schooling. These impediments are, however, greater for double-orphans than paternal or maternal orphans.

Introduction Drinking alcohol while pregnant is dangerous for health. Fetal alcohol syndrome is the most severe form of alcohol use-related damage, and is responsible for physical, intellectual, cognitive and behavioral impairments in the child . Europe consistently registers the highest levels of alcohol use during pregnancy in the world, estimated at 25.2%, and the highest prevalence of FAS, at 37.4 cases per 10,000 people . In France, where this research takes place, alcohol use during pregnancy is a major public health issue, as an estimated 27% of mothers drink alcohol and the prevalence of FAS is 41.4 per 10,000 people . To tackle this issue, the French authorities have introduced three main measures: an annual mass media campaign under the banner "no alcohol while pregnant, " consumption screening and prevention tools for health professionals, and pregnancy warning labels on alcoholic beverages. This paper deals with the on-container warnings measure, which the World Health Organization advocates as inexpensive for governments to deploy and effective at informing the public . Warnings targeted at women have been compulsory since 2007 in France , where producers are to label all alcohol containers with a warning text or a prohibition sign-like pictogram representing a pregnant woman drinking alcohol crossed out with a red line . In practice, most beverages carry the pictogram that is barely visible at an average size of 0.4 cm , often placed at the back, embedded in other on-beverage marketing . Equivalent warnings to inform on health have been adopted in only very few countries, such as Australia, New Zealand, Turkey, and the Republic of Moldova . One reason of this weak label implementation in the world may be strong lobbying by the alcohol industry against this measure . In Australia for instance, the AI has managed to delay the introduction of a new larger warnings that target pregnant women by 3 years . In France, two research studies stressed that the pictogram currently displayed on containers is poorly noticed and thus ineffective in grabbing attention and informing people . As a result, in 2018 an evolution was proposed as part of the 2018-2022 "National Mobilization against Addictions" plan. This evolution in format was designed to improve the visibility of the pictogram by increasing its size and imposing a color or contrast . This change met with strong opposition from the AI and has so far not been adopted. Research is needed to better understand the lobbying activities of the AI first against adoption of the French health warning label and during the recent warning expansion project . There is scant research into lobbying on alcohol warnings in general and only four studies on strategies against warnings that target pregnant women . Analysis of these strategies found that the AI claimed such health warnings are ineffective, because there is no evidence that pregnant women stop drinking; should be replaced by alternative measures ; are unneeded, because women already know the risks of consuming alcohol; will have negative unintended consequences ; restrict international trade and create added costs for producers; are illegal because they curtail the AI's freedom of speech and there is no legislative authority to compel warnings on alcoholic beverages. The aim of this research is to bring insight into the AI's lobbying against warnings aimed at pregnant women by analyzing the arguments advanced by the AI in the French press from 2000 to 2020 in an effort to block original implementation of the warnings and the later warning label expansion project . Our research makes several contributions. It brings fundamental insight to the limited literature on lobbying against alcohol warnings in general and those aimed at pregnant women in particular. The AI lobbies against the development of effective alcohol control policies around the world , so it is important to analyze the strategies it employs in order to better understand them. The analysis proposed here spans a period of 20 years in France, which covers a period that includes the original adoption of alcohol warnings and the later warning label expansion project that was not adopted. Analyzing differences and shifts in the AI's arguments may explain why this health measure got implemented in 2007 and why its subsequent expansion got blocked. Another theoretical contribution is to enrich the sets of arguments identified in the extant literature on Anglosphere countries and to propose an analytical framework that may be better adapted to countries like France that have long and deeply embedded ties with alcohol in general and wine in particular. France offers a specific case-setting regarding alcohol and drinking culture, for five reasons: alcohol is widely consumed in France compared to other countries : on average, in 2017, the French consume 11.7 L of pure alcohol per year per capita ; cultural acceptance of wine is very high in French society ; the "French paradox", a myth-strongly criticized by the scientific community -asserting that moderate consumption of red wine is good for health remains deeply embedded ; the economic dimension is important: in 2020, France was the second-largest producer and third-largest exporter of wine worldwide ; the AI has strong connections with President Emmanuel Macron , who was elected "Personality of the Year 2022" by the Revue du Vin de France for "his constant commitment to wine and its culture, " an award that he came to collect in person . Beyond these theoretical contributions, our research is also useful for NGOs and health advocates campaigning to raise awareness of the AI's lobbying against health measures and for countries that are planning to introduce their own alcohol warnings or want to maintain or increase the effectiveness of alcohol warnings in place. The discussion section provides recommendations, based on our findings, on how to deal with and counter lobbying by the AI. --- Materials and methods A qualitative analysis of media documents was conducted to identify the arguments advanced by the AI on French alcohol warnings targeted at pregnant women. The AI is understood to mean "a multi-national business complex that includes not only producers of beer, wine, and distilled spirits but also a large network of distributors, wholesalers, and related industries" . Trade associations that promote alcohol producers' interests and "social aspects and public relations organizations" are thus considered part of the AI , along with elected representatives of wine-producing regions or parliamentary groups identified in previous research that publicly defend winegrowing interests within the French government . --- FIGURE --- Flow chart of the inclusion process for press articles identified by searching the Europresse database. Frontiers in Public Health frontiersin.org . /fpubh. . Press articles were analyzed from the 2000-2020 period in order to identify the full spectrum of discourses during this relatively long period and arguments used prior to the warnings adopted in 2007 until 2 years after 2018, the year that the pictogram expansion project that never got implemented . --- Data collection The mainstream press is a good tool for analyzing industry lobbying and arguments , as media coverage can influence and frame public debate, public opinion and policy outcomes, especially when governments attempt to implement restrictive measures to improve public health . Here we used the Europresse documentary database, which provides access to a large number of leading print-press sources ranging from newspapers, magazines, journals and online news content to press releases and newswires . The search terms queries in Europresse were " AND AND " ET ET "). Polling the database returned 559 articles: 75 were excluded after deleting duplicates and articles from non-analyzed media ; 200 were excluded based on title and content relevance ; 199 were rejected because they did not express the point of view of the AI . A total of 85 press articles met the inclusion criteria and were included in the final analysis . --- Data analysis A quantitative analysis was first carried out in order to identify the number and trend-curve of articles covering the research theme over the 2000-2020 period. The profiles of the AI actors who spoke on the issue were identified. In each press article, the names of the interviewees expressing their views, their company and their alcohol sector were coded. They were then classified into the following categories: wine, spirit, beer, SAPROs, alcohol sector and other . An inductive method was used to analyze the data without "trying to fit the data to pre-existing concepts or ideas from theory" . Because it was the first French analysis of the lobbying of the AI against warnings, and because the French context toward alcohol is very specific , an inductive approach was preferred compared to the use of Anglo-Saxon frameworks suggested previously to analyse the lobbying of the AI against warnings . This approach provides detachment from the existing literature and contexts that are different from those of French culture. A thematic content analysis was then conducted to map content and topics across the data by identifying key themes . This analysis consists of an initial reading of all press articles. For each article, the argument used by the AI were identified and then grouped into categories of arguments . Once the analytical framework was finalized to highlight the content of the AI's arguments, the occurrence of the same category of arguments was counted. To conduct the analysis, a researcher independently carried out a manual coding of all press articles. A second researcher coded also independently this material using NVivo12 qualitative research data analysis software. The two analyses were compared, and if any divergences appeared, a team meeting with a third researcher was held to reach a consensus. --- Results --- Pattern of change in the press articles published over the years studied A majority of the 85 press articles included in the final analysis came from the written press . The number of articles published differed by year, with less articles published around the original implementation in 2007 than around the time of the pictogram expansion project in 2018 . Both periods coincided with peaks in the number of publications, which is evidence that the AI response to a public health proposal is to react via the mainstream press. The rise in the number of articles between 2007 and 2018 may be due to the development of the online press in 2018 compared to 2007 and/or to a professionalization of the lobbying of the AI between the two periods . It is also interesting to highlight a drop of articles after 2018 . It may be explained by a favorable political context at that time for the AI, especially for the wine industry with which the French President Emmanuel Macron has very close ties . --- Identification and evolution of who spoke via press articles A large majority of the AI actors who spoke in the press came from the vine and wine sector followed by the broader "alcohol" sector and SAPROs . Note, however, that the spirits sector , the beer sector , and another agent also voiced opinions via the press . Between period 1 and period 2, the number of winesector actors using press remained stable, at 48 vs. 45 occurrences , whereas other voices had a weaker and more isolated presence. This could be explained by the fact that the wine sector is well-perceived in France, and is increasingly becoming the front group for all the alcohol actors in the media . An identical phenomenon was observed during the lobbying against the Evin Law of marketing regulation in France: the winegrowers were a very visible front group to fight against the law compared to other alcohol actors . --- Arguments advanced by the alcohol industry in French media The majority of the arguments used by the AI from 2000 to 2020 were raised against the original introduction and subsequent evolution of the warning . There was nevertheless a small minority of AI arguments in favor of the measure . Different subcategories of arguments emerged from the analysis and are described below. --- FIGURE Evolution of the actors who spoke in the media. --- Arguments against the French pregnancy warning label Our analysis suggested there were three categories of arguments against the measure: pregnancy warning labels are a questionable measure, pregnancy warning labels would have counterproductive effects, and better alternatives exist. These categories are outlined below, with examples provided. --- Pregnancy warning labels are a questionable measure Exaggeration and overzealousness of actors in health The pictogram clearly links alcohol to mortality was one of the main arguments advanced by the AI , essentially in period 2 : "The sector is opposed to what it publicly describes as a 'deadly pictogram'" The AI claims that this measure is driven by hygienefirst logic, i.e., dictated by the medical perspective . This was the argument most used during period 2 : "the winegrowers fear that this enlarged pictogram will lead to a 'hygiene-first' logic [. . . ] where the ultimate form would be a bottle similar to the plain tobacco packaging" It also claims that this measure only serves to reassure the health authorities : "Does the reminder on wine bottles serve any other purpose than to reassure and hypocritically relieve the health authorities of all liability?" To prove this point, the AI signals that France is one of the only countries to implement the pictogram, which they see as further evidence that the measure is exaggerated : "This constraint is a feature specific to France and is not used in other European countries" --- FIGURE --- Taxonomy of alcohol-industry arguments. Finally, it was also mentioned that the measure was disproportionate and would open the floodgates to more virulent messages , such as those for tobacco , arguing that alcohol should not be treated in the same way as tobacco : "And the media coverage of it is odious. Soon, we will see 'alcohol kills' on labels of good Burgundy wine. Just like on cigarette packets" "It is even a 'total idiocy' that will 'further fuel the idea that wine is a dangerous product like tobacco'" --- Ine ectiveness of the measure The AI also put forward the ineffectiveness of the warning, arguing that the measure is ineffective in changing behaviors among pregnant women : "Women who are addicted to alcohol will continue to drink, just as smokers continue to smoke despite the warnings displayed on tobacco packages. The rest [of the women] already know not to drink during pregnancy." "To date, no comprehensive study has been produced to demonstrate the effectiveness of this measure." This ineffectiveness is explained by the poor design and content of label. The AI claims that the warning is not precise enough and could thus create confusion , and is poorly crafted : "When our Chinese customers saw it, they thought our wine was a contraceptive. . . " "Yves d'Amécourt [an elected representative] is outraged by the 'starkness' of the pictogram, as according to him, 'what could be more beautiful than a pregnant woman?'" The AI also adds that it is unreadable because container labels are already overloaded with information : "For the Vignerons Indépendants [professional association of independent winegrowers], it is yet another feature to fit on already overcrowded label" Pregnancy warning labels are touted as ineffective because they are perceived as a "cosmetic" measure : according to the AI, pregnancy warning labels are considered as superficial and therefore useless. AI actors also consider this measure as inappropriate to fight against alcoholism among pregnant women: "winegrowers deplore a 'primarily cosmetic measure'" "The measure is an 'inadequate response to a real public health issue'" --- Minimization of the severity of the alcohol problem The AI minimizes the severity of the issues tied to alcohol use by arguing that women are already informed and responsible : "Question: Do pregnant women know that alcohol is dangerous for their unborn child? Answer: To not know, either they'd have to ignore it on purpose or spend their pregnancy lost in a cave in the woods." The AI also claims that wine is not alcohol : "Is wine an alcohol like any other? 'Wine, consumed in moderation, is part of the traditional French foodways. It has to be kept apart from other alcoholic drinks'" They also added that alcoholics do not tend use wine and that FAS remains rare : "It is even more ridiculous for wine: I have never considered myself as trading in alcoholism, and in any case, alcoholics are not loyal customers of the winegrowers!" "Fetal alcohol syndrome is exceptionally rare " --- Skirting the alcohol problem The AI claims there are other more important problems than alcoholism during pregnancy and therefore regrets that the pictogram draws all attention onto this one specific issue. It thus proposes setting up pictograms for other problems : "If I have to add this pictogram, I think I will also add 'forbidden for diabetics', 'forbidden for people under 16' and 'forbidden for idiots'." The pictogram is also argued as questionable on the grounds that all human activities are dangerous and so other more serious issues should be addressed first ,: "Just as I'd let you have fun coming up with all the pictograms that could be put everywhere in our environment to remind us that the most banal human activities-breathing, eating, driving, sports-all carry risks and that, ultimately, living is bad for your health." "Wine kills fewer people than pharmaceutical drugs, but it is not politically correct to say so" --- Other arguments Two other more marginally-used arguments claimed that the measure was ridiculous and unpopular : "It's a bit ridiculous and very Franco-French" "The issue is unpopular" . /fpubh. . --- The warning would have counterproductive e ects On the economy The AI argues that this measure attacks the wine sector and weakens producers : "these 64 wine-growing estates denounce 'the transformation of a product that vectors excellence and is sold across the globe into some kind of contraband [. . . ]'" "The labeling of bottles [with the pictogram] also appears to be the last straw for a wine industry already in crisis." To a lesser extent, the AI raises the point that the producers need time to implement a warning label , that the cost will be high and that containers carrying the pictogram will be harder to export : "For the president of Brasseurs de France, 'implementing the measure will necessarily take some time, given the time needed to print new labels for our 400 different product references'." ("Femmes enceintes: les fabricants d'alcool résignés à apposer un pictogramme, " AFP Infos Françaises, --- 2006) "Adding labels or creating back-labels increases our costs" "How can we grow exports if wine is considered a dangerous product in France?" The measure is seen as binding and unfair for producers. They fear that the measure will cause a drop in sales due to lower consumption : "new constraints in terms of labelling" "If we have to put it somewhere on the bottle, then all European countries should do it too." --- On women The AI claims that displaying a pictogram that targets pregnant women stigmatizes women and causes guilt and anxiety : "But what about the risk of guilt-tripping women by labeling the risk?" "It is true that the alcohol lobby considers that informing women would be 'anxiety-provoking'" --- Better alternatives exist Promoting responsible consumption and its benefits for health The AI claims that responsible and moderate alcohol consumption is not dangerous, even for pregnant women , and that people should drink responsibly : "Alcohol misuse is not recommended for pregnant women, but an occasional drink is not forbidden" ("Une table ronde pour goûter si le vin est bon, " Le Progrès -Lyon, --- 2005) "We should each make sure we are careful, rather than continue calling for new regulations" They highlight the "benefits" of alcohol consumption that get forgotten with this prohibitive warning: "André Dubosc [in charge of development at Producteurs Plaimont, a grouping of several wine cooperative structures] also deplores the fact that only the harmful effects of alcohol are taken into account.'I cannot accept that when studies show positive effects, they are not listened to'." --- Promoting preventive programs The AI proposes launching wide-reaching education and prevention programs aimed at pregnant women, rather than just a simple warning : "We [the major alcoholic beverage companies in the SAPRO 'Entreprise et Prévention'] advocate targeted prevention, which is always more complex and often more expensive but ultimately more effective than simple regulatory measures" Promoting national-scale education on how to responsibly drink wine is another argument put forward : "Instead of educating people, showing them how to taste [the wine], to enjoy it in moderation, we make them feel guilty. It's shameful." --- Promoting the role of health professionals The AI emphasizes that prevention of alcohol exposure in pregnancy is the role of health professionals : "Warning pregnant women is not the legislator's role but the role of doctors" ("Femmes enceintes et alcool professionnels du vin furieux, les autres partagés, " AFP, --- 2004) The AI also argues that the warning label is not the right way to display health information as it must not be construed as a medical prescription : "Are we going to turn our bottles of wine into medical prescriptions?" --- Promoting alcohol-industry collaboration on prevention The AI would like to be part of the policy-making process and criticizes the fact that decisions on the pregnancy warning label were taken without consulting with business : "The councilors denounced what they saw as 'rushedthrough implementation' of a 'unilaterally imposed' modification brought in 'without consulting with the wine industry, whereas there were 500,000 vine and wine jobs set to be directly affected by these brutal changes'" The AI has moved to prove its concern for public health issues. It voiced readiness to engage in prevention initiatives and proposed a "Moderation Council" in period 1 as a public health-private business partnership : "We [members of the interprofessional bureau of Burgundy wines] are absolutely ready to engage in consumer education and awareness action on the dangers of alcohol abuse." ". . . the creation of a 'Moderation Council' that deliver targeted communication to pregnant women" Beyond these arguments against the warning-label measure, there was a fairly marginal minority of AI agents in favor of the measure. Their arguments are described below. --- Arguments in favor of the pregnancy warning label Two categories of pro-pictogram arguments emerged: positions in favor of implementation , positions in favor of an expanded pictogram . These categories are outlined and materialized via the examples given below. --- In favor of implementation Some of the actors were unconditionally in favor of implementing the pregnancy warning pictogram and sometimes even promised to spread the "abstinence during pregnancy" message even more widely. This pro message was mainly promoted by individual winegrowers. They were ready to implement the pictogram after seeing shocking media coverage of the "Lille affair" in 2004, a lawsuit brought by mothers whose children were victims of FAS. The mothers went to court because of the lack of information about the dangers of alcohol consumption during pregnancy at the time of the trial : "Woken up by the Lille affair [during which an association filed a complaint for failure to inform pregnant women about the dangers of alcohol use], he did not want to wait for the new legislation. His 600,000 bottles of Château Puech-Haut will now carry a statement that reads 'not advised for pregnant women' in four languages." One other actor, a SAPRO that represents the AI , was also in favor of warnings on condition that the effort was a shared one: "We are not against labelling bottles with health messages", says Arnaud Lassince, general manager of Entreprises et Prévention that federates 18 of the largest French alcohol industries and represents 150 brands of alcoholic beverages. "We are not lobbying against it we are well aware of the issues. But we don't want to be the only ones who have to get the message through." --- In favor of expansion During period 2, some actors were in favor of doubling the size of the pictogram and doubling the prevention effort if a compromise was proposed , if the size did not exceed a certain threshold , or if there was grace period before implementation : "The profession proposes that the sticker should double in size, to 0.8 mm instead of 0.3 or 0.4 mm, and be printed with greater contrast. The profession also commits to widely disseminate the 'no alcohol while pregnant' message" "If we're proposed a two-centimeter pictogram, that might be a problem, because our labels are small. But we should be able to find a compromise by playing on the colors to make it more visible; the issue is not opposition to the measure" Some actors were sometimes supportive without specifying conditions : "Deemed too discreet by the health authorities, the expansion of the pictogram prohibiting alcohol for pregnant women is one of the measures proposed [proposed by the three alcohol sectors in their contribution to the national public health plan]" --- Discussion A growing public health literature on the influence of companies selling unhealthy products like alcohol that are bad for population health has converged around the emerging concept of commercial determinants of health . Lobbying is one of the channels through which companies influence public policy . The AI mobilizes lobbying strategies in an effort to counter effective alcohol prevention measures , including labeling . It is well-known that in the context of tobacco, warnings improve consumer knowledge and influence smoking behaviors . However, in the context of alcohol, warnings are an under-developed health measure: few countries have adopted alcohol warnings, and the warnings that have been adopted tend to be poorly designed . One reason for this situation is AI lobbying against this measure . Lobbying strategies and arguments against alcohol warnings have rarely been analyzed in the literature. Our research fills this gap by analyzing the arguments used by AI lobbyists in the French mainstream press in their defense against alcohol warnings targeted pregnant women. --- Main results of the research This research reveals that the AI responds to new public health proposals via reactions in the mainstream press. Indeed, peaks in number of publications were found to coincide with the warning-label implementation and with the pictogram expansion-project schedule . This shows that the AI uses the mainstream press as an indirect tool for lobbying, as previously identified . As the number of articles through the press is higher in 2018 compared to 2007, it may be due to professionalization of strategies of lobbying of the AI that could consist of a wider use of the press to spread arguments and thus indirectly influence opinion leaders and decision makers. This research also found that the AI tends to develop transnational arguments to counter warnings in general and those that target pregnant women in particular in different countries. Some arguments identified in this French research have also been identified in the extant research on lobbying against alcohol warnings in Anglosphere countries , i.e., that the warnings are ineffective in changing behaviors, that warnings would have counterproductive effects on women and the economy, and that there are better alternatives, such as public information campaigns which are actually less effective than population-based interventions. Interestingly, we also surfaced arguments that emerge specifically in France, showing that lobbying seems to be adapted to nation-specific features . For instance, some arguments emphasize how wine and winemaking hold a special place in the French culture, with the idea that "wine is not an alcohol like others" , and a special position in the French economy . As wine is deeply embedded in French culture and given the strong proximity between the French President and this wine sector , the AI may have adapted its strategies and uses the voice of winemakers in the press as a kind of credible and appreciated front group. Our research has also emerged new arguments not previously captured in research on lobbying against warnings. We identified arguments in favor of the measure , some of which were voiced by independent winemakers in period 1. They were ready to implement the pictogram after seeing shocking media coverage of a lawsuit brought by three mothers whose children were victims of FAS on the grounds that they had been under informed on the dangers of alcohol consumption during pregnancy . During period 2, the AI also agreed to display a pictogram with a minimum size of 8 mm and enhanced by color contrast . Contrary to period 1, this discourse in favor of the warnings mainly came from organizations like "Vin et Société, " funded by the AI and identified as a strong front-group lobbyist in France . Those pro-measure arguments may be part of a process of professionalized lobbying by the AI that includes the recent rise of corporate social responsibility strategies . CSR strategies include launching prevention campaigns , a willingness to collaborate with governments to secure solutions , engagement with safer drinking by offering low and no-alcohol products , and proposed self-regulated marketing in order to protect young and vulnerable people . Considering this CSR trend that is an indirect lobbying tactic used to improve image of companies , the arguments that emerged in France in favor of pregnancy warning labels may stem from this trend developed by the AI worldwide, in a similar way to those employed by the tobacco industry . Our research highlights that whereas arguments used during the two periods were fairly similar in terms of content and number, the AI's lobbying failed in period 1 and won in period 2 . It is difficult to interpret this finding, but part of the explanation may be different political and social contexts . Period 1 was marked by high media coverage of a lawsuit brought by three mothers whose children were victims of FAS . This media coverage may have put added pressure on government to implement the alcohol warnings. Period 2 was characterized by an extremely favorable political context, especially for the wine industry. It is well-known that the French President Emmanuel Macron has very close ties with the AI and regularly expresses himself publicly as a defender of wine, which he qualified as "inseparable from our art of living, this art of being French" and he claimed to journalists that he drinks wine "daily, lunch and dinner" . For his engagement in favor of wine, he was elected "Personality of the Year 2022" by the Revue du Vin de France , he accepted this prize and received it in person . From 2017 to 2019, he also hired Audrey Bourolleau, the former manager of the SAPRO "Vin et Société, " that joined the French government as an agricultural adviser . Regarding alcohol prevention, Emmanuel Macron publicly disavowed the former Health Minister Agnès Buzyn who wanted to increase the size of the pregnant warning in 2018 . These different examples may explain the success of the AI in the period 2 in blocking the enlargement of the pictogram. --- Limitations of the research This research has some limitations. First, it has taken into account the position of the AI in order to identify the arguments it uses. It would be instructive for future research to analyze the arguments relayed by actors in health, both to determine their weight in the mainstream press and to analyze the pro-health arguments used . Second, this analysis was based on the mainstream press. Other potentially-relevant media should be analyzed, such as the trade press, parliamentary documents, or social media in order to gain external validity and identify arguments from other sources. Finally, the study may be exposed to the biases inherent to the use of a qualitative methodology, such as interpretation bias. We tried to limit this bias by working among the team to find a consensus when doubts emerged on some press articles. --- Theoretical contributions This research makes two main contributions to the scientific literature. First, it provides a larger framework for analyzing the arguments of the AI lobbying against alcohol warnings in general and those aimed at pregnant women in particular. Second, the analysis led here was conducted over a long period of time on a health measure proposed by actors in health that has been adopted or not in a non-Anglosphere country where the context is a priori favorable to alcohol . --- Public health contributions and recommendations Based on our findings, we make five public health recommendations to tackle the issue of AI lobbying against health warnings . First, given the strength of lobbying at national level, international treaties are needed to counter national-level influence on individual governments. A European Union Directive on alcohol warnings would be a relevant option, as already done for tobacco in 2014 requiring a combined health warning consisting of a picture, a text warning, and information on stopping smoking, covering 65% of the front and back of cigarette packs . Beyond the European Union level, a "Framework Convention on Alcohol Control" would also be relevant, as already done for tobacco in 2005 wherein article 11 stipulates that each Party signing and ratifying the treaty is to adopt and implement effective labeling measures within a period of three years. For countries similar to France very close to the AI, the article 5.3 of the FCTC that stipulates that each Party has to protect its "policies from commercial and other vested interests of the tobacco industry in accordance with national law" should be replicated to the AI in order to protect alcohol policies. Second, the French example appears to highlight strong links between the AI and government that could explain why certain public health measures do not get adopted. To counter this problem, citizens and health actors need to be better informed on these links via transparency instruments that could limit interference in public decisions. National legislation could be proposed on alcohol, following the example of the French law on transparency of the tobacco industry's influence relations, in particular on expenses related to influencing or representing the interests of tobacco product manufacturers, importers and distributors and their representatives . This could go further by compelling the AI to disclose any and all expenses tied to indirect lobbying channels . Third, more research is needed on the arguments used by the AI against alcohol warnings. To counter the AI's argument challenging the effectiveness of pregnancy warning labels, more studies should be conducted on the content and design of effective warnings. There is limited research on alcohol warnings compared to tobacco warnings at international level and especially in France where only three studies have been published on these issues. Concerning the argument around the economic costs for the AI, more research is needed on the economic burden of alcohol for society. In France, the first research that estimated this social cost dates from 2010 and arrived at a figure of 120 billion euros per year . No research has been carried out since to update this figure. Fourth, it is vital to provide actors in health with more training and skills in order to make them more effective in lobbying tactics and press relations and adopt similar strategies of the AI to better counter them. Finally, given that denouncing the industry's marketing and lobbying tactics seems to be effective, counter-marketing campaigns should be implemented, as was done in tobacco with the "Truth" campaign in the USA . It could be useful to develop a campaign via social media to denounce AI lobbying. The effectiveness of counter-marketing campaigns is explained by inoculation theory, which posits that people can be protected from attempts at commercial manipulation if they are warned against them with counter-arguments . --- Author's note Our research fills a gap in a limited and mainly Anglosphere research on lobbying against alcohol warnings. We conducted an analysis of the arguments advanced by the alcohol industry in the French mainstream press over a 20-year period covering a failure of lobbying and a success of lobbying . Our research found some arguments that were similar to those already identified in the literature, but it also surfaced other arguments that have emerged, probably due to the specificity of France as a pro-wine country and probably also linked to the emergence of CSR strategies. An analytical framework of arguments used by the alcohol industry is suggested, which could be helpful for countries that have a long history of pro-alcohol culture. Various recommendations are suggested in order to counter these alcohol industry's arguments: implementation of international treaties in order to counter national-level influence on governments; implementation of national legislation for more transparency around the alcohol industry's influence; development of research on health measures against alcohol use; educating actors in health on lobbying; implementation of counter-marketing campaigns to expose and delegitimize the alcohol industry's lobbying practices. --- Data availability statement The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. --- Supplementary material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh. 2022.933164/full#supplementary-material

Background: Drinking alcohol while pregnant is dangerous for health. To inform on this issue, various countries have adopted pregnancy warning labels on alcoholic beverages, including France since , where wine holds deep cultural consonance. The aim of this research was to analyze the arguments put forward by the alcohol industry (producers, distributors, wholesalers, allied industries, trade associations, social aspects and public relations organizations, councilors who publicly defend wine-sector interests) via the press in France: ( ) in when pregnancy warnings were first implemented, and ( ) in when larger pregnancy warnings to increase visibility were proposed but not adopted.We used documentary method to analyze the arguments advanced by the alcohol industry in mainstream (national, regional and specialized) press in France from to , using the Europresse documentary database. Quantitative analysis (number and trend curve of articles, mapping alcohol-industry actors who spoke in the press) and inductive thematic content analysis (analytical framework of the arguments identified) using NVivo software were carried out. Results: We found a total of relevant press articles in the database, of which were included in the analysis. Peaks in number of publications were found to coincide with the warning label implementation and with the expansion-project schedule. A large majority of the arguments promoted by the alcohol industry contested the pregnancy warnings measure (very few were in favor). They argued that ( ) pregnancy warnings were a questionable measure (e.g., ine ective, or the pictogram clearly links alcohol to mortality), ( ) pregnancy warnings would have counterproductive e ects (on women and the wider economy), ( ) better alternatives exist (e.g., targeted prevention programs, prevention by health professionals). A large majority of the actors who spoke in the press came from the winegrowing sector. Conclusion: This study fills a gap in the Anglosphere research on lobbying against alcohol warnings by analyzing lobbyists' arguments over a -year period covering both failed and successful industry lobbying. New findings Frontiers in Public Health frontiersin.org Millot et al. . /fpubh. . have emerged that are likely related to the wine-oriented culture of France. In order to counter the alcohol lobbying practices we conclude with a number of public health recommendations.

Introduction Work is an inseparable part of human lives. Through work, human beings fulfill a large part of their needs and inspirations. Therefore, the atmosphere and relationships at work are extremely important. A positive and healthy atmosphere in the workplace is necessary for the well-being of employees while the impact of a workplace disruption on working patterns has a negative influence on all work actors. This antagonistic behavior is not only unacceptable in the psychological domain, but is also forbidden by the law . Undoubtedly, workplace abuse is a growing phenomenon, a prime reason for the importance of scrutinizing it. For example, the International Labour Organization reported that 26% of workers worldwide have experienced workplace violence or harassment in the past year. This statistic is up from 22% in 2010. ILO's study also found that women are more likely to experience workplace violence or harassment than men. Other examples include a study by the Workplace Bullying Institute that found that 60% of American workers have experienced workplace bullying; a study by the European Union Agency for Safety and Health at Work that found that 55% of European workers have experienced workplace harassment; and a study by the Australian Human Rights Commission that found that 40% of Australian workers have experienced workplace bullying. Generally, despite the crystal-clear association between them, these two fields rarely intertwine academically. Reasons stem from different research goals and foci, different research methodologies, funding issues, and more. A call for more interdisciplinary research on workplace abuse was given as early as 2005, and to this date, and to the best of our knowledge, none has picked up the gauntlet of this crucial workplace phenomenon. That is to say, labor law and occupational psychology scholars can learn from each other and may help develop more effective interventions to prevent and address workplace abuse . Specifically, however, by studying the phenomenon of abusive workplaces, researchers may gain better and finer insights into its prevalence, causes, consequences, potential interventions, legal implications, obligations, and protections surrounding workplace abuse, in addition to the psychological processes, individual and organizational factors, and the impact of such behaviors on employees and organizations. It is the point where both occupational psychology and labor law domains collide and intertwine. The importance of such interdisciplinary endeavors cannot be overstated, as there is a plethora of vital implications to exploring this phenomenon further, for example : organizational productivity and performance; employee performance and productivity; legal and ethical considerations; cost implications; social responsibility; employee protection; promoting fair and inclusive work environments; assessing legal remedies; employer liability and risk management; psychological well-being of employees; work engagement and motivation; organizational climate and culture; leadership and management practices; and more . This phenomenon is of paramount importance to a healthy society and economy. Abusive behavior in the workplace is a very dangerous issue both to the employer and the employees and can be traced and found in every organization. The main component of this phenomenon is its identification as abuse since different types of daily behaviors cause stress and mental suffering. Workplace abusive behavior can take different forms., for instance, bullying. Bullying among co-workers includes physical abuse, intentionally spreading false information to hurt colleagues at work; telling inappropriate jokes in the workplace that can offend others; telling false stories about colleagues to humiliate and intimidate them. Bullying against inferior employees under supervision includes physical abuse or threats of such abuse; consistently insulting and criticizing employees; belittling and devaluing the opinions of others in a collaborative or discussion setting; excluding or ignoring an employee in the office or even at social events; acting aggressively either physically or by abusing the intimidator's position; intentionally assigning too challenging or even impossible tasks to complete to justify punishments; spying, or invading the privacy of inferiors by hacking into their emails . Workplace bullying has been considered an important social problem for over four decades . It is costly to organizations and individuals and is considered one of the most difficult areas for human resource professionals to manage . Workplace abuse represents one form of workplace mistreatment behaviors such as violence, aggression, bullying, and incivility . Tepper et al. estimated that approximately 10% of employees experienced supervisory abuse. Bowling et al. reported that 10 to 41% of employees faced workplace aggression acts in the United States, as well as 8 to 26% in Austria,3 to 20% in Belgium,2 to 27% in Denmark,5 to 24% in Finland,8 to 10% in France,23% in Ireland,5 to 9% in Lithuania,20% in South Africa,4% in Sweden,55% in Turkey,and 11% in the United Kingdom. Therefore, workplace aggression is a global issue . Feldblum and Lipnic claimed that as many as 50% of U.S. women experience sexual harassment during their working lives, but only a minority report it. Anyone connected to the organization's employees can be the aggressor, including managers, coworkers, subordinates, and even nonorganizational members, such as customers, clients, patients, and contractors. Consequently, workplace aggression is common . Tepper et al. suggested that supervisors were the leading sources of abuse in organizations. Ramdeo and Singh investigate the goal of understanding abuse from different sources by comparing the consequences of abusive supervisor and abusive co-worker behaviors. They explain the process by which procedural justice perceptions shape employee reactions to workplace abuse. Their study intends to demonstrate the relevance of the social exchange theory to the domain of abusive co-workers and further expand our understanding of procedural justice perceptions. De Cieri et al. study examined the impact of power imbalance in organizational relationships and claim it explains associations between perpetrators and those who experience bullying. Stress, anxiety, and depression, as well as lower levels of work attitudes , are a few of the adverse employee outcomes that come from workplace aggression. Workplace abuse and stress are related to poorer mental health, including sleep disorders, depression, anxiety, post-traumatic stress disorder and symptoms, and psychological distress . This can be the case even for co-workers who are not directly victimized . Exposure to workplace stress has also been associated with increased musculoskeletal injuries and disorders and a higher cardiovascular risk score among flight logistic workers and flight attendants . The systems that are currently in place have proven insufficient to prevent workplace abuse , and its victims are often left without support, within their job or from clinicians, while navigating the fallout of these experiences . Recent studies indicate that abusive workplace behavior is a widespread social phenomenon that cannot be ignored or trivialized. The issue is prevalent, for example, in countries ranging from the U.S. and Canada to Europe, the UK, Israel, and South Africa. The importance of addressing this issue is evident, given the link between the quality of employment and mental and physical health, and developments in the labor market . The latter include globalization processes, the COVID-19 epidemic's influence on economic development, technological innovation, and changes in labor market demographics . Existing theoretical frameworks to define abuse are based on two approaches: harm to a person's dignity and harm to the physical and mental health and safety of the individual . However, these two aspects insufficiently capture the injustice involved in workplace abuse; the framework does not sufficiently provide guidelines to those who counter workplace abuse, whether management, policymakers, and regulators. Indeed, these areas of research and inquiry are relatively new in the research field . In general, however, it can be said that abusive behavior in the workplace is inappropriate conduct that causes damage to people's dignity, and mental and physical health and safety (see Hawkins v. United States, 2019). Following Guerrero , there is, consequently, a striking need to set appropriate behavior norms in the workplace. Furthermore, we would add that all parties associated with eliminating workplace abuse cooperate in standardizing its definition to ease these guidelines into court legislation, where necessary. Notably, protecting human dignity and freedom is one of the essential constitutional values of most legal systems; accordingly, the law is committed to preventing such harm. Abusive workplace behavior touches upon individuals, but other units in the workforce are likewise exposed to its nefarious consequences. Subsequently, although many concerns prefer to keep these issues among themselves, we propose that because of the severe damage to people's lives and livelihood, the treatment of the phenomenon cannot remain solely at the organizational level. Abusive behavior in organizations appears to be a widespread phenomenon. Although previous publications attempted to shed light on the causes and mechanisms through which it develops globally ), they have not addressed the full range of abusive behaviors but rather solely bullying, and have not looked into particular difference factors . In this paper, we present a brief overview of several aspects of workplace abuse in the hope that the paper will enhance awareness of the issue and the desired outcome of increasing workers' protection and keeping their dignity. Moreover, we conducted an empirical investigation to examine the association between two difference factors and the manifestations of abusive behavior of superiors to subordinates, using Tepper's inventory. The underlying thinking was that if the instrument proves effective in detecting distinct associations between the gamut of abusive behaviors and different organizational factors, it could be used as a diagnostic tool to prevent the phenomenon of abusive behavior. --- Components of abusive workplace behavior Work relationships can sometimes be painful and lead to actions that are inappropriate at least and, at most, abusive and violent. Nobody is excluded from this phenomenon, whether they are employers, managers, supervisors, or employees. The latter includes regular employees and other workers, irrespective of contractual status. Indeed, even trainees, whether interns or apprentices and contingent workers such as independent contractors and freelancers are subject to abuse in their respective places of work. Although there is no universally accepted definition of workplace abuse , an observer can detect the hostility expressed in many ways. These damaging activities include psychological abuse; humiliation; ostracizing fellow workers; sabotaging colleagues' work; slandering colleagues to co-workers or superiors; and shaming co-employees in social media groups. Examples of workplace abuse abound. They include shouting, overly harsh or unjustified criticism, threats, intimidation, and excessive monitoring. Of note, despite that during the COVID pandemic, work interactions were transformed because so many people worked from home, abuse nevertheless continued and intensified on the electronic networks . On-screen communications allow people to be less inhibited than during face-to-face dialog. However, offenses on social media can be easily documented and thus provable and should undoubtedly be incorporated into definitions of workplace abuse. Electronic communication, however, leaves scope for ambiguity and represents but one of several unclear representations or contexts of what constitutes workplace abuse. Consider, for instance, that such abuse incorporates direct and indirect negative behaviors that embody intimidation, hostility, and harm. Generally persistent, the abuse is manifested by individuals or groups toward other persons or groups at work. Moreover, the molestation can occur in a public or private domain, in real-time, or virtually . Adding to what we have labeled "ambiguity, " we might consider that while abusive behavior at work cuts across all sectors and genders , those actions might occur at varying frequencies. To what extent does frequency circumscribe the negative quality of abusive behavior? Similarly, some recorded manifestations of workplace abuse, such as excessive workloads, tend to be more common than others, while others are rarer, such as threats of physical violence. To what extent, we might question, are the two cases equal in severity? To add to the lack of clarity: workplace abuse is sometimes labeled as harassment or emotional abuse . Following D' Cruz and Noronha , known types of emotional workplace abuse incorporate abuse that is interpersonal or depersonalized that, in turn, may be internal or external to the workplace . As indicated, this typology is consistent with face-to-face and cyber communication described as traditional and virtual bullying . Because of the various representations and manifestations of workplace abuse, we find it helpful to group them into three major categories -incivility, harassment, and bullying. • Incivility. Following Anderson and Pearson , incivility is low-intensity, interpersonal, deviant behavior. Offenders scorn their targets, doubt their judgment, and address them unprofessionally with rudeness and disrespect. By way of illustration, a survey questionnaire consisted of twelve manifestations of supervisor incivility that included such items as "Shouted at you, " "Ignored or failed to speak to you, " and "Accused you of incompetence" . When targets were subjected to these incivilities over an extended period, victims expressed low job satisfaction, increased withdrawal from work, and intention to quit ). Among the crushing effects of bullying on targets are high stress and anxiety levels, sleep impediments, depression, and even suicidal thoughts . Workplace abuse is thus a multifaceted concept. Furthermore, there are international dilemmas and peculiarities involved in its delineation. Notably, research to delineate workplace abuse accurately is in its preliminary stages and has yet to achieve the conceptual analysis that sexual abuse has engendered. We believe, however, that accurate delineation of abuse that informs an employee's quality of life at work should ultimately be based on three foundations: physio-psychological, administrative-sociological, and legal. The physio-psychological foundation deals with aspects of personal-mental, physical, and safety-related injury of the kind discussed so far. The administrative-sociological foundation is anchored in Schneebaum's proposal that the state should regulate the power embedded in the authority of office rather than being left to the mercy of self-regulation: the employers should not be free to determine whether and to what extent internal authority positions should be regulated. To prevent abuse, the legislator should dictate a web of authority relationships that neutralizes the empowerment of employers and other positions of power . The legal-constitutional aspects of damage to human dignity are the meat of the legal foundation. While the theoretical foundations for examining workplace abuse as legal harm are usually not addressed directly in the literature , there are, however, two core frameworks for conceptualizing workplace abuse within that conceptual framework. The first defines workplace bullying as a safety issue dealt with by health and safety regulations focused on preventing injuries and burdening the employer with the responsibility for all risk-creating practices. The second framework focuses on the humiliation of the bullied persons and identifies abuse based on the right to dignity . Although the administrative and legal aspects of workplace abuse go beyond the scope of this paper, we propose, nevertheless, that this trinity of foundations should help to accurately outline the borders of the definition and serve to operate the mechanisms that identify workplace abuse on the local, national, and international arenas. Finally, in the interests of economy and brevity, we employ the catch-all term "abusive behavior" to describe these negative actions in the workplace. --- Various facets of workplace abuse 2.1. The trust gap, antecedents, and outcomes of workplace abuse 2.1.1. The trust gap Although abusive behavior is commonplace worldwide, employees hesitate to complain about their supervisors because they fear reprisals. This "trust gap" is particularly prevalent in the workplace because expressed grievances are frequently ignored and not treated with the appropriate level of concern and responsibility. For example, Vault Platform conducted a study of both American and British office workers. They noted that the trust gap not only applied to workers unwilling to say their piece for the reasons stated above but also that their employers were worried about damage to their reputation and jobs in the light of complaints about abuse. Of note was their comment concerning the gap between workers' expectations of the workplace and reality at work which raises the question of subjective versus objective perceptions of abuse and the degree of offense that constitutes mistreatment. Besides the psychological effects and related concomitants on productivity at work, we might add that the answers to these questions have consequences in courts of law where cases of abuse are heard. By way of illustration, we might ask to what extent overly strict or abusive supervision can be included in the definitions of abusive behavior. We cite Tepper , who, concerning the gap in expectations at work, defined abusive supervision as a subjective evaluation resting on "subordinates' perceptions of the extent to which supervisors engage in the sustained display of hostile verbal and nonverbal behaviors, excluding physical contact" . The stress is on subordinates' perceptions of their supervisors' conduct toward them and not on their managers' behavior, per se. This distinction is particularly germane because the same behavior can be interpreted differently by different workers or even by the same worker at various times . This phenomenon, consequently, might influence statistical research on the incidence of behavior abuse. For example, Fischer and colleagues found that in the U.S., 13% of interviewees reported they had experienced psychological abuse at least once a week. However, a review of specific studies on abusive supervision found that less than 2% of these incidents had been officially reported. Fischer et al. suggested that the empirical results of their survey were due to statistical analysis challenges. However, we might variously interpret their finding of rarely reported incidences of abusive behavior in the workplace. We propose that the result stems from the fears faced by both the perpetrators and their targets and their reticence to report , and the reservations faced by the relevant organizations facing negative publicity. --- Antecedents We have indicated that abusive behavior arises through possible deficiencies of character and upbringing. Indeed, Cowan recorded that bullying, in particular, has been perceived by H.R. professionals as a consequence of internal factors, such as management style, personality, and communication skills, and external factors, including culture and norms pervasive in contemporary society . Furthermore, these influential forces were perceived as not under the control of either the bully or the target in unfavorable exchanges. For an instance of a relevant external force, we cite Baker's observation that today's competitive work milieu, replete with promotion schemes and comparative performance review systems, creates a "winner-takes-all" culture that promotes workplace bullying. --- Outcomes As noted, research has documented the outcomes of abusive behavior in the workplace regarding psychological damage to targets and the broader economic implications for the victims of abusive behavior, the workplace, organizational productivity, and the entire economy. For example, Heilbrunn and Itzkovitz found that economic implications arise for organizations when abused victims embrace extended breaks, slow work, missing workdays, and reporting more workplace accidents -actions that harm productivity. In that context, and more germane to our discussion, Lazarus and Folkman's theory, for instance, throws light on the psychological effect of workplace abuse on horizontal solidarity and employment security. The theory distinguishes between emotional reactions to stressors that inhibit facing up to issues and problemcentered responses that confront specific problems. Emotional reactions reflect a pessimistic outlook on the likelihood of coping with the issue, while the latter response indicates a positive approach to overcoming challenges. Notably, Lazarus and Folkman distinguished between whether the Source of the abuse stemmed from supervisors or co-workers and their respective weightings on workplace insecurity and horizontal solidarity. Following this line of discussion, we propose that workplace abuse causes employment insecurity when victims feel they cannot cope with the threats posed by their perceived ill-treatment. Moreover, if the abuser is the worker's manager, the latter's behavior is construed as employment insecurity because managers control most of the employee's socioeconomic resources. On the other hand, abuse by co-workers who do not control resources would not be so construed. In a related study, Cortina et al. used the Workplace Incivility Scale to measure "experiences of disrespectful, rude, or condescending behaviors from superiors or co-workers" on a seven-item list of incivilities using a four-point scale ranging from 1 = "Almost never" to 7 = "Most of the time. " The results indicated that uncivil supervision creates a sense of employment insecurity, while incivility by horizontal co-workers does not, although it has some effects. Furthermore, uncivil supervision does not impair or damage the horizontal solidarity between work colleagues; however, if the incivility continues for extended periods, it affects the welfare and social fabric of the workplace . The supervisor, it appears, is the key player to be watched in instances of abusive behavior. --- The problem statement A positive and healthy atmosphere in the workplace is necessary for the well-being of employees while the impact of a workplace disruption on working patterns has a negative influence on all working actors. The present study aims to address this issue. The statistics presented in the introduction section provide evidence of the negative influence of workplace bullying on the workplace. The general problem we are addressing is workplace bullying . Specifically, we are trying to identify managerial abuse in organizations regarding the type of organization , gender of the perpetrator, gender of the target, and the managerial role, of each of the abuse items and by that to suggest directions to facilitate the diminution of behavioral abuse in the workplace. Though abusive behavior in organizations appears to be a widespread phenomenon, previous publications have not addressed the full range of abusive behaviors but rather solely bullying, and have not looked into particular difference factors. Therefore, the gap this study attempts to fill is by referring to the full range of abusive behaviors rather than solely bullying, as well as examining particular difference factors that weren't examined in previous studies. --- Our investigation Finally, in an additional attempt to conclude supervisors' abusive behavior in its various manifestations, we conducted an exploratory study in Israel, employing Tepper's Abusive Supervision Survey Questionnaire [ASSQ]. This exploratory investigation is timely because, despite Israel's stringent regulations, workplace harassment has not lessened in recent years, and no law is yet in place that frames bullying or harassment as an actionable cause for a claim and compensation. We focused on the supervisor-subordinate dimension because conclusions drawn from the aforementioned studies and a more recent investigation confirm that the incidence and degree of deleterious effects of supervisor-employee abusive behavior are of more significance than that of colleaguecolleague abuse. Notably, ASSQ is a valuable tool with which to track incidences of abusive behavior in organizations and to provide data that contribute toward mechanisms that attenuate the trust gap at work. Furthermore, we propose that ASSQ taps into the three components of abusive behavior described above: incivility, harassment, and bullying. In passing, we note that Tepper's questionnaire may also be used regarding employer/employee behavior toward contingent workers not on the payroll, such as independent contractors, freelancers, other outsourced workers, and even job seekers and job applicants. In this investigation, we additionally chose to focus on the various incidences of abusive behavior managers display toward their subordinates in organizations, with specific attention to the managerial/supervisory role, the victims' gender, managers' gender, and type of organization , --- Procedure Employees of two large organizations and another small national organization in Israel were approached via their WhatsApp groups, requesting their assistance in this investigation. We introduced the questionnaire , noting its anonymity and academic context and that subject participation was voluntary. Frontiers in Psychology 07 frontiersin.org --- --- The questionnaire The ASSQ questionnaire consisted of fifteen statements on a fivepoint response scale : 1. I cannot remember him/her ever using this behavior with me; 2. He/she very seldom uses this behavior with me; 3. He/she occasionally uses this behavior with me; 4. He/she uses this behavior moderately often with me; to 5. He/she uses this behavior very often with me. --- Statistical analysis We conducted a factor analysis to ensure statistical validity. We found, however, that the items did not split into well-formed and distinct factors, so the findings are based on a global variable, "abuse" . It should be noted that the loadings of the items on one factor were very high , which overcomes the relatively small sample size . We performed Pearson correlations to examine the relationships between the general latent variable and the abuse items and their intercorrelations . Additionally, we conducted t-tests for independent samples to compare the managerial role with each of the abuse items, respondents' gender, managers' gender, and type of organization , including the general abuse variable . --- Results Table 1 reveals that all the correlations are high, positive, and significant. Below is a list of items that predicted abuse in descending order. The numbers 1-15 in square brackets denote the item numbers as shown in Tables 2,3: • [4] Puts me down in front of others . • [8] Blames me to save himself/herself embarrassment & [10] expresses anger at me when he/she is mad for another reason . • [9] Breaks promises he/she makes, and [12] is rude to me . • [7] Doesn't give me credit for jobs requiring a lot of effort, and [15] lies to me . • [11] Makes negative comments about me to others . • [3] Gives me the silent treatment and [5] invades my privacy . • [6] Reminds me of my past mistakes and failures . • [2] Tells me my thoughts or feelings are stupid . • [14] Tells me I'm incompetent . • [1] Ridicules me . • [13] Does not allow me to interact with my co-workers . --- Discussion of findings Because the mean of the reported abuse was relatively low , we conclude that the incidence of abusive behavior in the All correlations are significant at p < 0.001; a full description of items may be found in Tables 2,3. 10.3389/fpsyg.2023.990501 Frontiers in Psychology 08 frontiersin.org respondents' organizations is not high. We explain the result by presuming that a particular behavior interpreted as abuse by one individual is not deemed so by another. Furthermore, circumstances might also prevail: a statement perceived as incidental in one situation might be considered harassment in another. However, the meager reporting does not necessarily mean that abusive behavior is not extant. We reiterate that the finding matches Fischer et al.'s conclusion that abusive supervision is underreported. As noted, these researchers saw that phenomenon as a statistical-research challenge rather than an indication that the phenomenon is nonexistent; we variously implied that the fears associated with the exposure of the abuse mitigated against reporting. Specific to the four areas of association we investigated, our findings indicate that workers reported more cases of abuse than managers, but, again, the mean differences between managers and employees were minor . On the other hand, men reported more cases of abusive behavior than women. Again, however, the differences found were insignificant and cannot serve to form conclusions that indicate a trend. Women managers exhibited fewer abusive supervisory behaviors than men. However, the mean differences between men and women were modest . Despite finding no backing for this outcome in the literature, the finding appears to be explained by Miller et al. supposition that in the organizational arena, the female management style is characterized by a normative ethical theory, "ethics of care" Gilligan -an approach that centers on interpersonal relations and cares as a virtue. Consequently, women are more likely than men to be creative, supportive, and innovative when supervising subordinates, even involving their workers in decision-making. The lower incidence of reported abuse among female managers could also be understood in line with studies that indicate that men display and report aggression more frequently than women , that women reflect more stable personality dispositions, and that men are more likely than women to aggress against others without provocation seemingly playing out their expected respective stereotypical roles in society . However, because the differences found in all the variables in our investigation investigated were not significant, we cannot conclude that the results indicate a trend. We further found that abusive behavior was less prevalent in the two public concerns than in the private organization. Again, we found no support for this result in the literature, but we assume that regulation in public organizations accounts for reducing the possibilities of abuse. --- Conclusions and recommendations There is increasing awareness that abusive behavior in the workplace is a widespread social phenomenon . Given the changing and developing labor market, the importance of addressing this topic is apparent . In Israel, for instance, there is as yet no appropriate solution for this issue, which has organizational, economic, and legal consequences beyond the psychological and mental health implications. We have recorded, however, that reports of the phenomenon are meager, despite increasing awareness, and thus we presumed that various phenomena account for that finding. They include: the same behavior could be interpreted as abuse by one individual but not by another; Statements perceived as incidental in one circumstance are regarded as harassment in others; and There is a propensity not to disclose instances of abuse due to the fear of consequences. --- Implications for future research and recommendations On the research level, we propose developing methodological tools adapted to further exploration of abusive behavior. As noted by a reviewer of this paper, we need to look more deeply into the sources of abusive behavior at work. Thus, future research endeavors could consider the variable of cultural values on the relationship between abusive supervision and the workplace and their impact on employees, customers, shareholders, and the community. We could envisage, for example, the effect of racial or sexual prejudice on the incidence of abuse at work. Furthermore, based on the findings in this paper, future investigations could shed further light on whether female managers display more or less abusive behavior at work than male managers. For instance, despite the overarching conclusion that women are less aggressive than men, when operative in an all-male milieu, the answer could be "more abusive" . Could the reason be the existence of the "Queen bee syndrome" whereby women with authority in a male-dominated environment treat subordinates more critically: that rigid stance then flies in the face of the "ethics of care" theory discussed above. Researchers could also further investigate how various HRM practices, such as organizational hierarchies, performance management procedures, and tournament promotion, foster abusive behavior at work. Additionally, research into the impact of the narrower organizational culture on behavioral abuse is imperative in efforts to attenuate abusive workplace behavior. We argue that despite the lesser reported manifestation of abuse by co-workers accounted for in the cited research findings, abusive behavior between employees deserves to be further investigated -if for no other reason, practical recommendations to circumvent abuse can be derived from the research findings. On a local level, management can institute a series of deterrence mechanisms that include tools to prevent inappropriate behaviors, such as punishment of abusers, correlated to the degree of exposure to the abuse by the victim. But, unless the root causes of workplace abuse are investigated, sanctions and legal remedies will never be sufficient. As stated in the Introduction, we recommend cultivating a positive organizational culture that eradicates the "trust gap" and encourages employees to speak up without fear of retribution or threat to their jobs. We urge companies -so much more aware of workplace sexual abuse and the need for diversity and sustainability in their organizations to raise consciousness and provide practical tools for eliminating work abuse and promoting staff cohesion. Examples include training measures, workshops, and sensitive supervision. We add that on the community and legal fronts, public discourse should continue to raise awareness of the issue and the desired outcome -increasing workers' protection and protecting their dignity -while governments should promote further relevant legislation to protect workers from abuse. To this end, we further recommend the appointment of an officer in charge of eradicating abuse in the organization, to whom employees can complain and who, additionally, would help courts identify abusive behavior when employees file lawsuits against a perpetrator. We also propose assigning court-appointed experts to examine abuse allegations utilizing the statements in Tepper's questionnaire. The experts would assist the courts in establishing the incidence of abusive activity, as well as its substantive nature and frequency in the workplace, especially where particular behavioral aspects of abuse are not included in legal definitions. Ultimately, all these measures aim to protect workers' dignity, the integrity of the organizations, and the welfare of society as a whole. --- Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. --- Ethics statement The ethics committee of Peres Academic Center headed by Professor Amos Drory has approved the study following a diligent examination of all ethical considerations. The approval carries the ordinal number of 1001 and is dated 27th of December 2021: Abusive Workplace Behavi. The patients/participants provided their written informed consent to participate in this study. --- --- --- Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Objectives: This paper explores definitions, incidences, and outcomes of workplace abuse -a widespread, worldwide social phenomenon impinging on the well-being of workers in a developing labor market -and possible directions for delineating and standardizing classifications of the phenomenon that facilitate workers' protection. Furthermore, we employ Tepper's Abusive Supervision Survey Questionnaire [ASSQ] to identify managerial abuse in organizations regarding the type of organization (public/private), gender of the perpetrator, gender of the target, and the managerial role, of each of the abuse items. Finally, we suggest directions for further research and practical measures designed to facilitate the diminution of behavioral abuse in the workplace in the foreseeable future. Hypotheses: No hypotheses are submitted in this exploratory paper. Methods: Employees of two large organizations (the Katzrin Local Council and Tigbur Ltd.) and another small national organization in Israel were approached, requesting their assistance in this investigation. We introduced the questionnaire, noting its anonymity and the academic context, and that subject participation was voluntary. The questionnaire consisted of fifteen statements on a fivepoint response scale ranging from 1 = "I cannot remember him/her ever using this behavior with me" to 5 = "He/she uses this behavior very often with me" (alpha = 0.97, M = 1.81, SD = 1.03). One hundred five respondents comprised the final sample: men (39%) and women (61%) aged 23-66. 66.7% of the respondents were employed in a public organization, 28.5% in a private organization, and 4.8% in a national organization. Years of education: less than 12 (14.3%); 12 full years (17.1%); tertiary education (10.5%); bachelor's degree (40%); and master's degree and above (18.1%). While 28.6% held managerial positions, 71.4% did not. Job tenure ranged between 0.5 and 42 years. The participants in national and private organizations were combined to avoid the problem of unbalanced groups. Notably, the gender balance of the managers was men (50.5%) and women (49.5%). To ensure statistical validity, we conducted a factor analysis and performed Pearson correlations to examine the relationships between the general latent variable and the abuse items and their intercorrelations. Additionally, we conducted t-tests for independent samples (with Bonferroni corrections for multiple comparisons: see, for example) to compare (1) the respondent's gender; (2) manager's gender; and (3) the managerial role with each of the abuse items, and (4) the type of organization (private/public); including the general abuse variable. Results: High, positive, and significant correlations were recorded between each questionnaire item and the total score of abusive behavior. The results ranged from r = 0.92 for the statement, "Puts me down in front of others" (r = 0.92) to r = 0.69 for the statement, "Does not allow me to interact with my co-workers" (r = 0.69). Reported abuse in the respondents' organizations was relatively low (1.81), while among women managers marginally less than their male equivalents (in absolute value, the mean difference in the mean scores fell between 0.28

INTRODUCTION China has witnessed rapid economic growth and social development since the 1980s, and many women have entered the labor market. Consequently, in most Chinese cities, it is common for middle-class women to have jobs as well as families of their own. This has led to an increasing demand for domestic workers to do housework and to care for children and the elderly. Responding to this demand, migrant women from less developed areas have entered cities, and currently provide a large supply of domestic workers. Owing to the lack of professional norms, the expansion of China's market for domestic work faces two prominent obstacles. First, employers currently lack standardized and effective guidelines for managing domestic workers. Second, domestic workers in China typically have low job stability and satisfaction. It is important to study domestic workers' job satisfaction because it affects key areas of care needs and regulation of domestic services. In contrast to other service-based occupations, the rate of informal employment among domestic workers is very high -approximately twice the share of informal employment of other employees -which is typical of unstable work. The high instability of domestic work is not only caused by this high rate of informal employment, but also by the ways that employers interact with domestic workers, and specifically their strategies of labor control. From an employer's perspective, the outsourcing of housework and the care of family members to a domestic worker can help alleviate the problems associated with the need to provide private care or the use of public care services, making it possible to achieve work-life balance . However, most employers have very limited knowledge about the means for controlling the labor process to maximize the benefits of outsourcing domestic work. Furthermore, labor controls shape the work experiences of domestic workers. Strategies of labor control that are too harsh may reduce the job satisfaction of domestic workers and lead to other negative consequences, such as low job performance , low organizational commitment , and a high turnover rate . These negative effects also represent the feedback of domestic workers on labor controls. Job satisfaction may also affect a domestic worker's motivation and service quality, which in turn affects the health and wellbeing of the recipients of domestic care . In general, low work satisfaction is unconducive to the formalization of domestic work and the development of service quality in this specific industry. The literature suggests that the factors influencing the job satisfaction of domestic workers can be categorized into two areas: the personal characteristics of the domestic worker , and the working conditions . Several scholars also emphasize the influence of social support, migration trajectories, work trajectories, and other factors . However, most of these studies suffer from at least one of the following problems: they lack the perspectives of the immediate employers of domestic workers and their labor control strategies, and instead tend to focus on the employers at the organizational level ; they tend to be limited in their scope, focusing exclusively on domestic workers that are employed by institutions, or on those who are employed directly by private employers, but never both; and they suffer from methodological limitations; for instance, most studies have been qualitative, and a few quantitative studies have used convenience sampling. Owing to the problems outlined above, limited generalizations can be made about the job satisfaction of domestic workers. China is presently undergoing rapid change. On the one hand, traditional customs still shape social life; on the other hand, modern technology-especially the Internet-has permeated all sectors of society. In a field survey for the present study, we find that in their interactions with domestic workers, employers in China use modern video-monitoring technology and also engage in the traditional Chinese custom of "hongbao gifts" . Accordingly, in this paper we seek to answer the following questions: How do these two contradictory behaviors achieve labor control over domestic workers? Specifically, we examine the impacts of employers' video monitoring and hongbao gifts on the job satisfaction of domestic workers. As we do not apply a specific sampling frame, we use the Respondent Driven Sampling approach to overcome the lack of representation caused by convenience sampling. --- LITERATURE REVIEW AND HYPOTHESES The concept of "discipline" has a long history in the field of domestic work. Many studies document control strategies used by employers, such as fictive kinship, schedules, and accommodation arrangements . However, relatively few studies analyze the impacts of labor controls on the job satisfaction of domestic workers or examine the mechanisms by which these impacts unfold. --- Tough Control: Monitoring via Video Cameras Labor control is a core concept in labor process theory. Studies on labor control have mainly focused on organized workplaces . Although workplace monitoring has been proposed to improve employees' task performance , it can also have negative effects on employees , such as reducing their job satisfaction, wellbeing, and motivation, increasing their mistrust and other negative emotions, as well as promoting turnover . Similarly, studies show that video monitoring can reduce the performances of the monitored persons in simple tasks , and create undesirable tension between managers and subordinates . Another negative consequence of video monitoring is a reduction in employee job satisfaction ; this occurs especially when an employee perceives the monitoring as a violation of his or her privacy . Disputes over labor monitoring are rife in domestic work owing to the privacy of the workplace as well as the lack of work standards and privacy protection. Thus, video monitoring may more likely have a negative impact on domestic workers. Domestic workers working in private householdsespecially those who live with their employers-frequently encounter difficulties in distinguishing between public and private boundaries. Many stay in their employers' houses throughout the duration of their contracts, where they may be made to share a room with the recipient of their care . Domestic workers are therefore vulnerable to privacy violations, and some may even lack privacy completely . As an "intimate stranger" in the home, the personal life and moral character of a domestic worker is constantly subject to the scrutiny and supervision of their employers. In recent years, the popularization of low-cost monitoring equipment has provided numerous convenient methods for employers to monitor domestic workers . The use of telephones for supervision and the discreet installation of "hidden" recorders or cameras in homes to monitor domestic workers have become common practice . Such monitoring of domestic workers allows employers to visualize the labor process, though they typically justify their behaviors by claiming that the monitoring is necessary to ensure the safety of individuals in the home . Our survey finds that many employers use cameras because of their low cost and convenience, and their ability to provide real-time monitoring with both images and sound. While employers insist that the cameras are being installed to monitor the recipients of care, domestic workers have shown mixed attitudes toward such installations . Only a few qualitative studies document domestic workers' views on video surveillance, such as their dissatisfaction with the installation of cameras , which violate their privacy and undermine their trust in their employers . However, it is not well known whether such video monitoring affects the work attitudes of domestic workers. We believe that examining the impact of video monitoring on the job satisfaction of domestic workers provides a more indirect indication of domestic workers' genuine attitudes toward video monitoring. Accordingly, we propose the following hypothesis. H1: The installation of video cameras in the home of the employer significantly lowers the job satisfaction of the domestic worker. --- Soft Control: Hongbao Gifts In traditional Chinese culture, "hongbao" are gifts in the form of cash in envelopes, which are presented for the purpose of maintaining relationships . Hongbao are generally used to consolidate social relationships. For instance, they can be used as a means for strengthening one's connection with relatives and friends, or to confirm one' s personal relationships. In China' s transformation from a traditional society of "acquaintances" to a modern society of "strangers, " the custom of giving hongbao has extended beyond the scope of one's relatives and friends; it has penetrated the industrial and commercial environment, where it commonly modulates the relationship between employer and employee. Monetary rewards have shown to improve employee job satisfaction in many different sectors of work , and in particular, financial returns have been shown to enhance employee job satisfaction in emotional labor . In China, employers often give hongbao to their employees at the beginning and end of the year, as well as during major festivals and celebrations. Such hongbao-giving not only serves the purpose of providing the employee with an economic incentive, but also-and more importantly-enhances the emotional connection between employer and employee, thereby increasing the likelihood that the employee will adhere to the employer's subsequent instructions. For instance, one study finds that presenting cash in the form of a hongbao rather than as ordinary cash, leads to a greater improvement in employees' overall productivity, willingness to participate, and work quality . The custom of giving hongbao can likewise be beneficial to the construction of an informal relationship between employers and domestic workers. Some domestic workers are willing to establish an intimate relationship with their employers, and they may proactively pursue this. They may regard the development of intimate relationships as a sign of their employers' respect . From these intimate relationships, domestic workers may also derive emotional recognition, which they cannot obtain simply from their employment contracts . Research shows that acknowledgment indicates that the recipient is worthy of trust, respect, and appreciation . Additionally, employees' job satisfaction tends to increase when they perceive respect from their employers or managers . We expect that similar direct psychological rewards induced by hongbao gifts will stimulate the transformation of relationships based on paid work to emotional relationships. In our opinion, as the medium of personal relationships, hongbao is the means used by employers to control the labor of domestic workers. Domestic workers who accept hongbao are more likely to be controlled by their employers. On the one hand, hongbao may have a positive impact on the labor process and results of domestic workers; on the other hand, it could consolidate unstable employment relationships of domestic service. Therefore, we predict that hongbao gifts will positively affect the job satisfaction of domestic workers. H2: An employer's hongbao gift has a significant positive effect on the job satisfaction of the domestic worker. --- Mediating Variable: Perceived Discrimination Domestic workers are a stigmatized group . Often, this discrimination relates to the gender or race of the domestic worker, or the enslavement traditions and class structure in the country of work . The position of domestic workers on the fringe of the labor market also reflects the exclusive labor policies in some countries or districts . These factors also cause domestic workers to have low professional statuses in society, and make them a group vulnerable to discrimination. Furthermore, domestic workers are vulnerable to discrimination in the workplace. Interpersonal discrimination, a component of workplace discrimination, includes negative verbal and nonverbal behaviors that occur in daily social interactions in the workplace . In comparison with other employees in the service industry, domestic workers spend greater amounts of time with their employers, interact more frequently, and are more likely to be the subjects of interpersonal discrimination. For instance, employers who play the roles of benevolent mothers believe that their domestic workers are weaker than themselves, and therefore need to be protected. Such employers confirm their class status and superiority in daily dealings . In contrast, their domestic workers may perceive discrimination. One meta-study shows that perceived discrimination is significantly related to job satisfaction . Numerous laws make clear provisions on discrimination in the workplace to prevent workers from being treated unfairly. Nonetheless, workplace discrimination continues to be manifested in inconspicuous behaviors and attitudes, placing the recipient of such discrimination at a considerable disadvantage. As suggested by Jones et al. while public discrimination is resisted by society, subtle discrimination is difficult to define and eliminate. Because employers in the private sector possess considerable power to decide and interpret their actions, it is difficult to judge their true intentions owing to the unclear boundaries between that which is "public" and that which is "private." It is also difficult to determine whether certain behaviors of employers toward domestic workers are shaped by prejudices. Therefore, domestic workers' perceptions may influence their understanding and interpretation of the behaviors of their employers. Cameras capture the real-time labor process of domestic workers. Such video monitoring may cause domestic workers to feel as though they are being discriminated against by their employers, thereby reducing their job satisfaction. For instance, a recent study shows that electronic monitoring affects employee satisfaction by changing employee cognition . Likewise, we expect that video monitoring will affect the job satisfaction of domestic workers . In contrast, we expect that the potential signs of gratitude and the recognition that employers convey through giving hongbao will reduce the discrimination perceived by domestic workers, thereby improving their job satisfaction. Accordingly, the following hypotheses are proposed. H3: Perceived discrimination mediates the effect of video monitoring on job satisfaction. H4: Perceived discrimination mediates the effect of hongbao gifts on job satisfaction. Material factors, such as an adequate compensation or a safe and comfortable physical environment, are key to ensuring the job satisfaction of domestic workers . If video monitoring-which is a "tough" means for controlling the labor of domestic workers-destroys the safety and comfort of the working environment; in contrast, hongbao gifts-a kind of "soft" control strategy-are a form of economic and psychological compensation that employers present to their domestic workers. We expect that employers' labor control over domestic workers can be strengthened via the use of either "tough" or "soft" strategies. Based on the above discussion, we propose a research framework . --- METHODS --- Data and Respondents The research data originated from a questionnaire-based survey on domestic work, which was conducted in four cities in China from June to August 2019. The surveys were conducted by the "Research on the Employment of Domestic Workers in the Internet Era" research group in the School of Social and Behavioral Sciences, Nanjing University. Domestic workers are usually employed in an informal capacity and many are not officially registered. For this reason, a sampling frame of domestic workers was difficult to obtain. We used the RDS approach in the survey to address this problem, "because standard probability sampling methods produce low response rates and responses that lack candor" . RDS is a network sampling technique that is typically employed to sample hidden populations that lack a sampling frame . The results of the RDS indicated that the dataset was representative. We explained the purpose of the research and obtained the informed consent from each participant prior to the survey. The respondents were domestic workers who had been serving in their employers' houses for extended periods of time . To reduce potential sources of bias in our analysis, our surveys excluded domestic workers who had worked for their employers for short periods of time , or who were paid by the hour, as such domestic workers usually served multiple employers during the same period. We also excluded eight domestic workers who were male, because these samples insufficiently represented the population of male domestic workers. Our final sample consisted of 699 domestic workers who ranged between the ages of 28 and 70 years . Approximately 88.4% of the respondents were married, and more than 70% of them originated from rural areas. The majority of respondents had a junior high school or lower educational level. --- Measurements --- Dependent Variable Job Satisfaction The "overall job satisfaction scale" developed by Tsui et al. is among the most widely used by researchers 1 . The scale includes six items that measure an individual's satisfaction with his or her work performance, organizational superiors, relationships with co-workers or peers, income, and promotion opportunities, as well as the overall job satisfaction. Domestic workers typically do not maintain any relationships with their co-workers or their superiors, and job promotions are extremely rare. Accordingly, we modified the scale by eliminating the item measuring the individual's satisfaction with a co-worker-relationship, and replacing the items measuring the individual's satisfaction with the superior-relationship and promotion opportunities with items that measured the individual's satisfaction with the employee-employer relationship and the occupational status. Thus, our modified scale included a total of five items; it measured a domestic worker's satisfaction with her job performance, income, employee-employer relationship and occupational status, as well as her overall job satisfaction. Respondents' answers to each item were rated on a five-point Likert scale that ranged from a value of 1 to a value of 5 . The Cronbach's alpha for the scale was 0.75, and the KMO was 0.763 ; this indicated that the scale had satisfactory reliability. --- Independent Variables Video Monitoring The participants were asked "Does your employer have a video camera installed in the home at present?" The response items were "yes, " "no, " and "unsure." Among the domestic workers surveyed, 25% of the participants clearly stated that their employers had video cameras installed in their homes, 66.48% said that their employers had not installed such devices, and 8.06% said that they were unsure as to whether any such devices had been installed. We combined the data for the "no" and "unsure" responses and coded them with a value of 0; we coded the "yes" responses with a value of 1. --- Hongbao Gift The participants were asked "Did you receive a hongbao from your employer in the previous year?" The response items were "yes" and "no." If a participant answered "yes, " she was subsequently asked to state the monetary value of the hongbao she received. Among the domestic workers surveyed, 62% reported receiving hongbao from their employers. The values of hongbao received ranged from 5 to 10,000 yuan . We used hongbao value as an independent variable, coded participants who had not received hongbao with a value of 0, and performed logarithmic processing . --- Mediating Variable Perceived Discrimination We measured the discrimination the domestic workers perceived by asking the question "Have you ever felt discriminated against because of your status as a domestic worker?" For the response items, we used a five-point scale, which ranged from a value of 1 to a value of 5 . Among the domestic workers surveyed, 61.39% reported "never" feeling discriminated against; while 21.92%, 10.33%, and 6.36% of the domestic workers reported feeling discriminated against "occasionally, " "sometimes, " and "often" or "always, " respectively. We treated Perceived Discrimination as a binary variable, that is, coding a response of "never" with a value of 0, and all other responses with a value 1. --- Control Variables Our model controlled for any variables that corresponded to participants' sociodemographic characteristics and work statuses . The social demographic variables included the age, years of education, household registration , and marital status of each participant. The variables corresponding to participants' work statuses included the number of years for which they had worked as domestic workers , the type of work they performed , whether they were living in their employer's house , whether they had signed a contract with their employer , and their monthly salary . --- Data Analysis We used a structural equation model in the software Stata16 to estimate the relationships between Video Monitoring, Hongbao Gifts, Perceived Discrimination, and Job Satisfaction. As the mediating variable Perceived Discrimination was a binary variable, we used the "gsem" command in Stata16 to fit this binary variable to estimate the model parameters. However, the "gsem" command could neither directly report the goodness-of-fit for the model, nor calculate the direct, indirect, or total effects. Therefore, in a second analysis, we included Perceived Discrimination as a continuous variable. We used the "SEM" command to estimate the model parameters in a linear probability model, and calculated the model's goodness-of-fit, and the direct, indirect, and total effects. Finally, we used the bootstrap method to estimate the direct, indirect, and total effects of the independent variables. --- RESULTS The results of the SEM estimated using the maximum-likelihood method are presented in Table 2. Perceived Discrimination was estimated using a logit model in Model 1 and using an OLS model in Model 2. Based on the results of the goodness-of-fit indices, both the SRMR and RMSEA of Model 2 were <0.05, indicating a good fit. However, the values of log-likelihood, AIC and BIC in Model 1 were smaller than those of Model 2, indicating that a better fit was achieved using the logit model. In both Models 1 and 2, the factor loadings of the four latent variables were all significant at the 5% level, indicating that the measurement model was effectively accepted. Models 1 and 2 both showed that the installation of video monitoring in employers' homes had no significant impact on domestic workers' job satisfaction. However, Video Monitoring showed a significant positive effect on perceived discrimination . That is, domestic workers who were subject to video monitoring were more likely to perceive themselves as being discriminated against. This result indicated that video monitoring did not directly impact the job satisfaction of domestic workers, but had an indirect negative impact on their job satisfaction through the mediating effect of perceived discrimination. The estimated coefficient for Hongbao Gift indicated that it had a positive effect on the job satisfaction of domestic workers . Specifically, for every 1% increase in Hongbao Gift, Job Satisfaction increased by 2.8%. Hongbao Gift also had a significant negative impact on Perceived Discrimination . That is, a domestic worker who received a hongbao gift of greater value from her employer was less likely to feel discriminated against. Furthermore, Perceived Discrimination was found to have a negative impact on the job satisfaction of domestic workers . Domestic workers who had perceived discrimination experienced 25.8% lower job satisfaction than others who had never perceived discrimination. The results showed that in addition to directly promoting job satisfaction, hongbao gifts could also promote job satisfaction indirectly by reducing perceived discrimination. To obtain the total effect of Hongbao Gift, Video Monitoring, and Perceived Discrimination on Job Satisfaction, we reported their standardized coefficients based on Model 2 . With reference to the standardized coefficients in Table 3, we found that the direct effect of Video Monitoring on Job Satisfaction was not significant. Instead, we found that its indirect effect and total effect were both significantly negative at the 0.05 level . This indicated that Video Monitoring reduced Job Satisfaction through Perceived Discrimination, in a complete mediation effect. Hongbao Gift had both a positive direct effect and a positive indirect effect on Job Satisfaction. The sum of the two variables-Video Monitoring and Hongbao Gift-yielded the total effect on Job Satisfaction . These results show that the role of Hongbao Gift on Job Satisfaction is part of the mediating effect. --- DISCUSSION There is an urgent need to promote the job satisfaction of domestic workers in China due to the increasing demands for housework and care, as well as the low professional standards of domestic work and the high mobility of domestic workers in the country. In this study, we investigate how domestic workers' job satisfaction is influenced by two labor control strategies used by Chinese employers: video monitoring, which has rapidly become popular in the private sector, and hongbao gifts, which have the function of embedding personal relationships within formal relationships. The former strategy of labor control uses cameras to extend panopticism to the private space; the latter is a means for employers to exploit "strategic intimacy" and control domestic workers to ensure their service quality, or to seek unpaid labor . In addition, we find that domestic workers' perceptions of employer behaviors, specifically whether domestic workers feel discriminated against, play an important role in the impact of labor control on job satisfaction. Applying SEM to data from surveys of domestic workers in four cities in China, we examine how the two labor control strategies-the installation of video cameras in the home for monitoring purposes and the giving of hongbao to domestic workers-affect the job satisfaction of domestic workers. The results verify our hypotheses that video monitoring and hongbao gifts significantly impact domestic workers' job satisfaction. Specifically, we find that video monitoring negatively affects domestic workers' job satisfaction and this relationship is completely mediated by perceived discrimination. In contrast, hongbao gifts not only directly and positively affect domestic workers' job satisfaction but also have an indirect positive effect by reducing perceived discrimination. The installation of video cameras in the home completely exposes the labor process of a domestic worker to an employer. Decisions on whether and where to install such cameras-and whether to inform and obtain the consent of domestic workerslie with individual employers. The uncertainty about the number and locations of cameras, as well as the unknown identities of the observers and the duration of monitoring make video monitoring more invasive than other forms of monitoring . The professional and private lives of domestic workers overlap considerably in both time and space, and it is difficult for these individuals to find "safe spaces" where they can escape the camera's uninterrupted gaze. The installation of video cameras in homes therefore causes domestic workers to worry about privacy violations and to believe that they are not trusted by their employers. It is worth noting that the domestic workers' perceived discrimination completely mediates the effect of video monitoring on their job satisfaction. This suggests that the video camera stimulates a domestic worker's perception of discrimination, and negatively affects her job satisfaction through this mechanism. Prior to entering employment in a private household, some domestic workers may already be aware that their occupation is discriminated against, as domestic work is sometimes treated as a "dirty" or "inferior" form of work. This perception may change in response to the behavior of the employers. More than half of the domestic workers surveyed indicated that they found it acceptable for their employers to install video cameras in their homes. However, this does not eliminate the negative effects of such monitoring. Domestic workers' attitudes toward cameras reflect their recognition of their own identity as "others" and their lack of "voice" in the private family. This represents a helpless and passive acceptance of the asymmetrical power relations. Some domestic workers even see video monitoring as a means to prove their integrity and to advertise their service quality. For instance, one worker caring for newborns and their mothers squared her shoulders and stated that she "very much supported " because "a clean hand wants no washing, " and thus video monitoring could help her employers address any suspicions they might have had concerning her integrity. Hongbao gifts are not only means for employers to encourage domestic workers to work hard and work more, but are also concrete representations of personal relationships in the Chinese context. The efficacy of hongbao gifts in domestic services can operate through the following channels. First, hongbao have economic value. In contrast to non-monetary gifts with less tangible economic values, the economic returns or compensation represented by a hongbao is clear. Second, hongbao act as incentives, conveying employers' recognition and endorsement of the work quality or personality of domestic workers, and their expectations for subsequent high-quality services, which may motivate the domestic workers. Third, hongbao act to embed emotional connections within a contractual relationship based on market logic. In traditional Chinese culture, the giving and receiving of hongbao takes place primarily during the Spring Festival, and is an intergenerational gift. In this context, hongbao represent the blessing of the giver on the recipient, and serves to reinforce the bonds with one's family, relatives, and friends. In contrast, the timing of employers' hongbao are not fixed; they may be presented on holidays or on the birthdays of domestic workers. In this context, hongbao represent employers' blessings and gratitude to domestic workers, strengthening the emotional connections between the two parties. This emotional connection, as well as the variable timing at which hongbao are presented, blurs the perception of unequal status in the employeremployee relationship, even providing domestic workers with some psychological return and increasing their loyalty. Finally, hongbao can function as a medium for personal relationships and thus reduce the mobility of domestic workers and stabilize their employment. In contrast to foreign domestic workers who operate in conditions of "legal servitude" , domestic workers encounter low costs in changing jobs in mainland China, allowing them relatively high mobility. The service quality of domestic workers in the current market varies considerably, and it is difficult to find "good" domestic workers. Labor controls such as video monitoring that potentially risk violating human rights may become reasons for domestic workers to actively terminate their contracts. In contrast, hongbao gifts are taken to represent the favor and face of the employers. In Chinese culture, gifts help to maintain the informal relationship between the two interacting parties. A rejection of a gift is interpreted as the rejection of the giver's kindness; such actions risk damaging the face of the giver and destroying the relationship. Therefore, employers use hongbao gifts and foster personal relationships to ease the inherent tension in domestic services. These labor controls can act as soft constraints on domestic workers' abilities or willingness to leave their jobs, and therefore reduce their mobility. In addition to increasing the income and internal motivation of domestic workers, which increase directly with their job satisfaction, the psychological return of respect and recognition brought about by hongbao gifts reduces domestic workers' perceptions of discrimination, and indirectly promotes their job satisfaction. In summary, our research demonstrates the distinct effects of tough labor controls and soft labor controls on the job satisfaction of domestic workers and verifies the mediating effect of domestic workers' cognition . In addition, our analysis of domestic workers' perceptions of discrimination aligns with the conclusions of previous studies, which show that discrimination at work negatively impacts an individual's job satisfaction (4, [67][68][69]. The two strategies continue to strengthen the labor control of domestic workers. The camera has introduced a novel, accessible means of labor control to employers-precisely at a time when few are able to supervise the work of domestic workers because of their own work commitments. The positive signals conveyed by hongbao gifts substantiate employers' gratitude in exchange for the work of domestic workers. However, hongbao are nonetheless a disguised form of labor control, designed to offset or alleviate the negative perceptions and work attitudes brought about by video monitoring. Our study makes three main contributions. First, we incorporate video monitoring, hongbao gifts, and perceived discrimination into an analytical framework for understanding the labor processes of domestic workers; we discuss employers' tough-and soft control strategies and the mechanisms through which domestic workers' cognition impacts their job satisfaction. Second, our study highlights how domestic workers' identities are shaped by modernity and society and emphasizes the influence of employers' behaviors on the attitudes of domestic workers to their jobs. Third, our quantitative analysis using the representative samples obtained by RDS sampling complements qualitative research on domestic work. To be clear, we are not calling for the prohibition of video monitoring or recommending that all employers present hongbao gifts to their domestic employees. Instead, by examining the impacts of two labor control strategies on the job satisfaction of domestic workers, we wish to draw attention to the subjective experiences of domestic workers. As one important limitation, the present research does not account for differences in the types of employment relationships, which have shown to play an important role in shaping employers' choices and motivations for labor control strategies, as well as the responses of domestic workers . However, owing to limitations in the data, we do not consider different types of employment relationships and their complex interactions; neither do we account for the emotional experience between employers and employees in this study. In addition, the symbolic meanings of hongbao gifts vary across different regional cultures in China. Such regional and cultural differences may influence domestic workers' perceptions of hongbao, and thus altering hongbao effects on job satisfaction and job performance. Future studies may therefore expand on the scope of the present research by exploring how the effects of different types of employment relationships and their interactions, the emotional experiences of employers and employees, and regional culture influence labor controls of domestic workers and their attitudes toward these controls. --- DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be provided by the corresponding author upon request. --- --- Conflict of Interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Publisher's Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Copyright © 2022 Yang, Fu, Liu, Fan and Jilili. This is an open-access article distributed under the terms of the Creative Commons Attribution License . The use, distribution or reproduction in other forums is permitted, provided the original author and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

There is a rapidly growing demand for domestic services among urban families in China. However, domestic work remains a low-status occupation with a high turnover rate. Focusing on the job satisfaction of domestic workers is useful to interpret this phenomenon. We investigate how the job satisfaction of domestic workers in China is affected by to two distinct labor control strategies used by their employers: the installation of video-monitoring devices in employers' homes (a "tough" control strategy), and the Chinese custom of giving monetary gifts, or "hongbao" (a "soft" control strategy). By analyzing data from surveys of domestic workers in four cities in China (N = 699), we find that video monitoring in employers' homes negatively impacts domestic workers' job satisfaction, and that hongbao gifts from employers significantly promote domestic workers' job satisfaction. The analysis of the causal mechanism based on a structural equation model suggests that video monitoring can increase the discrimination that domestic workers perceive, which in turn reduces their job satisfaction. In particular, we find that domestic workers' perception of discrimination completely mediates the effect of video monitoring on their job satisfaction. However, we also find that hongbao gifts significantly reduce domestic workers' perceptions of discrimination, and thus promote their job satisfaction; that is, the relationship between hongbao gifts and job satisfaction is partially mediated by discrimination. Our study provides a more comprehensive understanding of Chinese employers' labor control strategies and their effects on the job satisfaction of domestic workers.

POPULATION ATTRIBUTABLE FRACTION OF POOR ORAL HEALTH FOR THE INCIDENCE OF DEMENTIA: A POPULATION-BASED COHORT STUDY Xiang Qi, and Bei Wu, New York University, New York City, New York, United States Emerging evidence indicates a positive association between suboptimal oral health and increased risks of dementia. Despite the preventable and treatable nature of many oral health problems, oral health has not been considered in the risk factors for dementia that shape public health policy and research priorities. We aimed to quantify and compare the fractions of incident dementia attributable to poor oral health and other modifiable risk factors. The research included a nationally-representative sample of 15,474 cognitively intact adults aged≥50 from the Health and Retirement Study, with the outcome of incident dementia between 2006 to 2016. We estimated the multivariate-adjusted hazard ratio and population attributable fraction of edentulism and nine risk additional dementia factors. Additional analyses were conducted on 1,239 individuals who completed the "Dental Health Module" in 2008, examining oral health problems as an exposure. Edentulism was associated with a higher risk of dementia for both men and women and hypertension . Oral health problems were also a significant factor for men and women , with a PAF of 11.84% among men, the second-largest following less education . The substantial estimated proportion of dementia attributed to poor oral health in the aging population, particularly in men, warrants more attention. Dental care should be an integral component of a comprehensive approach to promoting cognitive health. Informal dementia caregivers tend to report more negative health impacts from the caregiving role, higher levels of unmet needs, and lower levels of service utilization compared to caregivers of persons with other types of illnesses. There remains a gap in knowledge of understanding unique correlates of dental care utilization among informal dementia caregivers. As guided by the Andersen model, this study examined correlates of dental care utilization among U.S. informal dementia caregivers. Data came from the Behavioral Risk Factor Surveillance System collected in 2016, 2018, and 2020 . Respondents were selected when they reported providing regular care or assistance to a friend or family member who has Alzheimer's disease, dementia, or other cognitive impairment disorder. Dental care utilization was dichotomized into "yes" versus "no" within the past year. Caregiving-related characteristics included caregiver role , intensity of care, duration of care, and type of care. Other correlates included caregiver characteristics, health behaviors, and physical and oral health needs. Findings from stepwise logistic regression models demonstrate that being a spousal caregiver, providing more than 20 hours of care per week, being male, having high school or less education, being low-income, unemployed, current smoker, having no health insurance, no current alcohol use, or less than adequate sleep was associated with a lower likelihood of vising a dentist within the past year. The findings suggest targeting spousal caregivers and those who provide intensive care to improve dental careseeking behaviors in dementia caregiver support intervention strategies. Abstract citation ID: igad104.0052 --- SOCIAL STRUCTURAL DISPARITIES IN DENTAL CARE UTILIZATION AMONG ADULTS IN CHINA'S MEGACITIES Xiaomin Qu 1 , Xiang Qi 2 , and Bei Wu 2 , 1. East China University of Political Science and Law, Shanghai, Shanghai,

Although the oral health status of the world population has improved over the years, structural and societal inequities in dental care and oral health are pervasive with older adults. This symposium discusses findings from three studies that examine structural and societal determinants of dental care and oral health inequities among diverse populations in the US and China. Mao's study investigated correlates of dental care utilization among informal dementia caregivers in the US. Findings demonstrate that spousal caregivers and those who provide intensive care are less likely to visit a dentist. These findings suggest that dementia caregiver support intervention strategies should target these populations to improve dental care-seeking behaviors. Qi's study examined the population attributable fraction of poor oral health on incident dementia in comparison with other modifiable risk factors. This study found that poor oral health, including edentulism and oral health problems, is a significant risk factor for incident dementia. The study highlights the need to integrate dental care as an integral component of a comprehensive approach to promoting cognitive health. Qu's study explored social structural disparities in dental care utilization among adults in China's megacities. This study revealed that social structural factors, such as hukou status, occupational status, and city of residence, are associated with disparities in dental care utilization. Overall, these studies shed light on important factors that impact dental care utilization and oral health. Addressing disparities in dental care utilization and oral health should be a priority in promoting overall health and well-being, particularly among vulnerable populations.

Introduction -The Impact of Intersecting Risks The co-occurrence of risk factors is a topic of study across multiple fields in public health. We have long recognized patterns of co-occurring mental health diagnoses and patterns of co-occurring medical diagnoses as age increases . Furthermore, research has highlighted the interplay between clusters of both mental health and physical health diagnoses. Patients coming to terms with a diagnosis of a chronic illness like HIV, for example, face an increased risk of depression [1][2]. Moreover, individuals struggling with depression are more likely to self-medicate through substance use, including alcohol consumption, and are more likely to engage in other risk taking behavior [3]. This additive effect of negative health statuses may be further complicated by a sociocultural context that supports potentially harmful behaviors. For example, there is evidence that individuals residing in communities with elevated rates of substance use, and increased poverty and assault rates are more likely to experience multiple negative health outcomes compared to relatively safer communities [4][5]. In short, the negative health outcomes an individual may face are complex as they often result from a chain of events situated in environments that foster such outcomes. The recognition that the accumulation of physical and mental illnesses is more than just additive -i.e., that the total burden of illness is greater than the sum of each negative health outcome-is referred to in the public health literature as a syndemic. The term syndemic was introduced by Singer [6] to address the relationships between health problems and the contextual influences that shape them [7]. One of the key goals in considering syndemics is understanding how the intersection of health problems unfold and, ultimately, affect general well-being. Syndemics occur in populations situated in particular locations, anchored in specific moments in time -hence the recognition and prevention of syndemics requires an awareness of the contextual factors that connect each disease outcome [8]. This paper considers the intersection of three epidemics: Alcohol use, interpersonal violence , and HIV infection. In particular, we focus on how these epidemics intersect among pregnant women in South Africa. We chose to concentrate on these epidemics among these women as rates of alcohol use, IPV and HIV among South Africa women are, on the whole, among the highest rates in the world. Furthermore, rates of fetal alcohol syndrome among South African children are also among the highest in the world. And finally, the presence of these epidemics is consistent with the synergy of disease, social context, and behavioral patterns that Singer [6] initially presented as forming a syndemic; one which he refers to as SAVA -Substance Abuse, Violence, and AIDS). A recent review of the SAVA syndemic in American samples confirms that the impact of substance use and IPV are synergistic -in particular, among studies focused on women, the SAVA syndemic emerges as a stable pattern: there are powerful relationships between substances use, the impact of IPV, poor decision making, and risk taking behavior [9]. There is great need for context-sensitive interventions, particularly for the world's most vulnerable populations. The development of context-sensitive interventions for pregnant women is urgently needed to improve not only the health of pregnant women but to create environments most advantageous for their infants. When pregnant women live in contexts that place them at a socio-economic disadvantage, support risk-taking behaviors, and normalize negative health outcomes, health consequences are compounded, each enhancing the next, thus amplifying the harmful effects for both the mother and potentially her unborn child. --- Critical intersecting epidemics among pregnant women The Impact of Alcohol use With regards to world regions, we observe the overall highest rates of heavy episodic alcohol consumption among men and women living in sub-Saharan Africa. [10] Although Africa has a per capita consumption rate consistent with world averages, men and women in this region report high rates of abstaining from alcohol use, therefore, alcohol consumption among those who do drink is alarmingly high. [11] Similarly, the proportion of South Africans who drink alcohol is fairly low , however, South Africa has one of the highest per-capita rates of alcohol consumption in the world among those who do consume alcohol [12][13]. Heavy alcohol use in informal drinking establishments or shebeens and taverns within many townships in South Africa occurs frequently [14][15][16][17][18][19][20]. Furthermore, South Africa has high rates of hazardous drinking and patterns of detrimental alcohol use tend to be more severe in South Africa [21][22][23], compared to other parts of the world [12]. It is widely believed that heavy drinking among poorer South Africans, particularly Coloureds , is deeply-rooted in the legacy of the 'dop' system. The dop system, named after the colloquial Afrikaans word which means an alcoholic drink, has a long history dating back to European colonialism. In the Western Cape province , the agricultural economy centered around grape and wine production. Within this system, the payment of farm workers in part with crude wine considered unfit to drink was established [24]. This system allowed farmers to get rid of surplus wine, and kept workers dependent and loyal due to their alcoholism. The dop system played an important role in farm workers' dependency on alcohol and in the social control exercised over the indigenous people of the Cape region. Indigenous people were enticed to work on farms with the promise of payment in the form of tobacco, bread, and wine [25]. Under this system, farm workers were well controlled [26], but the shackles of their addiction had far reaching effects. In particular, since women did not refrain from alcohol use during pregnancy, the greatest impact has been on mothers and the children born to them [24]. The dop system was officially banned in 1960 but continued under apartheid; the ban was only enforced in the 1990s following South Africa's democracy. Despite this change, there have been claims since then that the dop system is still active. However, in 2007, The Black Association of the Wine and Spirits Industry called for the establishment of an industry fund to change people's attitudes about drinking alcohol and to set up an institution to treat alcoholism. This effort led to judicial action in order to fight against the agricultural industry and, ultimately, to encourage government to address the outcomes of the dop system [27]. Some argue that farmers are still practicing the dop system today but the pattern of remuneration by means of cheap wine is cloaked under the pretence of farmers providing workers with a 'gift'. The amendment of the 1928 liquor act in 1961 did not address the provision of free wine as a gift to farm workers, which left a loophole through which the practice continues [25]. Perhaps more devastating than the possibility of the currently functioning dop system practiced by farmers have been the claims that farm workers across the Western Cape have created their own dop system since the official system was outlawed [28]. In a newspaper article published in 2012, Prof Viljoen , was quoted as stating that the dop system has been replaced by shebeens and on farms far from shebeens or from the bottle stores found in townships, there now exist mobile shebeens. Trucks stocked with alcohol travel to farms and communities to sell alcohol and will even sell alcohol on credit [29]. A cycle of drinking and paying off drinking debt persists to the extent that farm workers themselves have expressed concerns about the operation of mobile shebeens [30]. Therefore, despite the abolishment of the dop system, dependence on alcohol continues to play a role in enmeshing farm workers in a cycle of poverty and dependence [25] and beyond the rural areas in the townships of urban areas, regular and heavy alcohol consumption persists. During apartheid, illegal shebeens in black and coloured townships played an important role in the lives of the oppressed [26]; today, shebeens continue to thrive in impoverished environments. The legacy of excessive drinking continues to afflict these communities, causing widespread social damage and can only be addressed by changing the pattern of drinking in South Africa -a shift which has been argued as essential if the alcohol-related burden is to be reduced [26]. --- The impact of Interpersonal Violence Women of childbearing age are at the greatest risk of experiencing IPV across the lifespan [31]. The World Health Organization's [32] multi-country Study of Women's Health and Domestic Violence found prevalence rates for IPV as high as 28% -it is noteworthy that among the 10 countries included in the survey, the two African countries hold the highest rates of IPV. This international study also reports that women who experience IPV are more likely to categorize their health as poor or very poor compared to women who have not experienced IPV. The negative outcomes of IPV during pregnancy pose particular risks to mother and child including increased risk of sexually transmitted infections/HIV, miscarriage, preterm labor, induced abortions, pregnancy complications, hypertension, delivering at a low birth weight, physical injuries, stress, and -in extreme cases-is associated with maternal death [33][34][35]. Worldwide, rates of IPV are highest within impoverished communities and women with less education and fewer financial resources are at an increased risk for victimization. Efforts to improve employment and education rates can be obstructed by IPV, as assault and the fear of violence contribute to reduced school enrollment for girls [36]. Gender inequalities like these are common in many African cultures, which tend to organize social power around patriarchal systems, and underpin generational cycles of IPV [37]. Substance use -and alcohol use in particular -is also highly correlated with IPV [38]. Cultural attitudes that condone violence against women, in combination with attitudes that minimize the risks of the consumption of drugs and alcohol, create a context wherein violence against women is normalized, resulting in exceptionally high rates of IPV. Remarkably, estimates of IPV during pregnancy among women in Africa range from 23-49% and these rates have been identified as among the highest reported globally. Major risk factors for IPV among African pregnant women include being HIV infected, and having a history of violence, and alcohol and drug use [39]. This high prevalence rate may be unique to pregnancy because expecting a child is likely to put increased demands on limited family resources and on relationship dynamics as individuals adjust during the transition to parenthood [40]. In South Africa, violence is the second leading cause of premature death and homicide accounts for more than half of all fatal injuries among 15-34 year-olds [41]. South Africa's epidemic of community and personal level violence is fueled, in part, by alcohol use and attitudes condoning violence against women. More than half of female homicide victims are killed by their relationship partners, and both perpetrators and victims of violence often have high blood alcohol content [42]. Reviews across results from over 80 IPV studies have identified alcohol use as a significant predictor of violence perpetrationin particular, that there is a significant and large correlation between attitudes condoning violence and violence perpetration [43]. In South Africa, IPV history has been linked to an increased chance of problematic alcohol use for men and women [37], and drinking has been associated with greater odds of men perpetrating IPV and sexual assault [43,31]. These results combined suggest the relationship is cyclic: IPV histories increase alcohol use, and alcohol consumption contributes to IPV -this relationship most likely occurs through the process of lowering inhibitions, and increasing impulsive and risk-taking behavior. Indeed, research suggests that IPV victimization and perpetration are both better predicted by past IPV trauma histories than by substance use histories or any demographic variable [37,44]. "In South Africa, exposure to violence in childhood is as ubiquitous as it is in adulthood" [37] and, overall, reports of IPV demonstrate its remarkably high prevalence in South Africa: between 25% and 42% of study respondents report currently being in a violent partnership [45,[46][47][48]. --- The impact of HIV transmission The burden of HIV infection disproportionately affects women in Africa: it is estimated that 60% of the individuals who are living with HIV in Africa are girls and women. Furthermore, in sub-Saharan Africa young women are eight times more likely to be HIV infected than men. In all of sub-Saharan Africa, 40% of women living with HIV reside in South Africa [49][50]. Efforts to prevent maternal-to-child transmission have expanded globally with an estimated 53% of HIV positive pregnant women receiving PMTCT treatment. However, only 15% of HIV positive pregnant women accessing maternal health clinics are receiving antiretroviral therapy for their own health [49][50] The limited care available for many HIV positive pregnant women in sub-Saharan Africa is consistent with the lack of services accessible for these same women who are using and abusing substances. Although options for HIV positive pregnant women who abuse substances are limited, it is well known that substance use during pregnancy is a major concern in multiple sub-Saharan countries [51]. Recently, work completed by Chersich et al. [52] has outlined the intricate relationships between alcohol use, HIV transmission, and unintended pregnancies. Cherisch focuses on the detrimental influence of alcohol in the role of secondary HIV transmission and unintended pregnancies-arguing for a focus on the broader context these behaviors occur in. Similarly, Rotheram-Borus [53] has developed a community-level, behavioral intervention around addressing the role of HIV infection, alcohol use and pregnancy in the Cape Town, South Africa area. This work has paved the way for defining approaches to delivering intervention content relating to complex risk and is delivered by lay health care workers residing within the community where they are working. Recent studies focusing specifically on the use of alcohol among HIV positive pregnant women are limited in sub-Saharan Africa. However, available research suggests that the role of alcohol likely varies considerably across cultures, with studies from South Africa [54] and Nigeria [55] highlighting the different influences of alcohol among HIV positive pregnant women. Desmong et al. [54] notes that alcohol use among HIV positive South African pregnant women varies greatly, with those residing in the KwaZulu-Natal area reporting relatively lower of rates of alcohol use than other areas, yet women who do consume alcohol report high rates of alcohol consumption. However, Etukumana et al. [55] notes a lack of association between testing HIV positive and alcohol abuse. Overall, the impact of HIV on South African communities continues to bolster the presence of a syndemic among pregnant women in South Africa. Without considerable advances in access to HIV prevention education and tools, and without improved options for HIV treatment, the presence of a syndemic is likely to remain a key factor influencing the overall health outcomes of pregnant women. --- Synergistic Co-occurrence Taken independently, each of the epidemics of alcohol use, IPV and HIV transmission are a noteworthy and formidable threat to the health of pregnant women. When public efforts broaden to a global view , it is clear that threats to maternal and child wellbeing are synergistic. Their combination results in an alarming set of circumstances for South African women in particular. As depicted in Figure 1, HIV prevalence rates are greatest in Africa, with Southern Africa experiencing the highest prevalence of HIV infection compared to any other region of the world. Likewise, maternal mortality ratios are the highest in Africa, with central Africa being most affected. Furthermore, heavy episodic alcohol consumption among women is greatest in Southern Africa, including South Africa. On the whole, as is evidenced by Figure 1, South Africa represents the intersection of the highest rates of the aforementioned epidemics-a scenario characterized as a syndemic. By taking this broader view, prevention strategies can begin to address the bio-social context in which this syndemic is sustained. Moreover, Figure 1 depicts a public health scenario for Southern Africa that leaves little doubt that improvements in maternal child health are desperately needed. However, what we understand about the health of pregnant women in Southern Africa is also greatly limited. The information on maternal fatalities does not break down cause of death to distinguish between disease mechanisms and inflicted injuries , nor can we present reliable information on the rates of alcohol consumption among women who are pregnant specifically. Indeed, the lack of national-level data with much detail is an obstacle to forming timely, relevant, and effective intervention strategies. Remarkably, we were unable to identify any database containing FAS information: There appears to be no prevalence data on FAS for any of the 57 African countries. Sources have extrapolated estimates for FAS prevalence using North American incidence rates , but these seem vastly underinformed given the clear differences in health and human service resources available on each continent. The limited data available exists within single or specific regions within single nations , and these numbers indicate on a national level the potential for each generation to be suffering disturbingly high FAS morbidity burdens. [28] In the absence of more detailed information, first steps towards improving maternal-child welfare outcomes must be informed by what is at hand. Thus, it is imperative that we prioritize improvements in the understanding of various maternal and child health outcomes, and that we make efforts to assess the prevalence of FAS in developing countries. --- Prioritizing goals for future prevention efforts Meeting the needs of HIV positive pregnant women is challenging in resource limited settings. However, there are potentially multiple entry points into care, or during ongoing care, where intervention around substance use and HIV transmission risk behaviors can be prioritized. For example, recent work by Peltzer [57], lays out the framework needed to address reasons that pregnant women do not receive HIV testing or HIV test results. Although, in this work, a majority of pregnant women received their results , a significant minority did not, and, importantly, reasons for not receiving results may be addressable through counseling. Basic concerns around fear of learning one's results and confidentiality were most commonly reported. In fact, in a meta-analysis of 5 recent studies [39], IPV increased when a woman's HIV status was positive, hence some of the reluctance to engage in testing may center around basic safety concerns. "We need to understand how HIV status operates in a culture where female subordination is the norm and how together with other factors it increases pregnant women's risk for violence". [39] In our evaluation of available services for both HIV positive and negative pregnant women residing in informal settlements around the Cape Town, South Africa area, we suggest that relatively basic interventions could possibility have a profound effect on improved health outcomes. These interventions include: transportation to substance use treatment centers, social support groups for people actively engaging in substance abuse, the addition of non-alcoholic beverages at shebeens and greater accessibility to pregnancy tests for early detection. Moreover, safe places for women and families to reside who are exposed to IPV could serve as important options for breaking the cycle of violence. --- Conclusion Development of community-based, and therefore contextually-grounded interventions for vulnerable populations are desperately needed in Southern Africa. Ideally, these interventions should be informed by thorough needs assessments and through reliable estimates of the scope of risk behavior. This process will allow for targeting and addressing factors that support the HIV, alcohol use, and IPV syndemic . While the most effective and enduring strategies will be those that challenge attitudes that condone violence against women, educate the community about the risks of substance use -particularly during pregnancyand create an environment of support and treatment for HIV positive individuals, these cultural shifts will likely take decades. Positive steps to improve maternal-child health outcomes, however, can be taken in the immediate future. For example, efforts to estimate the scope of each epidemic can be extended beyond the few regions currently collecting incidence and prevalence data. Moreover, structural interventions that provide access to community-based support for both HIV treatment and substance use recovery -both with an eye towards IPV screenings -can serve as safe havens and provide much needed health education resources. Screening and brief interventions delivered in prenatal services may help improve access to substance use interventions for pregnant women, thus preventing FAS and other negative consequences of substance use during pregnancy. Visual graphics of global HIV prevalence, maternal mortality ratios, and heavy episodic alcohol use among women.

A critical factor for understanding negative health outcomes is acknowledging the synergistic quality that clusters of health problems create. An important step in addressing clusters of health problems involves gaining an awareness of the contextual factors that connect them. This paper considers the intersection of three mutually reinforcing health problems: alcohol use, interpersonal violence (IPV), and HIV infection among pregnant women residing in South Africa. We explore how SAVA (substance abuse, violence and AIDS)-a syndemics related theory-underscores the dire need to intervene in various areas of psycho-social health and general well-being. Based on World Health Organization data, we highlight the remarkably high rates of alcohol use, IPV, and HIV infection among South African women compared to women residing in other countries around the world. We conclude by highlighting the need for improved recognition of the intersection of these epidemics and for improved surveillance of the prevalence of alcohol use among pregnant women. Finally, based on the literature reviewed, we provide recommendations for future interventions.

Introduction Job satisfaction refers to the feeling that work can bring psychological and physical satisfaction for employees [1]. Organizational management prioritizes enhancing job satisfaction because a high level of satisfaction attracts, recruits, motivates and retains a company's workforce [2]. Although Internet use has become a part of most jobs, only a few scholars have estimated the direct impact of the Internet on workers' job satisfaction and explored the mechanisms of the effects that are exerted [3]. Most existing literature on job satisfaction discusses individual satisfaction by analyzing the antecedent variables, including salary income, career prospects, education, skill levels, work organization and nature, and work relationships [4][5][6]. Few studies have explored the direct relationship, and there is a gap in China. The Chinese context provides an ideal setting for studying the relationship between Internet use and job satisfaction for several reasons. First, China has achieved rapid economic growth, but the work time and intensity increase at the same time. The high-intensity work of laborers has brought about social problems such as sudden death and suicide from overwork, which have been fermented through the Internet and attracted widespread attention to the subjective welfare of laborers such as job satisfaction and happiness [7]. Second, the way Chinese people use the Internet is different from that of Western workers. For example, the Chinese used to communicate with colleagues by WeChat [8]. WeChat is an instant messenger, and new messages will likely be received during non-working time, affecting workers' lives. However, most Western workers used to communicate by email. Emails can be responded to centrally based on the recipient's wishes. The difference between the two is likely to lead to differences in job satisfaction. Third, Chinese people are excellent targets for heterogeneity analysis because of the huge urban-rural differences in Internet use. According to the 2021 Statistical Report on Internet Development in China, 70% of Chinese residents already use the Internet, but in rural areas, only 30% have this skill [9]. This leads to a huge difference in the ability of people in different regions to access information or search for jobs through the Internet, which provides a rare opportunity to estimate the heterogeneous impact of Internet use on job satisfaction. It is important to note that the impact of the Internet on work patterns and life may have a negative effect on job satisfaction. Internet use prolongs working hours, which brings greater time efficiency [10]. In addition, excessive Internet use will also reduce people's real-life information interaction, thereby reducing work enthusiasm [11]. Finally, the Internet is not yet the main work tool for migrant workers who are characterized by manual labor and repetitive tasks, so Internet use have little impact on job satisfaction [12]. Seashore and Tobor believe that job satisfaction is determined by a combination of external environmental factors and personal factors. The former includes the work environment and workplace relationships, while the latter includes family, education, income, age, and other personal characteristics [13]. Specifically, workers who are highly educated or have high incomes tend to have high job satisfaction [14,15]. Research findings on age are not uniform, for example, Peng et al. concluded that the effect of the Internet on older adults is not significant and negative emotions are mainly generated by younger groups [16]. However, more scholars believe that there is an inverted U-shaped relationship between age and job satisfaction because younger people are more likely to use the Internet for work and older people can learn new skills through the Internet [17]. Similar to Easterlin's paradox, a large part of workers' job satisfaction stems from comparisons with those around them. By making comparisons with others easy, the Internet reinforces workers' feelings of superiority or psychological disparity. How is Chinese workers' job satisfaction affected by Internet use? What exactly are the differences between people with different personal characteristics? We need to explore those questions. We use the China Family Panel Studies , a nationally representative and longitudinal survey, which tracks job satisfaction and Internet use of 83,012 samples biennially from 2010 to 2018. Identifying the causal effect of Internet use on job satisfaction has two challenges. The first is omitted variable bias . While the survey's panel structure allows us to control for individual fixed effects, there are unobserved and time-varying confounding factors that can potentially bias OLS estimates. Another concern is that the estimates on the effects of Internet use may be biased due to reverse causality. The higher evaluation of job satisfaction may also affect people's preference of Internet use. To address these two challenges, we refer to Sabatini and Sarracino instruments, using the urban optical cable line length as an instrumental variable [11]. On the one hand, due to the difference in the use of Internet communication technology, labor force Internet use and Local Area Network base station establishment and optical cable transmission speed, which is closely related to the length of the optical cable line; On the other hand, optical cable construction is greatly affected by regional topography, which is generally related to crustal movement and is not affected by job satisfaction. In addition, to address omitted variable bias, we flexibly control for economic variables and include fixed effects at individual and province-year levels. The baseline estimate suggests that compared with those who do not use the Internet, job satisfaction increases by 3.2 percentage points, which is consistent with the growing Internet use in China. Mechanism analysis found that Internet use weakened the negative impact of time efficiency on job satisfaction but strengthened the positive impact of work autonomy on job satisfaction. Compared to the existing literature, we have made three main contributions. First and foremost, we are the first to examine the mechanism of Internet use on job satisfaction in China. In recent decades, studies have been conducted mainly in Europe, the United States and Southeast Asia. Scholars found that the Internet, as a work tool, can significantly improve job satisfaction [18][19][20], but the impact of the Internet in China is still unknown. Second, we use the five-period micro-panel data of the CFPS from 2010 to 2018 to analyze the relationship between Internet use and job satisfaction in the urban and rural labor force, which provides important evidence at the micro level for evaluating the economic and social effects of Internet use. The database is widely used to study issues such as Internet use [16,21,22], mental health [23], and income inequality [22,24]. Third, we use the length of urban optical cable lines as an instrumental variable to solve the omitted variable and reverse causality problems and to improve the reliability of the findings. The heterogeneity analysis can help to understand the differences in the impact of the Internet on people with different incomes, education, etc. Finally, the PSM method was used to ensure the robustness of the results. This paper is organized as follows: Section 2 provides the theoretical analysis and research hypotheses. Section 3 describes the research model and the dataset, followed by the empirical analysis in Section 4. Section 5 proposes two possible mechanisms by which Internet use may affect users' job satisfaction. Section 6 provides the conclusion. --- Theoretical Analysis and Research Hypotheses Castellacci and Tveito summarize four channels through which the Internet affects job satisfaction. First, using the Internet may save workers' work time and increase productivity. Second, through social software, the Internet provides workers with a convenient and inexpensive channel for remote communication. Third, the development of Internet technology has created a large number of new jobs. Fourth, on the Internet, there is a large amount of information and data for workers [18]. We summarize these four channels as two mechanisms that influence job satisfaction: the information acquisition effect and the technology acquisition effect. The information acquisition effect refers to the fact that the Internet enhances the ability of individuals to access information, communicate information, and information selection in the labor market. First of all, the Internet can showcase and flaunt people's daily life, which undoubtedly enhances the well-being and job satisfaction of high-income workers. It can also lead to comparison and a desire for material things, which leads to dissatisfaction among low-income employees [15,25]. Second, the knowledge and information that exists on the Internet benefit a segment of the population that lacks information sources. For example, information on agricultural products facilitates farmers to share digital dividends, but it may be false or misleading. Communicating risks can lead to a decrease in well-being [26]. At the same time, the Internet has expanded the scope of information transmission and promoted remote work, thereby enhancing the individual's sense of well-being. For example, Bloom et al. conducted an experiment with employees of a Chinese Internet company in 2015 and found that job satisfaction increased by 7.22-15.8% for employees who worked from home through the Internet. The authors suggest that this can be explained in terms of both quieter and more convenient work from home [27]. However, there are also studies that suggest that family chores lead to decreased productivity [28]; employees have difficulty distinguishing between family and work boundaries, leading to overwork [29]. Both situations can reduce employee satisfaction. After the COVID-19 pandemic, many companies tried to encourage employees to work from home via the Internet [30], which provided a natural environment for experimentation. A large number of scholars have pointed out the possible negative impact of remote work on job satisfaction. Vyas and Butakhieo believe that during the home working period caused by the pandemic in Hong Kong, employees were under increased stress and felt exhausted [31]. Gibbs et al. studied a large Asian Internet company and found that workfrom-home hours increased significantly during the pandemic, while unit output decreased slightly. They believe there are three explanations: first, working from home is easily interrupted, second, it is more difficult to communicate with colleagues remotely, and third, there is less guidance between leaders and employees [32]. We argue that the above differences are caused by the specific type of work as well as the home environment. Specifically, Bloom et al. studied outcall workers, where the home would be quieter relative to the company. In addition, remote workers who live alone would clearly be less likely to be disturbed than workers with children. The technology acquisition effect refers to the impact of the Internet on individual skill levels and labor efficiency; the Internet, as the main source of acquiring external knowledge, has improved workers' skills, so that they can find solutions to tasks without relying on other colleagues [33]. This autonomy is conducive to improving employees' work motivation and job satisfaction [34]. However, there are also studies that have come to different conclusions. Chung-Yan, for example, suggests that the effect of job autonomy on job satisfaction is nonlinear. The positive effect rises with job complexity, but for low difficulty repetitive tasks, job autonomy may lead to lower job satisfaction [35]. Therefore, over-simplicity of work is one of the reasons for negative satisfaction with work autonomy. Based on the above theoretical analysis, we specify a hypothesis concerning Internet use on job satisfaction as follows in Figure 1. home is easily interrupted, second, it is more difficult to communicate with colleagues remotely, and third, there is less guidance between leaders and employees [32]. We argue that the above differences are caused by the specific type of work as well as the home environment. Specifically, Bloom et al. studied outcall workers, where the home would be quieter relative to the company. In addition, remote workers who live alone would clearly be less likely to be disturbed than workers with children. The technology acquisition effect refers to the impact of the Internet on individual skill levels and labor efficiency; the Internet, as the main source of acquiring external knowledge, has improved workers' skills, so that they can find solutions to tasks without relying on other colleagues [33]. This autonomy is conducive to improving employees' work motivation and job satisfaction [34]. However, there are also studies that have come to different conclusions. Chung-Yan, for example, suggests that the effect of job autonomy on job satisfaction is nonlinear. The positive effect rises with job complexity, but for low difficulty repetitive tasks, job autonomy may lead to lower job satisfaction [35]. Therefore, over-simplicity of work is one of the reasons for negative satisfaction with work autonomy. Based on the above theoretical analysis, we specify a hypothesis concerning Internet use on job satisfaction as follows in Figure 1. --- H1. Internet use can increase job satisfaction. As a key factor affecting job satisfaction is work efficiency, most studies show that individual weekly work efficiency positively correlates with job satisfaction, and longer working hours are not conducive to accumulating job satisfaction [36]. Using the European Quality of Life Survey , Pichler and Wallace found that workers who were constantly under tight deadlines and low productivity had lower job satisfaction on average [37]; similarly, Lopes et al. found that workers who did not adequately complete their tasks had lower job satisfaction [38]. From the perspective of the information acquisition effect of the Internet, the Internet helps employees perform work tasks in a fast manner through high-speed information dissemination and effective communication; and from the perspective of the technology acquisition effect, the Internet as a powerful office tool can help workers increase productivity and quality of work, and the time saved allows employees to allocate more time to high-reward and interesting tasks, thereby increasing job satisfaction [37,39]. The Internet is a powerful source of access to work information and external knowledge. On the one hand, Dettling Carlson believes that the Internet has expanded the scope of information transmission and promoted the process of telecommuting [40]. When employees perform tasks in a flexible and autonomous way, they are usually more motivated [41]. To a large extent, employees' job autonomy is defined by their organizational leadership, workplace flexibility, and the way work tasks are organized. The autonomy --- H1. Internet use can increase job satisfaction. As a key factor affecting job satisfaction is work efficiency, most studies show that individual weekly work efficiency positively correlates with job satisfaction, and longer working hours are not conducive to accumulating job satisfaction [36]. Using the European Quality of Life Survey , Pichler and Wallace found that workers who were constantly under tight deadlines and low productivity had lower job satisfaction on average [37]; similarly, Lopes et al. found that workers who did not adequately complete their tasks had lower job satisfaction [38]. From the perspective of the information acquisition effect of the Internet, the Internet helps employees perform work tasks in a fast manner through highspeed information dissemination and effective communication; and from the perspective of the technology acquisition effect, the Internet as a powerful office tool can help workers increase productivity and quality of work, and the time saved allows employees to allocate more time to high-reward and interesting tasks, thereby increasing job satisfaction [37,39]. The Internet is a powerful source of access to work information and external knowledge. On the one hand, Dettling Carlson believes that the Internet has expanded the scope of information transmission and promoted the process of telecommuting [40]. When employees perform tasks in a flexible and autonomous way, they are usually more motivated [41]. To a large extent, employees' job autonomy is defined by their organizational leadership, workplace flexibility, and the way work tasks are organized. The autonomy of work brought by Internet use can help employees choose a more comfortable working environment and working place, expand the limitations of traditional work in space and time, and improve employees' job satisfaction and work quality [42,43]. On the other hand, the Internet allows employees to carry out their work independently without relying on other colleagues by providing technical help, thus improving job satisfaction. To sum up, using the Internet improves employees' work autonomy, thus improving job satisfaction. Based on the above theoretical analysis, we specify a hypothesis concerning the mechanism of Internet use on job satisfaction as follows. --- H2. Internet use increases job satisfaction by increasing work efficiency. H3. Internet use increases job satisfaction by enhancing job autonomy. --- Dataset and Empirical Strategy --- Data Source and Variable Description We use five waves of the CFPS biennially conducted from 2010 to 2018 to estimate the relationship between Internet use and job satisfaction. The survey reflects the changes in China's society, economy, population, education, and health by tracking and collecting data at three levels: individual, family and community. We mainly select labor force samples in the age range of 16-64 years and eliminate invalid samples. Among them, Internet usage, job satisfaction, and individual characteristic data come from the adult database, and family characteristics such as family income, family size, family property, and the number of children in the family come from the family database. The adult database and the family database were matched based on the corresponding object code and year, and finally, 83,012 observations were kept. The explained variable is the job satisfaction of the respondents. Respondents were asked to rate their level of job satisfaction on a scale from 0 to 5, where 0 indicates very dissatisfied and 5 indicates very satisfied. The explaining variable is Internet use. According to "Do you surf the Internet" in the questionnaire, it includes computer surfing and mobile surfing. The respondent is set to 1 if they do and 0 if they do not. In addition, the control variables include individual characteristics and family characteristics. Among them, individual characteristics include respondents' age, age squared term, male, marriage, years of education, health status, popularity of work; family characteristics control family income, family size, family property and the number of children in the family. The higher the value of health and popularity, the better it is. Table 1 summarizes the key variables. 36% of people use Internet, their average age is 45 and they are in good health and popularity, and most of them are married. The mean of job satisfaction is 3.5, which means most of the workers are satisfied with their jobs. Note: This article is abbreviated at the 1% level, as in the following tables. --- Model Settings Since job satisfaction uses 5-point scales, we use the Ordered Logit model to study the impact of Internet use on job satisfaction of urban and rural laborers. The specific model is as follows: y * it = F Formula , Internet it is whether the urban and rural labor force uses the Internet, the subscript i represents the individual, and t represents the year. X it is a series of control variables and γ it and ∅ it are the fixed effects of individual and county by year. µ it is the unobservable error term. F is a nonlinear function whose specific form is as follows: F =            1, y * it ≤ µ 1 2, µ 1 < y * it ≤ µ 2 3, µ 2 < y * it ≤ µ 3 4, µ 3 < y * it ≤ µ 4 5, µ 4 < y * it In Formula , µ 1 , µ 2 ... µ 4 are the parameters to be estimated, and y * it are the unobservable labor force i's true job satisfaction in the year, and this variable is a latent variable t. There can be underlying endogenous issues with Internet usage. On the one hand, there may be a reverse causal relationship between Internet use and job satisfaction. Unobservable omitted variables such as acceptance level; on the other hand, the model may have a self-selection problem when performing regression, and workers may actively choose whether to use the Internet to improve job satisfaction according to their own conditions, preferences, or job nature. To alleviate the estimation bias caused by the self-selection problem, we also adopt the propensity score matching method for estimation. Using the Internet as the treatment group and not using the Internet as the control group, Equation is the probability of the worker sample in the treatment group, and Equation is used to calculate the average effect of using the Internet on job satisfaction. P it = Pr = L[h] ATT 1 = E -E --- Empirical Results --- Baseline Regression Result --- Table 2 reports the Ordered Logit estimates of Internet use on job satisfaction. Column of Table 2 studies the regression results of Internet use on job satisfaction of the full sample. The results show that Internet use improves workers' job satisfaction on an overall level. Column and Column consider the individual fixed effect as well as the county-year fixed effect, with the addition of individual and household control variables. Although the effect of Internet use on job satisfaction decreases from 19.3% to 3.2%, the coefficient is always positive. The relationship between age and job satisfaction is an "inverted U shape", indicating that age can improve workers' job satisfaction, but with the increase in age, its influence on job satisfaction gradually weakens. The higher the education level, the better job satisfaction. Finally, an increase in family size and the number of children in the home can, as expected, negatively affect job satisfaction. This is likely because workers have to take time out of their work to care for children or families, which leads to easy interruptions and reduced efficiency. In brief, our baseline regression results are consistent with most studies. We decided to conduct heterogeneity analysis for further exploration. --- Heterogeneity Analysis The results in Columns and of Table 3 show that Internet use significantly improves job satisfaction of urban workers, but Internet use has no significant impact on job satisfaction of rural workers. Urban workers are more educated and more likely to engage in multi-tasking and complex jobs that require multiple skills support, and Internet technology as a work tool can enhance urban workers' performance and job satisfaction [44]. However, these effects are arguably polarizing. Because the rural labor force is mainly manual labor and repetitive tasks, the Internet is not yet its main work tool, and its impact on job satisfaction is limited [45]. This empirical result is basically consistent with the research results of Salanova et al. [46]; that is, different groups are influenced by their educational background and economic status, which leads to large differences in the impact of different groups using the Internet on job satisfaction [47]. There may be differences in the impact of Internet use on job satisfaction among different income groups. Therefore, we divide it into five levels from low to high according to the "local status of personal income" in the CFPS data. The Grouped Ordered Logit regression results are shown in Table 4. The results in Table 4 show that the positive effect of Internet use on job satisfaction is mainly concentrated in the groups with income in the 20-80% quantile range. For both the highest and lowest income groups, there is a negative effect of Internet use on job satisfaction. We argue that the Internet reinforces material desires and income comparisons, which makes the lowest income group dissatisfied with their jobs. Unlike the low-income group, the highest-income group has a low marginal utility of Internet use and more spiritual pursuits, which may lead to insignificant results. Next, we also examine the effect of Internet use on job satisfaction under different educational backgrounds. The regression results in Table 5 show that with the increase in education level, the impact of Internet use on job satisfaction is also greater. The two are only significantly positively correlated in the group with college education and above and are not significant for primary school and below and middle school education. Maona et al. found that the less educated people have a poorer understanding and absorption of the Internet [48], and they are less likely to see the Internet as a work tool because they may not have sufficient technical literacy or Internet skills. Among the educational groups, the Internet has a limited effect on job satisfaction. --- Robustness Checks Considering reverse causality and missing variable problems, we use robustness checks. Column of Table 6 is the reduced form regression result, which is a direct regression of job satisfaction to instrumental variables. From the effect of optical cable length on job satisfaction through the Internet, the result is positive and statistically significant, indicating that the longer the optical cable length is, the wider the coverage of communication and information technology, and the greater the positive impact of Internet use on job satisfaction. Column reports the results of the linear two-stage least squares estimation of the instrumental variables and the results of the F value in the first stage, which shows that Internet use and job satisfaction are significantly positively correlated at the 5% level, and monthly post and telecommunications expenses in the first-stage regression are significantly positively correlated with Internet use. The F value of the first stage is 875.16, greater than the critical value 10, indicating that there is no weak instrumental variable problem. In addition, considering the heteroscedasticity problem at the district and county level, we select the heteroscedasticity robust DWH test with robust standard errors. The results show that the null hypothesis that there is an endogenous problem in Internet use is rejected at the 1% level. Therefore, these Instrumental variables are valid. In addition, the regression results of urban and rural laborers are basically consistent with those in Table 2 and will not be reported here. In addition to the endogenous problem caused by omitted variables and reverse causality, the model may also have self-selection problems. Referring to the estimation idea of propensity value matching proposed by Rosenbaum and Rubin, we adopt one-to-one nearest neighbor matching and radius matching based on Formulas and to further identify the causal relationship between the primary item of Internet use skills and the employment quality of migrant workers. As shown in Table 7, in the one-to-one nearest neighbor matching, the average treatment effect of the whole sample is 3.6%, which is significant at the level of 5%; the average treatment effect of the urban labor force is 5.2%, which is significant at the level of 10%; the average treatment effect of rural labor is 1.6%, which is insignificant. In addition, the estimated results of radius matching are basically the same, indicating that the estimated results in we are robust. The standardized deviation diagram of variables before and after matching shows that the standardized deviations of variables after matching are all less than 10%, indicating that the matching results meet the requirements of balance. Most of the observations are within the common range of values , so only a little sample is lost in the matching, which indicates that the model matches well. In addition, according to the fitting plot of the propensity scores of the treatment and Yes groups before and after matching , the fit between the treatment and Yes groups after matching was better than before matching. Only keeping a sample of respondents who participated in all five periods, we retain the sample of respondents who participated in all five periods to conduct a regression analysis on the relationship between Internet use and job satisfaction (see Table 8 --- below). Again, the results are basically the same, and the regression estimation results remain robust. Only keeping a sample of respondents who participated in all five periods, we retain the sample of respondents who participated in all five periods to conduct a regression analysis on the relationship between Internet use and job satisfaction . Again, the results are basically the same, and the regression estimation results remain robust. To comprehensively measure the impact of Internet use on urban and rural labor, we replace the core explanatory variable "Internet use" with respondents' "attitude to Internet use" for regression estimation. The specific operations are as follows: According to the data on the importance of using the Internet for study, work, social interaction, entertainment, and business activities among respondents in the CFPS adult database, the "Comprehensive Index of Internet Use Importance" is constructed using the entropy method. Table 9 reports the impact of the importance of Internet use on job satisfaction. The regression results are basically consistent with the benchmark regression results, and the results remain robust. --- Mechanism Analysis --- Internet Use, Work Efficiency and Job Satisfaction On average, employees who are constantly under tight deadlines have lower job satisfaction. Based on this, to explore urban and rural labor forces, Internet use will work by saving time to improve work efficiency, and have a positive influence on job satisfaction. We try to use "weekly working time " to measure the productivity of labor, using the "Internet to work", which shows that the greater the time pressure, the lower the work efficiency. The corresponding mechanism regression results are reported in Table 10. The regression results in Column of Table 10 show that labor force Internet use can reduce working hours and improve work efficiency. The results in Column show that labor force Internet use can improve individual job satisfaction by reducing working hours. --- Internet Use, Job Autonomy, and Job Satisfaction We select workplace flexibility that reflects one aspect of work flexibility, with autonomy as 1, and otherwise set to 0. Workplace autonomy can be regarded as one of the important mechanisms affecting Internet use and job satisfaction. Table 11 shows that Internet use significantly improves workplace autonomy, with a 5.4% increase in employee job satisfaction through autonomy. Our study supports the consistent findings of the majority of studies that concluded that the marginal impact of job autonomy ranged from 5% to 17.1% [3,49,50]. We believe the main reasons for the positive impact of job autonomy are innovative workplaces that can increase the diversity and interest of work and make those employed work harder. Meanwhile, workers can also self-coordinate their leisure time, thus increasing their overall job satisfaction. --- Conclusions We estimate the causal effect of Internet use on job satisfaction using the China Family Panel Studies. We find that Internet use significantly improves job satisfaction by 3.2 percentage points. Our results are robust to different specifications, instrumental strategies, controlling for workers' educational background and economic status variables, and using alternative measures of both Internet use and job satisfaction. We find evidence that suggests that Internet use induces job satisfaction through strengthening job autonomy and increasing time efficiency. The findings of this study are: Internet use can significantly improve the job satisfaction of workers, and this effect still holds after robustness analysis. Follow-up mechanism analysis found that Internet use increased job satisfaction through strengthening job autonomy and increasing work efficiency. There are differences in the impact of Internet use on job satisfaction among different income groups. The positive effect of Internet use on job satisfaction is mainly concentrated in the middle-income group, whose individual income is in the 20-80% quantile range. In the low-income class, Internet use will reduce job satisfaction; in addition, we also find that the higher the education level of workers, the more obvious the positive effect of the Internet on job satisfaction; finally, urban workers are more positively impacted by the Internet The impact of the Internet on job satisfaction is twofold. It will make workers interact less offline but will allow them to have more online interactions. It will both provide a great deal of knowledge for users to learn, but it will also tend to distract workers who become addicted to games, short videos, etc. Despite the many problems with the Internet, our research supports the opinion that the Internet will have more positive effects on job satisfaction. The Internet should be widely used by companies to improve job satisfaction. The simplest uses are online training, email notifications, etc. In the post-pandemic era, companies can also consider converting some of the work that does not require fixed office space, such as design and telemarketing, into voluntary telecommuting. This will not only reduce the company's staff density and improve staff motivation, but also reduce some of the site expenses, which is undoubtedly beneficial to the development of the enterprise. --- Data Availability Statement: The data used in this study are provided by the China Household Tracking Survey Database , which is open to the public . The datasets generated and analyzed in this study are available from the corresponding author upon reasonable request. --- Author Contributions: Data curation, D.Z.; formal analysis, D.Z. and X.L.; methodology, X.L.; validation, S.Y.; writing-original draft, D.Z.; writing-review and editing, S.Y. and X.L. All authors have read and agreed to the published version of the manuscript. ---

We explore the causal effects of Internet use on job satisfaction using a sample of 83,012 Chinese labor force members aged 16-64 years from the China Family Panel Studies (CFPS) from 2010 to 2018. We use ordered logistic estimation and find that Internet use significantly increases job satisfaction by 3.2%. Heterogeneity analysis shows that the Internet has a more positive impact on those who are in urban areas and have higher incomes and higher education. Our results are robust after eliminating endogeneity using instrumental variables and solving the self-selection problem using the PSM method. Our mechanistic analysis leads to similar conclusions to mainstream research, where Internet use induces job satisfaction by increasing time efficiency and enhancing job autonomy. Specifically, shorter working hours boosted job satisfaction by approximately 0.3%, while working in informal places boosted job satisfaction by 5.4%. Thus, employers may consider encouraging employees to access the Internet.

Introduction Historically, interest in the developmental significance of 'father absence' has come from a number of sources. Father absence is currently primarily the focus of Evolutionary Anthropologists and Psychologists with an interest in predictors of life history variation and reproductive development . However, there also exist substantial Developmental Psychology literatures addressing father absence from the perspective both of psycho-social wellbeing and, more pertinently to the current paper, gender development . The literature on father absence and gender development tailed off following Stephenson & Black's 1988 meta-analysis, with gender researchers growing interested in other developmental questions . More recent developments in the evolutionary literature, however, suggest that re-visiting the question of how father absence relates to sexually dimorphic behaviours may be warranted [5,6]. In the current paper, we first take an overview of extant data regarding links between father absence, physiology and gender development. Second, we test the proposition that father absence may masculinise children's developing sex roles and behaviour by comparing groups on multiple traits which are known to show sex differences in adults. --- Physical differences Timing of puberty and reproductive behaviour. At least 16 studies in the last 25 years have found that father absence prior to puberty is associated with earlier menarche in daughters in Western samples and there is evidence from studies of siblings that this association is environmental in nature [7]. Suggestions regarding the causation of this link have included the effects of stress on the endocrine system, the possible effects of exposure to familiar/stranger male pheromones and the effects of psychological reactions to stress on eating and therefore fat deposition . Because puberty is much harder to date retrospectively in men , there has been far less research into father absence and sons' development. Of what little exists, some has suggested that they too experience earlier puberty, while other researchers have found no links and others still have recently found father absence to predict later puberty in sons [9,10]. In addition to timing of puberty, father absence and parent-child relationships are also associated with timing of first coitus and even first pregnancy in offspring -perhaps because of earlier physiological readiness for reproduction, or perhaps for more psychosocial reasons arising from the parent-child relationship. One candidate psychological feature which might mediate a relationship between father absence and early coitus is impulsivity. Impulsivity has been linked not only with aggressive behaviour but also with willingness to engage in casual sex with multiple partners [14]. Physical masculinisation. Waynforth [15] found that amongst the Maya of Belize, men with separated parents showed significantly greater craniofacial masculinity . Similarly, Boothroyd and Perrett [6] found amongst a British student sample, that parental separation in early childhood was associated with increased ratings of facial masculinity in young women. Such differences could be of a genetic nature although the evidence of environmental causation above regarding puberty may make this less likely. Alternatively the link here may also be environmental such that either: stress or perhaps pheromone exposure in childhood affects pubertal hormone levels and thus affects facial structure; or, perhaps less likely, masculinised offspring could in some way be a catalyst to marital breakdown. In either respect, these data suggested that father absence may be associated with physical androgenisation and as such, that broader patterns of behavioural differences between father absent and father present offspring, may also be explicable in these terms. A key challenge in addressing this question, however, is the fact that the vast majority of research has focused on one or two measures at a time and as such falls short of giving a holistic view of whether father absence is indeed a moderator of gendered behaviour. In this following section, therefore, we will consider the different aspects of gender researched in this context so far: explicit gender role, and aggression, fearfulness, and impulsivity, which all show sex differences by adulthood. --- Gender role Research into father absence and gender role has historically produced rather mixed results. Some studies have found that father absence due to divorce was associated with low masculinity in boys and others found no effect . Stevenson & Black [3] attempted to synthesise the existing results in a meta-analysis of North American research up until 1986. They split studies on the basis of type of methodology, aspect of gender studied, age of subjects, and reason for father absence, amongst other factors. While there were insufficient studies to look at interactions between these factors, they did find that different groups of studies found different results. When split by participant age, confidence intervals confirmed that the significant association between father absence and feminine gender role was strongest in boys under 7 years of age, slight in 7-12 year olds and over 18s and absent in 13-17 year olds. This difference persisted in both lower and middle class samples . More importantly, when they compared different aspects of gender, they found that amongst boys and men, father absence was associated with feminine gender orientation and preference, while it was associated with masculine gender adoption . This is concordant with Beatrice Whiting's [20] suggestion that father-absent boys will have a feminine gender identity due to identification with the mother during childhood, and masculine behaviour as a reaction against this socially inappropriate feminine identity . However, Stephenson & Black also split the studies on reasons for father absence. They found that father absence due to unspecified reasons or military service was associated with a more feminine gender role, but there were no overall effects of father's death or parental divorce . Whether this means that father absence due to divorce has no effect on gender, or whether in fact half of the 'divorce' studies show masculine behaviour while the other half show feminine identity as above, is impossible to say. Furthermore, they found no effects amongst studies of white boys, only amongst studies of Black or unspecified populations. It is therefore very unclear exactly how father absence, particularly father absence due to parental separation, relates to gender role development in males. Similarly, there are mixed results in terms of the effects of father absence on girls' gender development, although there have been far fewer studies looking at girls. Hetherington [21] found that teenage girls with divorced parents , a dead father, or living with both married parents did not differ in terms of sex role. However, Vess et al [22] found that early parental divorce was associated with greater masculinity in girls, while late parental divorce was associated with femininity in girls . Stevenson & Black [3] found almost no overall effects of father absence on girls' gender development. When the studies were split by participant's age, 13-17 year old girls did show that father absence led to less sex-typed gender but there were no differences in any of the different groups for type of father absence, aspect of gender, race or socio-economic status . This apparent ambiguity in results is perhaps less surprising when one considers that the absence of the father from the family home does not necessarily imply the complete absence of the father from the child's life. Furthermore, in all types of family, the relationship between the parents and the child can vary greatly. Later research into fathers and child's gender therefore moved on to address the relationship with the father rather than his presence/absence per se. Unfortunately, however, this literature also found ambiguous results. For instance, Kelly & Worrell [23] and Jackson et al [24] both found that warm fathers were associated with androgynous sons , but while Jackson et al found the same pattern in daughters, Kelly & Worrell found no association between paternal warmth and sex role in daughters. Similarly, Orlofsky [25] found that rejecting fathers were associated with feminine sons and masculine daughters, but contrary to this, Stevenson [26] found that close fathers were also related to masculine daughters, and found no effect in sons. Most recently, Yang [27], looked at parent-child relationships and gender in Korean children and found no link between the two. Other researchers have considered whether there are different socialisation patterns in different kinds of households. For instance, Biller [17] looked at mother's encouragement of masculine/feminine behaviours and found that discouragement of masculine behaviours seemed to mediate the lack of masculine gender role in father-absent boys . On the other hand, some researchers have suggested that it is fathers which lead to greater gender typing in intact families. Siegal [29] conducted a meta-analysis of parental gender socialisation behaviours and found that although only half of 39 studies showed an effect of child gender on father's behaviour, there was a significant trend across the studies for fathers to treat their children differently depending on their gender. The effect size for fathers was significantly greater than that for mothers, with the latter not differing significantly from zero. Research in this area appears to have largely halted during the 1980s; a search of the literature only revealed the work of Yang [27] as cited above and [18]. The latter found that father absence was associated with less traditional gender roles in African American teenagers, particularly amongst lower SES families. A body of recent research, however, has been conducted with children of lesbian couples. These studies have reported no differences in gender identity and adoption between children raised by lesbian couples and children raised by heterosexual couples . This has two possible implications. Firstly it seems to support the notion that it is psychosocial stress and family attitudes which lead to differences between father-absent and father present individuals. Alternatively, it may suggest that some underlying genetic factor mediates both parental relationship success and offspring gender identity-such that children whose mothers have more lasting adult relationships may have different gender identity than children whose mothers experience relationship breakdown. More research is required, therefore, to determine exactly what about the absence of a father in the home causes differences in development. --- Specific sexually dimorphic behaviours --- Aggression and delinquency. A key dimorphic trait of particular relevance to father absence is aggression. Physical aggression is a strongly sexually dimorphic trait, with differences appearing before 2 years of age [31] and continuing into adulthood. Male-male homicide vastly exceeds female-female homicide in all societies for which data exists and men consistently show higher levels of physical aggression than women across both laboratory studies and questionnaires . Differences between children from intact and father absent/separated households in rates of physical aggression were first identified in the mid-C20th and there is evidence that father absence has continued to be associated with aggression, delinquency and problem behaviours across later decades. For instance, Cherlin et al [36] found in a 1960s British sample and a 1970s American sample that parental divorce was associated with greater problem behaviours in children . Hetherington [37] also found an association between father absence due to divorce and increased levels of aggression in teenage girls. Similarly, Berezckei & Czanaky [38] found using data from the late 1980s that boys whose parents were divorced and had been raised by their mothers engaged in significantly greater rule-breaking and delinquent behaviour during their teens than boys with intact families or with dead fathers. Again, however, they found no difference in girls. Stevenson & Black included aggression in their meta-analysis and found a significant overall association between father absence and aggression in both males and females . Furthermore, there is cross cultural evidence that amongst patriarchal societies, 'father absence' is associated with higher levels of assault and homicide [39]. Fearfulness and impulsivity. Fearfulness and impulsivity have a relevance to the father absence literature which is more secondary. Fearfulness is strongly sexually dimorphic, with women showing higher levels of phobias than men, rating identical situations as more fearprovoking, and being less likely to engage in physically dangerous activities [40,41]. Although men tend to show higher levels of social fears, women are more likely to be afraid of anything which could cause injury . Similarly, there is a significant sex difference in impulsivity across development . The gender differences in fearfulness and behavioural impulsivity have both been suggested as the psychological mechanism underlying the gender differences in physical aggression . One could therefore predict that the elevated levels of physical aggression in children of divorcees ought to be associated with lower levels of fearfulness and greater behavioural impulsivity. However, an extensive search of the literature yielded only one study which had measured fearfulness in children from different family backgrounds. Milne et al [45] observed that children living with both biological parents had lower levels of phobias than children who did not live with both parents. However, whether these latter children were specifically the children of separated parents , or whether they also included children who were adopted, fostered or living with step-parents was not clear. Regarding impulsivity, parental separation is associated with developmental disorders characterized by impulsivity and poor inhibitory control, such as ADHD and conduct disorder [46,47] and there may be a link between father absence and non-pathological impulsivity. In early research, Mischel [48] found a link between father absence and reduced delay in gratification amongst Trinidadian children, and Cortes and Fleming [49] found father absent boys were reported as more impulsive and irritable than father present boys using an economically deprived African American sample. Young and Parish [50] found no link between father absence and cognitive impulsivity in college age women, using the matching family figures test. More recently, however, Rostad and colleagues [51] reported that great father psychological presence/closeness is associated with less impulsivity in college women . Apart from these studies, it is difficult to find evidence using psychometric or laboratory assessments of impulsivity; research using other behaviors has however, found a link between father absence and, gambling and drug use [52], and sexual risk taking . --- Summary Overall, there appear to be four contradictory aspects of the development of father-absent offspring compared to father-present offspring: they appear to have no difference in, or feminisation of, their gender role identification; they have no difference or masculinisation of their gender role adoption ; limited evidence suggests they have increased levels of sub-clinical fears ; they may be more impulsive in some domains; they have a masculinised appearance in at least 2 studies. These factors appear to defy theoretical integration. Thus far it appears no research has sought to study these traits simultaneously within the same population, which is essential if we are to try and understand how they link together. The primary purpose of this paper was therefore to investigate the links between family structure during childhood and sexual dimorphism in multiple measures of behaviour and personality. Our hypothesis was that father absence and poorer quality family relationships are associated with greater general behavioural and psychological masculinity across both sexes. Participants were assessed using standardised self-report measures for aggression, impulsivity, fearfulness and sex role identity. These outcomes were then entered into a principal components analysis to investigate whether they would load onto a common 'masculinity' factor, such that a more masculine sex role would be associated with greater physical aggression, greater impulsivity, and reduced fearfulness. The factors derived from the principal components analyses, and indices of reproductive development were compared between participant family backgrounds. Should Boothroyd & Perrett's [6] hypothesis be correct, father absence is associated with hormonal masculinisation which drives both pubertal/physical changes and broader behavioural changes. We would therefore predict that both males and females with absent fathers will show a shift towards the more male-typical end of the general masculinisation scale . However, should other previous suggestions be correct, the higher levels of aggression in father-absent boys are a reaction to an initially feminised sex role identity, while father-absent girls have a similarly less gender-typical identity but no such later reaction against that identity. We would therefore predict that our sex role and behaviour measures would not share common variance and that men with absent fathers should show a more feminine sex role identity, coupled with higher levels of aggression, while women with absent fathers would show more general masculinity. --- Samples We tested the above hypotheses using a sample of young adult participants in the US and Australia, recruited via online repositories of social psychology research projects. Australia and the US are relatively high-parity nations in terms of gender empowerment: Their Global Gender Gap Index scores are 0.74 and 0.75, respectively [55]. The GGI ranges from 0, which represents the greatest possible inequality, to 1, representing perfect equality. Scores range from 0.51 to 0.86 for the 142 countries assessed in 2014. Australia and the US also have Global Gender Inequality indices of 0.11 and 0.28 respectively [56]. A GII of 0 represents 'perfect' gender equality and 1 represents the highest possible levels of inequality. Of the 155 countries with a GII in 2014, the range is 0.02 to 0.74 . Furthermore, Australia and the US have similar profiles in terms of Hofstede's national masculinity-femininity index , and the factor structure and sex differences found for the Bem Sex Role Inventory have been reported to be similar in Australia to those found in US samples [57]. Kashima et al [58] examined cross-cultural variation in sex differences in self-construal across five cultures including the mainland US and Australia . They performed multidimensional scaling on measures of seven traits including agency, assertiveness, and emotional relatedness to others. They reported striking similarities between the mainland US and Australian samples with regard to the pattern of sex differences. Some of our dimensions of interest show significantly greater sex differences within the US than other areas of the West, including Australia . However, some sex differences are larger in Australian samples . With regard to impulsivity, Cross et al [43] found consistent evidence that men score higher than women in US samples, but not in other Western countries . Their meta-analysis, however, only found one study with Barrett Impulsivity Scale data from Australia with a sample size of only 31 [62]. So, while the Australian estimate differed from the US mean, there is currently insufficient evidence that the two countries differ significantly with regard to sex differences in impulsivity. Perhaps most crucially, there is evidence that parenting may influence relevant traits such as aggression similarly in the US and Australia [63]. Furthermore, a meta-analysis examining differences in conduct problems between children from intact vs divorced or separated families found no difference between US samples and 'other' samples . --- Method 691 American and Australian participants aged 17 to 29 were recruited online via links on social psychology research sites. They were invited to take part in a study "investigating the links between childhood experiences and adult personality". The majority of participants indicated their ethnicity as 'white', and the vast majority reported being undergraduate students. Demographic details of participants from each country are given in Table 1. Participants reported whether they thought their parents' income when they were children had fallen into the top, upper middle, lower middle or bottom quartile ; the majority reported upper middle , or lower middle , followed by bottom and top . Ethical approval for the research was granted by the Science Faculty Ethics Committee at Durham University. Participants gave consent via button click and details of online sources of support for those experiencing distress relating to divorce were given at the close of the study. Participants completed a series of questionnaires online, firstly giving the demographic data above, and then reporting on measures of the following variables : --- Father absence Participants reported whether their parents had separated, who they lived with and at what age any separation occurred. Participants were classified as father absent if their parents had separated before they were 12 and they continued to live with their mothers, or if they had only ever lived with their mothers. Where parental separation occurred after they were 12, they were classified as father absent after 12. Those who lived with their parents until they left home were classed as father present. Age 12 was used in order to match the family relationship data below. Unfortunately, there were insufficient father-absent male participants to treat the two father-absent groups separately and so they were amalgamated for analyses with males. For women, we ran an analysis comparing all three categories, which was followed up with a secondary analysis which used the same two categories as were used for men. Furthermore, because there were only six father absent males in the US sample who reported age of first coitus and only three who responded to all behavioural questionnaires, all US males were excluded from father absence analyses. Anyone whose parents died, or who lived only with their fathers, or whose parents had joint custody were coded as missing. Furthermore, anyone reporting 'other' was also coded as missing. --- Family relationships Participants rated the warmth with which they recalled each parent, and the quality of their biological parents' relationship, before the participant reached 12 years, on a 1-9 Likert scale. This measure has previously been used by Boothroyd & Perrett [6,65] and has been shown to relate to bodily masculinity, age of menarche and facial preferences. It also correlates with Adult Attachment classification . All three variables were entered into a principal components analysis, which yielded a single component containing all three measures positively loaded and explained 71% of the variance. Scores from this component were used as a measure of family relationship quality thereafter. Aggression. Participants completed the Buss & Perry Aggression Questionnaire [66], a measure which was developed to assess aggression in school and college age samples in the United States, and which exhibits significant sex differences on the physical aggression scale both within and outwith the US such that males score higher than females . Participants rated the extent to which 29 statements were characteristic of them on a 5 point scale. Mean scores were calculated for each of the four subscales: verbal aggression , physical aggression , anger and hostility . --- Impulsivity Participants completed the Barrett Impulsivity Scale . As discussed above, although the BIS does not always show a sex difference, men score significantly higher than women on the BIS in the United States [43], and as such it was a pertinent measure for use in the current study. Participants reported on a 4 point Likert scale from Never/Rarely to Almost always/ Always the frequency in which they engaged in a range of behaviours. Mean scores were calculated for the three 'second order' subscales: inattentiveness , motor impulsivity and planning . Fearfulness. Fearfulness was assessed using the Fear Survey Schedule . The FSS was originally a clinical tool for identifying phobias, but has become a commonly used research tool in documenting both overall and specific fears in individuals. Participants reported on a 5 point scale from 'not at all' to 'very much', how much they were disturbed by a range of items such as loud voices, journeys by different modes of transport, tests, and being ignored. Mean scores were calculated for overall fearfulness , and for 'blood' fears and 'animal' fears which are higher in women than in men [40]. --- Sex role identity The Bem Sex Role Inventory was used to assess levels of masculine and feminine sex role identity. The BSRI was developed in the United States to document variation in behaviours typically ascribed to men and women . Individuals can vary in the degree to which they exhibit traits in each scale independently; however, generally men have tended to score higher than women on masculinity and vice versa for femininity. Although over time so-called 'masculine traits' have become less exclusively male in participant self-reports, there still remains a sex difference in femininity scores in the US , and we would anticipate likewise a sex difference in overall masculinity/femininity ratio. Participants rated how much 'like me' 60 personality characteristics were in 7 point Likert scales. From these, separate scores were calculated for masculine and feminine traits. --- Reproductive development Participants reported the age in years at which they had their first menstrual period , and first had sexual intercourse . --- Data processing Adjusting for age. All variables were assessed to determine whether age had any effect on an individual's scores. Because age is thought to relate nonlinearly to aggression in particular, peaking in the late teens and early twenties before declining again , variables were assessed using both linear and quadratic functions. No outcome measures were found to be significantly predicted by age , so no variables are adjusted for age. Principal components analysis on outcome measures. We hypothesised that all our gender-related psychometric measures would load onto a single factor. To assess this hypothesis, all outcome variables were entered into a principal components analysis with loadings below 0.3 suppressed. No rotation was used as we were specifically looking for a single factor. In fact the analysis produced two main principal components, as shown in Table 2. Factor 1 consisted of all aggression subscales, all impulsivity subscales, and fear. It also had a negative loading for BSRI femininity. We conceptualise this factor as general reactivity-i.e. the propensity of the individual to behave impulsively and react to situations negatively . Factor 2 consists of higher physical and verbal aggression, higher scores on BSRI masculinity, and low fear ; as such it can be conceptualised as general masculinity since it encompasses variables which are known to be higher in males than females. As well as loading positively on factor 1, planning and attention impulsivity loaded negatively on factor 2 -we return to this in the discussion. Factor scores were then examined for sexual dimorphism and country differences using linear models. For reactivity, there was a significant effect of country, such that Australians were more reactive than Americans , but no significant sex difference , or interaction between sex and country . For masculinity, there was a significant sex difference such that men scored higher than women , and an effect of country such that masculinity was higher in the US than in Australia , but no interaction . --- Results --- Father absence and factor scores We ran linear models with father absence and country as predictors of factor scores, for men and women separately. When women were classified as father present and father ever absent as the men were , there was a weak but significant effect of father absence, such that those who had ever experienced father absence were more reactive than those who remained father present . However, when father-absent women were coded as father absent before or after 12, neither of these groups had significantly higher reactivity than the father present group . When women were classified as father present and father ever absent as the men were, there was no significant effect of having ever experienced father absence on masculinity scores . When women were coded as either father absent before 12, father absent after 12, or never father absent, neither group had significantly higher masculinity than the father present group , but there was a significant interaction between country and father absence before 12 , which suggested that there might be a significant effect of father absence before 12 in one country but not the other. We therefore ran analyses separately for the US and Australia. In the US, the effect of father absence before 12 was clearly non-significant and in the non-hypothesised direction , whereas the effect in Australia of father absence before 12 was in the predicted direction , although very small in size and with a two-tailed p-value of .06 Australian men showed no significant effect of father absence for reactivity scores or masculinity scores --- Ratings of parents and factor scores Regression analyses were conducted separately with male and female participants, to assess the association between family relationship scores and both factor scores . Country and the interaction between country and family scores were added to these regression models. Overall, participant perceptions of parental marital quality were negatively related to reactivity scores, such that participants whose parents had a better quality relationship were less reactive; this was true for both male and female participants . There was no significant interaction between country and family relationship score for men or women . Participant perceptions of parental marital quality did not significantly predict masculinity scores for men or women . --- Reproductive outcomes Father absence did not significantly predict age of first coitus in Australian men . Women who were father absent before age 12 experienced first coitus significantly earlier than father present women in the United States . This effect was not found in Australia , and the significant interaction term for country indicates that father absence before age 12 affects first coitus differently in these two samples. There was no evidence of an effect of father absence before age 12 on age at menarche in women , nor a significant interaction between country and father absence . Regression revealed no significant associations between family ratings and age of first coitus in either sex , or menarche in women , except for an interaction term between family relationship scores and country in predicting age of first coitus in women: positive ratings of family relationships were associated with a higher age of first coitus in American women but not Australian women . --- Socioeconomic status Participants who were father absent before 12 , but not those who were father absent after 12 , reported a significantly lower parental income than those who remained father present. Reporting a positive family relationship was also associated with a higher self-reported income However, there were no significant associations between income and any of our outcome measures and further analyses controlling for income produced results that were the same as those reported above. --- Discussion The aim of the present research was to examine the behavioural and developmental correlates of childhood father absence in two similar Western countries, and establish whether these correlates could be explained as a general masculinisation of behaviour. Our principal components analysis indicated that this was not the case: two factors emerged, one of which did indeed show sex differences and which was categorised as 'masculinity'. The other factor, however, was not dimorphic and instead appeared to reflect 'general reactivity' . Although father absence before age 18 in women was associated with an increase in scores on reactivity, there was no link between father absence and 'masculinity' scores. Fatherabsent men were neither more reactive nor more masculine than father-present men. In both sexes, poor self-reported family relationships predicted higher reactivity scores, but were not associated with masculinity. We found no evidence that father absence was associated with changes in the gendered behaviour of female participants, which is concordant with the results of Stephenson & Black's [3] meta-analysis. Instead, father absence may predispose women to greater sensitivity to, and negative reactivity towards, the social environment in general. As such the data regarding elevated aggression in father-absent women/girls in the studies analysed by Stevenson & Black may in fact not be an index of behavioural masculinity but may instead be due to this greater reactivity. This also suggests that, for girls at least, the differing results regarding the association between father absence and gender development in previous studies may be mixed because the studies are accessing psychological constructs other than gender itself. Our results are, however, consistent with more recent suggestions that elevated stress reactivity may be an adaptive response to early experiences of highly stressful conditions . The association between poor quality family relationships and reactivity is particularly consistent with this hypothesis. In young adult men, our results showed no association between father absence and 'masculinity', indexed by sex role identity, low fearfulness, and high aggression. This is not consistent with the evidence for physical masculinisation in father-absent men and with Stevenson and Black's conclusions that behavioural measures at least showed a clear bias towards masculinity in father-absent males, nor with those studies which suggested that father absence was associated with a reduction in masculinity. A point of incongruity within our data was the difference between results using father absence as the predictor versus ratings of family relationship quality. While differences in outcomes between father absent groups were weak, correlations with ratings of family quality were much more robust and, as mentioned above, the results for reactivity were consistent across the sexes. This would suggest either that family conflict results in changes in emotional development , or that individuals who are more reactive in their late teens and twenties are more likely to view memories of their parents negatively. The fact that stronger results have been found using the subjective predictor variable than from the objective predictor variable would seem to suggest that recall biases might be influencing results. Alternatively, this analysis might suggests that poor quality parental relationships cause changes in development independent of any effects of father absence. We found effects of childhood experiences on reproductive outcomes in our American participants that did not replicate in Australia. In American females, father absence before age 12 and negative ratings of family relationships were associated with age of first coitus. To our knowledge, this study includes the first evidence that family background variables may relate differently to behavioural and reproductive outcomes in young adulthood in these two countries. Caution should be taken with interpreting these interactions, however, particularly in light of the modest size of the American sample and the fact that there were too few fatherabsent American males for analysis. Furthermore, our preliminary data processing suggested that Australian participants scored higher on general reactivity than American participants, and lower on behavioural masculinity. It is possible that higher 'baseline' reactivity and masculinity in the US is what causes father absence to have an effect on reproductive outcome here but not in Australia. Future work should confirm the extent to which these differences are not an artefact of self-report. A final point of note is that the relationship between aggression and fearfulness seems in the current sample to operate on two different bases-greater levels of physical aggression are indeed associated with lower levels of fear insofar as they both load on our factor of behavioural/psychological masculinity; this relationship may well be explained by underlying androgenisation and is concordant with the view of aggression posited by Campbell [35]. However, all forms of aggressiveness simultaneously load on our reactivity factor such that those who have more fears are also more hostile, angry and verbally and physically aggressive; these same individuals also seem to be more behaviourally impulsive. These two simultaneous relationships between aggression and fearfulness amongst a sample of young adults is consistent with conceptualisations of aggression as having different types: instrumental aggression is negatively related to fear-related factors such as harm avoidance, is not characterised by particularly high levels of emotional arousal, and is more typical of men than of women; while expressive aggression is positively related to fear, associated with high levels of emotional arousal, and is not consistently higher in men than in women [76]. Considering types of aggression may explain the sometimes differing findings regarding the relationship between aggression and fear, and between aggression and early life experience. Important caveats in the current study include the fact that most of our participants were students , and our reliance on self-report data. This is particularly pertinent considering our measure of fearfulness: while it is a standard measure of fears in self-report tests, it was not intended as a measure of trait fearfulness. New measures such as the Situated Fear Questionnaire [77] may prove useful for this type of work in the future. Related to this, although the Barrett Impulsivity Scale is one of the primary self-report methods for assessing impulsivity, it does not show reliable sex differences in all samples and may thus have failed to tap into the precise construct we were aiming to test . Finally, our family data were entirely retrospective child-reports; longitudinal, third-party assessments of family relationships would allow for a better test of whether the marital relationship itself is critical in the development of these outcomes, or father absence specifically. A final caveat concerns the age cut-offs we used. While considerable attention has been given to the 'effects' of father absence prior to 5-7 years, we selected a cut-off of age 12 for these family recall questions in order to give a realistically accessible window of memory for participants . We likewise used 12 as a cut-off for our father absence analyses in women to be concordant with that. Unsurprisingly, women who were father absent between the ages of 12 and 18 do not appear to differ in reproductive outcomes from women who were never father absent-father absence having probably occurred too late to influence the onset of menarche, for example. However, when we compared ever father absent with never father absent groups , we found effects of this variable on reactivity in adulthood. This suggests that father absence might have effects on offspring more generally even when it occurs too late to influence pubertal development. Alternatively, people who are generally reactive might be more likely to have parents who separate for other reasons. Our study revealed a relatively consistent picture of the effects of father absence in young women: women who were ever father absent showed slightly higher reactivity-but not higher masculinity-than women who remained father present until leaving home. This effect was not moderated by country. Women who were father absent before 12 also experienced earlier first coitus than father present women, but only in the US. A poorer quality of parental relationship, as assessed by subjective recall, was also associated with elevated reactivity in women, irrespective of country. These results suggest that father absence-and, more broadly, a stressful childhood or adolescence-result in a general increase in reactive behaviour in women, but not with 'masculinisation' of behaviour. For men the picture was less clear, partly because we did not have enough father absent men from the US for analysis. In Australia, however, father absence was not associated with increased 'masculinity' scores, increased reactivity, or earlier first coitus. Poor parental relationship quality was associated with increased reactivity in men, however, as it was in women. While our results illustrate the need for further cross-cultural research on the effects of father absence with large, representative samples-particularly of men-our results from women suggest that previous mixed reports of father absence in girls causing greater 'masculinity' might be attributable to greater reactivity being construed as masculinisation. --- All relevant data are within the paper and its Supporting Information files.

Research has previously found a number of apparently contradictory patterns in the relationship between 'father absence' (having a non-resident father during childhood) and the expression of gender roles, as well as other sexually dimorphic traits such as aggression. In the current study we measured a battery of sexually differentiated traits in relation to family background. 133 men and 558 women from the United States and Australia completed the Bem Sex Role Inventory, the Barrett Impulsivity Scale, the Fear Survey Schedule and the Buss & Perry Aggression Questionnaire. Principal components analysis found two main axes of variation in these traits. Firstly, a general 'reactivity' factor, on which aggression, impulsivity, and fear all loaded positively, was weakly associated with father absence in women. Secondly, 'masculinity' (consisting of high scores on masculine traits, low fear, and physical and verbal aggression) was not associated with father absence. Participants (except American males) reporting a poor childhood relationship with their parents also had high 'reactivity' but not higher 'masculinity'. We found some evidence of a link between father absence and earlier age of first coitus in American females (although not in Australia), but there was no link with age of menarche in either country. Overall, the current results suggest that previous findings linking gender development with father absence in girls may have arisen from a tendency towards greater externalising and reactive behaviour rather than a change in gender development per se.

Introduction This study focuses on the enthusiasm of contemporary young people for star-chasing. It is of great significance to study how young people understand their obsession with star-chasing and how the public can properly address the problem of excessive star-chasing. This study mainly discusses the aspects of the era background, characteristics of star-chasing, and solutions. By using the method of literature analysis, searching, and reading relevant materials and documents, it can deeply analyze the pros and cons of star-chasing, which is helpful for promoting research. The ultimate goal of this study is to explore the influence of star-chasing culture on young people and propose coping strategies. To achieve this, it has discussed fan culture and studied the comprehensive impact of popular culture on people's daily social lives as economic globalization, social informatization, and cultural diversification deepen. The star effect and the worship of star idols, as important forms of popular culture and carriers of dissemination, popularize popular culture and become an essential force [1]. People are the carriers of culture, which also means that culture shapes people. The worship of star idols is a transitional behavior during adolescence, since the reform and opening-up, teenagers mainly idolize and pursue star idols. --- Background In the era of the rise of the entertainment wave and the rapid development of new media technology, popular culture not only influences the living environment of the public but also shapes the image of star idols and popular culture through the cultural behavior of fans and the public, including the cultural relationship between fans and non-fans. Secondly, idol worship as popular culture has maintained and developed the relationship between idols and fans, as well as the emergence of fan culture into the public eye [2]. The development of fan culture has also been driven by the idol economy. With the booming entertainment industry, the number and influence of idol groups and celebrities have greatly increased. Idols promote their works through various channels to attract followers. At the same time, idol management companies have also realized the commercial value of fan culture and profit from fan activities and merchandise sales. This further promotes the development of fan culture. In summary, fan culture, as a new type of social culture phenomenon among young people, has rapidly emerged and continues to develop under the impetus of the Internet and the entertainment industry. It not only has a significant influence on contemporary young people but also has profound effects on the entire society and culture. Positive and uplifting fan culture plays a leading role among young people, but in recent years, distorted fan culture has had a negative impact on their sense of right and wrong, consumption outlook, moral standards, and independent personality formation. It also poses significant challenges to the ideological and political education of contemporary young people. Fans engage in commenting, supporting, and showcasing their love for idols in the "fan circle" to achieve self-identification and group identity [3]. Not only in China but also with the invasion of overseas popular star culture, this phenomenon has been exacerbated: the maturity of the Korean Wave in its star-making model has attracted a large number of loyal fans in China, especially K-Pop fans. The interactive process of fans interpreting, sharing, and discussing star-chasing practices in the online space has gradually surpassed the scope of ordinary entertainment and has become a discourse practice integrated into fans' daily lives. Alongside this interactive practice, a series of cultural phenomena have emerged in the process of media usage and consumption by K-Pop fans. Chinese K-Pop fans, in the process of cross-cultural idol worship, not only engage in localized interpretation and practice of foreign cultures but also maintain a high degree of political sensitivity. In recent years, the changes in the political and cultural context have made K-Pop consumption in China more complex. Chinese K-Pop fans face various challenges in the process of transnational fandom. On the one hand, there has been a rise in anti-Hallyu movements in China's online space, tarnishing the identity of Chinese K-Pop fans with nationalist stigmatization. The cultural disputes between the two countries hinder fan activities. On the other hand, Chinese idol groups that emulate the K-Pop style are rapidly developing, providing Chinese youth with more culturally familiar and politically safer choices for idol culture. Unexpectedly, despite these challenges, the number of K-Pop fans in China has not decreased. With the rapid development of new media, coupled with capital and traffic support, as well as the weak discernment of the youth group, the traditional fan culture has evolved into the fandom culture, which has many negative influences on adolescents. It wastes time and energy, affects personal cognition, leads to distorted consumption perspectives, and hinders the formation of correct moral awareness. Therefore, sorting out the impact of the current "fandom culture" on the values of young people, actively responding to and guiding new ideas for the education of young people's values, is conducive to promoting the correct formation of values from various perspectives and angles. --- 3. --- Overview The fan community has expanded into a massive Internet society with high levels of organization, actionability, and influence. Its culture and subculture deeply intertwine with various social issues, refreshing mainstream society and the academic community's perception of fan culture and fan communities. "Fandom" fans idealize idols as their own 'personal ideal,' projecting their empty selves onto the idols they aspire to be, seeking identity and self-affirmation. This projection often creates a significant gap between fantasy and reality, leading young people to indulge in their imagined 'perfect world,' resulting in a need for ideological transformation among adolescents. The idol worship culture in China has undergone a transformation from singularity to diversity, from elitism to populism, and from politics to culture. Chinese idols have developed from heroic idols to celebrity idols, from talent shows to grassroots idols, and from popularity-driven idols to nurtured idols. Currently, idol worship among Chinese youth exhibits generalization and high involvement. Celebrity idols are the mainstream, while second-dimensional idols reflect the characteristics of the times. Technological usage and consumer behavior have become norms in idol worship activities, with active self-awareness. Fandom culture is a typical form of youth culture, where young people who immerse themselves in idol chasing may experience irrationality, hollow life beliefs, and false collective behavior. The underlying reasons include the mismatch of satisfaction needed for adolescent growth, the implicit induction of group life, and the value influence of virtual environments. --- Influence Fandom culture has had a multifaceted impact on contemporary young people. Firstly, it provides a social platform for young individuals to find like-minded friends and share common interests and hobbies. This environment of mutual support and communication helps foster a close-knit community among young people, promoting emotional and psychological exchanges, and enhancing their sense of belonging. Secondly, fandom culture cultivates young people's teamwork and organizational abilities. Within fandoms, fans actively organize online and offline activities, create fan merchandise, and plan celebratory events for their idols' birthdays, among other things. These activities require fans to collaborate and work together, nurturing their spirit of cooperation and organizational skills. However, if "fangirls" are not guided in a targeted manner, they may easily become obsessed. To some extent, the chaos and behaviors that occur in the "fan circle" have brought negative impacts on the values of some young people, greatly affecting the positive cultivation and practice of their values [4]. Taking the recent highly popular Chinese boy band 10th-anniversary concert as an example, nine out of ten trending topics are related to them. One of the topics is about the excessively passionate behavior of fans: two groups of fans even got into a physical fight in order to secure the best cheering spot for their idols, resulting in further chaos at the venue, which was already in a disordered state. This disorder extends to the online arena, where fan wars are even more intense, featuring various means of verbal disputes and conflicts. In addition to the above-mentioned incidents, forced hugs, hotel voyeurism, airport delays, online verbal violence, and excessive obsession to the point of emotional involvement are common negative fan incidents. "Fangirls" and "fanboys" among young fans have sparked widespread discussions in society. These incidents reveal that excessive idolization and obsession with celebrities often lead to irrational and serious consequences and social issues [2]. It is necessary to strictly control online order, enhance media control, and also strengthen the guidance on fans' mentality to change their unhealthy obsession with idol worship. "Fan culture" not only has negative impacts but also brings psychological comfort to fans. Idols create a perfect image, which fans love and are influenced by. They spread positive energy to fans, some of whom take idols as role models and motivate themselves to become like them. At the same time, some fans add happiness and a mental pillar to their lives through idols. They see idols as mirrors, reflecting the person they most want to be. --- Strategies Teenagers are the future of a nation, and the subcultures they create are diverse and significant. What sets them apart from mainstream popular culture is that they emerge and spread rapidly within the youth community. Not only do they form a distinction from the general adult population through the use of new media technologies, but they also undergo rapid changes, elimination, and succession in cultural forms. "Fandom culture" originated on internet platforms and quickly rose to prominence with the development of internet technology. A healthy online environment is a vital external force for its healthy development. To ensure the healthy development of "fandom culture", it is necessary to create a clear online environment, regulate it, and govern unhealthy online environments. The "obsessive" nature of "fandom culture" not only disrupts normal communication order but also directly harms the healthy growth of young people who are in a crucial period of development, thus becoming a pressing issue that various sectors of society urgently need to address. In order to deal with the negative impact of fan culture, it is necessary to take some countermeasures and suggestions. Firstly, families and schools should strengthen the education of young people, guiding them to have a proper attitude towards fan culture and balance their interest in idol worship and drama-watching with the needs of normal life. Parents can provide children with more meaningful activities and hobbies to cultivate their diverse interests. Secondly, social media platforms and idol agencies should enhance guidance and management of fan behavior. By promoting correct values and moral standards, fans can be guided towards positive and constructive idol worship, while strict measures should be taken to combat online violence. At the same time, idol agencies should focus more on the healthy development of idols and avoid excessive packaging and commercial exploitation of young idols. Lastly, young people themselves should maintain a rational and objective mindset, avoiding excessive immersion in fan culture. While pursuing idols and dramas, one should always remember their own identity and goals, striving to balance entertainment with education, and work, in order to participate in fan culture in a healthier manner. Through sorting out the harm of the fandom culture, the "Clean and Rectify Fandom Chaos" special action aims to promote the healthy development of the fandom culture [5]. Adolescents are a vital force for advancing the modernization of national governance. To promote the healthy growth of adolescents, it is essential to guide the positive functions of the fandom culture in terms of values, purify the interpersonal ecological network of the fandom in terms of communication methods, and regulate idol and fan behavior to eliminate extravagance and waste [6]. With a scientific understanding of the fandom culture, while respecting differences and interests, it should govern uncivilized and non-compliant aspects of the fandom culture. By supervising and restraining negative fandom culture, it can create a harmonious and positive online environment. To address the problem of rational confusion among adolescents caused by the fandom culture, it is necessary to create a social environment conducive to rationality, cultivate a nurturing ground for rational incubation in schools, and establish a support system for rational growth within families. This will lead to the rational return of adolescents and help mitigate the negative impact caused by the fandom culture [7]. Public opinion should be positive, taking positive energy as the starting point, allowing idols to truly become role models so that university students can enhance their cognition and enrich their spiritual world in a relaxed and happy atmosphere [8]. No media should maliciously hype the uncivilized phenomena of the fandom culture [9]. Strategies and recommendations for dealing with the fandom culture include strengthening educational guidance, enhancing the management of social media platforms and idol agencies, encouraging diversified interests and participation, as well as promoting personal self-regulation and consciousness shaping. Through efforts from various aspects and shared attention, it can better guide young people to view the fandom culture correctly and let it become a part of enriching life rather than a source of negative influence [10]. --- Conclusion The research findings show that the influence of fan culture on contemporary young people is always a double-edged sword, but overall, the benefits outweigh the drawbacks. Therefore, the research conclusion indicates that with proper guidance, being a star-chaser does not generate the negative impact that the public often assumes, and positive advice can even lead to unforeseen positive effects. This research has provided valuable reference significance for future studies in this direction, mainly focusing on how to deal with excessive fanaticism toward star-chasing fan culture. Future research should delve into the positive guidance direction of star-chasing fan culture.

This study aims to explore the influence of star-chasing culture on young people and corresponding strategies. To achieve this goal, a discussion on fan culture was carried out, and the comprehensive impact of popular culture on people's daily social life was studied as economic globalization, social informatization, and cultural diversification deepened. The star effect and the worship of star idols, as important forms of popular culture and carriers of dissemination, popularize popular culture and become an essential force. The worship of star idols, as a part of popular culture, promotes the connection and development between idols and fans, fans and fan culture, gradually entering the public's view. With the rapid development of new media, supported by capital and traffic, the youth lacks discernment, and the traditional fan culture has upgraded to a fan circle culture, which has caused many negative effects on teenagers, such as wasting time and energy, influencing personal cognition, leading to deviation in consumer attitudes, and hindering the formation of correct moral concepts. Fan circle culture not only brings negative influences but also brings psychological comfort to fans. Idols create a perfect image that fans love and transmit positive energy to them. Some fans take idols as examples and motivate themselves to become better people.

Introduction Telework has been a beacon of hope for gender equality policy since it should potentially improve women's work-life balance [1]. However, in the late twentieth century, scholars were already issuing warnings about home-based teleworking representing a risk factor since workplaces remain inequal and gender regimens remained biased [2][3][4][5]. Since then, distance work has become progressively more widespread as the digitalization of the economy continues. The "domestication" of ICT [6] and globalization have substantially transformed work toward a kind of "work arrangement where workers work remotely, away from an employer's premises or fixed location, using digital technologies such as networks, laptops, mobile phones and the internet" [7], . Telework has expanded to embrace diverse work types, which has deeply transformed labor conditions as well as workers' lifestyles [8,9]. Technological and social transformations do not seem enough to erase gender bias [10]. Until today, men and women have had entrenched unequal conditions, with gender-based roles and tasks present both in the domestic and public domains, as described below in this paper. Androcentrism underlies female discrimination in the labor market and domestic domain, adding an ever-greater accumulation of social roles to accomplish, and public policy focused on erasing differences . The emergence of the concept of a work-life balance was a foregone conclusion, addressing women who work double shifts but, over time, has benefited both men and women, with the aim of equitable distribution of tasks and enjoyment of private time. The gender gap is still present in work culture [11]. There is evidence such as the gender pay gap, the glass ceiling in high positions at organizations, segregation of jobs by gender, and harassment. The discussion is still open about whether women are career-centric, arguing that they are less motivated to pursue professional goals, pushed or pulled by family [12,13]. The incorporation of women into the labor market is difficult and irregular, depending on their cultures, countries, and lifestyles [14]. Northern and Southern European countries are examples of different gender regimes and distribution of teleworking and values referring to work and lifestyles. Whereas Northern countries were historically leaders in advanced gender policy, teleworking is more widespread and time organization is tight, making it difficult to respect employee-friendly schedules and personal issues, Southern countries have worse labor conditions, a higher degree of presenteeism culture at workplaces, and deeper gender-role prescriptions. In addition, the work ethic of neoliberalism has been deeply ingrained in workers' lives, constantly pursuing achieving their goals and accomplishments at increasingly faster speeds and incessant multi-tasking. Social acceleration [13] affects workers' life satisfaction, particularly women workers , juggling multiple social roles and time regime constrictions [15]. Additionally, the COVID-19 pandemic has strengthened digitalization as the result of social distancing and lockdowns [16]. For the first time, men and women were at home undertaking work and family duties at the same time. However, who was responsible for childcare, household chores, and nursing, traditional roles of women? Did men take the opportunity to assume care duties? If men did indeed take on more roles, did men feel stressed and worried about the "broken social norms"? Conversely, did women feel hopeful that there might finally be changes in these social norms? Did men or women feel their well-being deteriorated? We asked these questions to learn about the certain transformation of gender-role balance and whether there was an unraveling of social roles, at least under the special crisis circumstances, and if so, what we should learn from a gender perspective. --- Are Changes Possible? Although social changes involve all men and women, most gender studies center on women as workers or how they are treated by institutions on the path towards equality, emphasizing the role of women in this social process. When scholars compare gender regimes and what kinds of transitions from domestic to public realms are transforming gender relations, what they are looking for is how switching roles affects women [17,18]. They are the social actors who are doing and undoing social changes, and they are also the vulnerable group in need of the public policy. This framework is rooted in the evidence that social conditions are primarily what push women into male circumstances. Over time, women are tackling spaces that were traditionally male domains, while men remain standing in the same place. All risks are taken by women, who move from one social role to another, juggling domestic and employment tasks, accomplishing demands with accelerated timelines, and increasing demands on well-being. Time distribution polls still report women spending more hours on childcare and housework, while men keep privileges as breadwinners and work-centric public domain actors [19,20]. No social forces are pushing men into the domestic realm, none so strongly as women are pushed by employment and social dynamics. Teleworking has different connotations for men and women as a result of traditional gender roles [8]. Whereas male teleworkers are highly mobile employees and self-employed, more women than men are teleworkers at home with poor conditions and low-ranking jobs in the company. In the same study, findings display quite similar satisfaction with work-life balance with telework, although, in the group of highly mobile workers and occasional teleworkers, women are more likely to report that working hours fit well or very well with their lifestyle than men. This suggests that men hold more critical opinions than women of their lifestyles and that women are more satisfied with their work schedules when teleworking. Although patriarchy characterizes all countries, the cultural differences across regions play a role in gender equality, work-life balance, and well-being [21][22][23]. Generally, Northern European countries present the lowest levels of work-life conflict because of gender and social inequality and the regulation of labor relations. Meanwhile, Southern countries reported more domestic division and worklife conflicts. The work-life conflict would be associated with poor self-reported health for men and women, although there are diverse interfering factors, such as the tradition of public childcare support for families, women's part-time regimes, dual-income couples, willingness for professional development, public policy and benefits, and so forth. Thus, cultural differences affect the self-reported well-being of men and women regarding the work-life conflict. Mensah and Adjei [23] found the smallest association between poor self-reported health indicators for both men and women in Nordic and Southern welfare states. Conversely, Liberal countries , Conservative countries , and Central Eastern concentrate the largest association between poor work-life balance and poor self-reported health. They explain these findings as a combination of factors related to a strong male breadwinner tradition, extension of social policy and provision of family-care benefits, employment regulations, and union protection. Lockdowns threw off social transformation for everyone, although the main change would be on the men's side, who traditionally have rarely settled in the domestic arena, and the pandemic led to them spending more time at home. Has this new scenario brought changes in gender relations? What changes have affected men locked down at home and exposed daily to domestic issues? Did they do chores and take care of children? What roles and tasks have replaced the time they once spent on work relationships and public affairs? Rubery and Tavora collected evidence from cross-sectional surveys during the pandemic that suggests a shrinking gap between mothers and fathers in childcare hours spent [24]. However, the authors are cautious with these findings and warned that work-life balance was undertaken differently between the health crisis and "normal" times. Conversely, data show information about women feeling resistant to assuming caretaker roles during the pandemic as professional and supported agents of family care and household well-being [16,25]. Differences in cultural and political traditions also influenced policy responses during COVID-19 and the populations' risk perception [24]. In that sense, have men and women from European regions revealed different roles and reactions with regards to work-life balance as the result of their exposure to care roles and working at home? Have they perceived the risk of well-being in the same way, or did they experience different worries? What kind of risk determinants felt related to juggling family and work chores? The question was posed about what men and women have felt about teleworking and handling family and jobs in the same place during the pandemic, and more importantly, what mattered to men? What did they do, and how did they feel spending more time at home exposed to care roles in domestic settings? Eurofound [25] has reported low female resilience more often than male respondents as a result of their vulnerable situation during the pandemic . Results were formulated on the basis of two questions in the survey rated on five levels of satisfaction: "I find it difficult to deal with important problems that come up in my life" and "When things go wrong in my life, it generally takes me a long time to get back to normal." This relevant insight is related to toughness to withstand adverse conditions but disregards what they felt when facing daily work-life balance experiences by gender, the focus of this work. --- Materials and Methods From April to July 2020, Eurofound launched a two-round online survey entitled "Living, Working and COVID-19," consisting of 11,575 total panel questionnaires across EU-27 countries. The crisis contextualization may also have affected the number and motivation of respondents at home to endorse the surveys' claims. Responses were gathered by uncontrolled convenience sampling through public dissemination among Eurofound contacts and stakeholders, social media advertising, and targeting hard-to-reach groups [26]. Therefore, the sample may have a self-selection recruitment bias. The dataset includes two rounds, where the fieldwork for Round 1 took place between 9 April and 11 June 2020 and Round 2 between 22 June and 27 July 2020 [26]. The analysis of this work aims to compare how men and women felt during the COVID-19 lockdown in various contexts depending on their gender regimes and worklife balance experiences. We were particularly interested in learning more about male teleworkers because we understand they were in a brand-new setting that may have led to their doing more in domestic realms. If they did, we want to know what risks they perceived to their lifestyle and well-being. We are aware that the population has preliminary cultural differences related to teleworking and gender regime-even more so in this period-with mortality rates and health-risk perceptions that may have swayed their thoughts and self-reported feelings during this period. In summary, we first needed to determine what combination of demographic factors best classified all the different individuals . Table 1 below details the characteristics of the study sample for the purpose of this analysis. A classification tree analysis was conducted to learn which factors are based on the categorical variable for life satisfaction. This technique is intended to divide the populations or samples under study, following a descending sequential process, for the identification of homogeneous subgroups in a variable of interest to the researcher , taking into account the characteristics of the subjects studied in a selection of the independent variables, predictors, prognosticators or most important explanations for them. This data mining technique aims to identify homogeneous subgroups of the population with respect to certain characteristics and which independent variables show the greatest differences with regard to life satisfaction scores. A dendrogram was obtained from the socio-demographic variables on the population by the greatest weight in the responses for the constructed indicators: gender, work at home, and European countries group. Population subgroups were formed by the importance of the relationships between variables, and a further characteristic was added according to its degree of importance based on statistically significant differences. The process classifies individuals in several groups according to the most important factors determined by responses with significant differences between variables, differentiating which factors are different among homogenous subgroups of the population. To learn more about the feelings of men and women during this period of time, the analysis addresses the distance between happiness and life satisfaction, an indicator proposed by Ruut Veenhoven [27,28] to develop a comprehensive indicator of the quality of life between nations. To analyze this distance, an ANOVA has been used, which is one of the most widely used statistical techniques to compare groups of measures, which is normally used to establish similarities and differences between three or more different groups. Through ANOVA, an analysis is established to comparatively evaluate results in different classifications or groups. In this way, it is possible to calculate if the mean values are the same in the different groups studied. Happiness is considered a subjective enjoyment of life, whereas life satisfaction is related to how much the person likes their life. As both variables correlate [28], subtracting happiness from life satisfaction, the measure would indicate a subjectivation of respondents' quality of life. --- Description of Variables Northern, Southern, Eastern, and Western European countries follow the United Nations' classification Eastern Europe: Belarus, Bulgaria, Czechia, Hungary, Poland, Republic of Moldova, Romania, Russian Federation, Slovakia, Ukraine. Northern Europe: Denmark, Estonia, Finland, Iceland, Ireland, Latvia, Lithuania, Norway, Sweden, the UK of Great Britain and Northern Ireland. Southern Europe: Albania, Bosnia and Herzegovina, Croatia, Greece, Italy, Malta, Montenegro, North Macedonia, Portugal, San Marino, Serbia, Slovenia, Spain. Western Europe: Austria, Belgium, France, Germany, Liechtenstein, Luxemburg, Monaco, Netherlands, and Switzerland). Well-being is constructed through two questions included in the survey about "Life satisfaction" and "Happiness," rated on a 1-10 scale to assess "How satisfied are you with your life these days?" and "How happy would you say you are?" The activity type they engaged in is made up of two variables: "Started working from home as a result of the COVID-19 situation" and "Employment status" , creating a six-range variable. Finally, the "Feelings about work-life balance" is a five-category variable through which respondents scored four items asked by the survey: "Kept worrying about work when you were not working", "Felt too tired after work to do some of the household jobs which needed to be done", "Found that your job prevented you from giving the time you wanted to your family", "Found it difficult to concentrate on your job because of your family responsibilities", and "Found that your family responsibilities prevented you from giving the time you should to your job". --- Results Findings are presented in three sections. The first section aims to capture workers' variation in responses by gender due to the lockdown situation in the COVID-19 pandemic, according to self-reported life satisfaction. The second section addresses men's and women's well-being by geographical areas, which may be related to the gender-regime tradition, public policy, social benefits, and labor market characteristics. The third section attempts to discover the well-being and worries of men and women from spending more time than usual at home. --- Life Satisfaction and Gender of Participants, General Overview The analysis finds an association between the life satisfaction of workers and the gender of respondents , where men have higher life satisfaction than women. They appear to be affected differently by the crisis and the situations both of working at home and taking care of family. While women's well-being is associated with work type , while men's well-being is related to geographic region. The classification tree analysis reveals that both men and women teleworking have higher satisfaction levels than those who remained in the workplace, making teleworking seem like an optimal solution to uphold performance in times of social acceleration. However, female responses suggest worries about their daily experience as front-line workers, while men's responses seem to reflect a perception of health risk. Northern countries show the lowest mortality rates, and male life satisfaction reaches 7.25, while higher excess mortality rates correspond to Southern countries, where men report the lowest life satisfaction, 6.21 ). Results point to risk perception being the baseline for life satisfaction, where men are more likely to value staying safe at home, and women who are not teleworkers rate the lowest life satisfaction . Women's worries seem to hinge on work, while the health situation appears more concerning for male respondents. This result suggests that segregation of the labor market matters to women, who expressed concerns about the possibility that working remotely would hinder their job performance. --- Life Satisfaction and Happiness by Gender Regime Broken down by European Regions By regions, workers from Southern countries who do not telework show minimum life satisfaction , while male teleworkers from Northern countries and female teleworkers display the highest levels of life satisfaction . The gender gap is clear according to these results . Men show higher selfreported life satisfaction than women in almost all regions, except Southern countries, where women show a slightly higher perception of life satisfaction than men. Southern European countries present the lowest life satisfaction ratio, while Northern and Western countries have the highest. To learn more about workers' experiences of spending more time at home than usual, we compared the life satisfaction of male and female teleworkers during the pandemic. In line with previous findings, we expected differences at least in Southern countries, leading us to search for likely variation by employment status . To learn more about workers' experiences of spending more time at home than usual, we compared the life satisfaction of male and female teleworkers during the pandemic. In line with previous findings, we expected differences at least in Southern countries, leading us to search for likely variation by employment status . The analysis finds higher rates of male than female satisfaction in Eastern and Western countries. Data may suggest that men found working at home satisfactory, with scores around 7. Gender differences are revealed for self-employed without employees in Northern countries and employees and freelancers in Southern countries, where women report higher life satisfaction than men. --- Spending Time at Home, Well-Being, and Worries In July, the second round of the survey included relevant questions for our aims on caring and household work during the last month. Results reported by Eurofound [25] reveal persistent gender differences, where women were generally more involved in spending time caring for family and doing household work . Therefore, considering the whole population, it concludes that lockdown did not entail a significant transformation of gender relations or transformation of traditional gender-role distribution at home. Still, over one-third of respondents worked fewer hours during the pandemic, where the decreased hours were more usually for men than women , which significantly changes the scenario for male workers. It raises a question about the daily home The analysis finds higher rates of male than female satisfaction in Eastern and Western countries. Data may suggest that men found working at home satisfactory, with scores around 7. Gender differences are revealed for self-employed without employees in Northern countries and employees and freelancers in Southern countries, where women report higher life satisfaction than men. --- Spending Time at Home, Well-Being, and Worries In July, the second round of the survey included relevant questions for our aims on caring and household work during the last month. Results reported by Eurofound [25] reveal persistent gender differences, where women were generally more involved in spending time caring for family and doing household work . Therefore, considering the whole population, it concludes that lockdown did not entail a significant transformation of gender relations or transformation of traditional gender-role distribution at home. Still, over one-third of respondents worked fewer hours during the pandemic, where the decreased hours were more usually for men than women , which significantly changes the scenario for male workers. It raises a question about the daily home lives of male teleworkers: How did this "spare time"-since they worked fewer hours and did not transform their roles at home-affect their well-being? We wondered if it would interfere with men's self-reported worries about life-work balance, at least, because they were exposed to spending more time in family care and housekeeping. If they did take on care responsibilities to a lesser degree than women, did men perceive this time as more relaxed than women who must juggle the work-life balance more? We look at the distance between happiness and life satisfaction to estimate what the subjectivation is of their quality of life during this time. As Figure 2 displays, they reported that they "always" had difficulties concentrating on their jobs because of family and "never" felt too tired after work to do some of the household jobs that needed to be done . According to the first indicator, they probably felt overwhelmed by the family around them while working, and according to the second, they did not do so much housekeeping that they felt tired. Jointly, this points to them keeping their routines for work and domestic duties unchanged. lives of male teleworkers: How did this "spare time"-since they worked fewer hours and did not transform their roles at home-affect their well-being? We wondered if it would interfere with men's self-reported worries about life-work balance, at least, because they were exposed to spending more time in family care and housekeeping. If they did take on care responsibilities to a lesser degree than women, did men perceive this time as more relaxed than women who must juggle the work-life balance more? We look at the distance between happiness and life satisfaction to estimate what the subjectivation is of their quality of life during this time. As Figure 2 displays, they reported that they "always" had difficulties concentrating on their jobs because of family and "never" felt too tired after work to do some of the household jobs that needed to be done . According to the first indicator, they probably felt overwhelmed by the family around them while working, and according to the second, they did not do so much housekeeping that they felt tired. Jointly, this points to them keeping their routines for work and domestic duties unchanged. Data suggest that women continue with all gender roles despite lockdown, coping with work-life balance in the new situation. Delving into the impact of gender on wellbeing, we compare men's and women's self-reported life satisfaction. According to Table 3 below, women felt too tired "sometimes" and "most of the time" more than men, but women self-reported higher proportional life satisfaction than men. Findings disclose a similar trend for the following items. Women found it more difficult to concentrate on their jobs because of family in comparison to men , but women reported proportionally more life satisfaction than men. Men and women felt worries about work, even though men are less centered on household and care responsibilities, and their life satisfaction ratios were close, and sometimes even lower, than that of women . This suggests that women exhibit more adaptable and resilient attitudes, while men who made very few changes to their habits involving work-life balance felt life satisfaction as low as women. Data suggest that women continue with all gender roles despite lockdown, coping with work-life balance in the new situation. Delving into the impact of gender on wellbeing, we compare men's and women's self-reported life satisfaction. According to Table 3 below, women felt too tired "sometimes" and "most of the time" more than men, but women self-reported higher proportional life satisfaction than men. Findings disclose a similar trend for the following items. Women found it more difficult to concentrate on their jobs because of family in comparison to men , but women reported proportionally more life satisfaction than men. Men and women felt worries about work, even though men are less centered on household and care responsibilities, and their life satisfaction ratios were close, and sometimes even lower, than that of women . This suggests that women exhibit more adaptable and resilient attitudes, while men who made very few changes to their habits involving work-life balance felt life satisfaction as low as women. --- Discussion According to data on teleworking during the pandemic, this work mode is not related to advances in greater gender equality [1,2]. Remote work is insufficient if the labor market and gender regimes remain biased [10]. If preceding findings highlighted that teleworking had a limited impact on women advancing in the labor market and the public domain, these results also suggest it is a poor mechanism for advancing toward a more equal gender regime in the domestic arena. Some women may find advantages in teleworking because of time management and the self-organization of tasks corresponding to work-life balance in the household [5,15]. On the contrary, men spending more time at home involves almost no impact on a more equal gender regime with equal gender-role distribution. If women have taken the risk of tackling work-life balance as they advance in joining the labor market, men do not have the risk of taking more responsibility for care and household work when circumstances place them at home. In the career-centric debate [12], women appear to be focused on careers without neglecting family care responsibilities. On the contrary, men seem to keep traditional gender privileges as breadwinners, even if the pandemic may have caused a social change, keeping them at home. Although the pandemic could have pushed men toward making social changes, they did not change. A possible shrinking gender gap points to there not being enough time caring for the family and household to modify the gender regime [19]. In that sense, the gender regime remains unbroken. Since men do not take the risk of tackling the lion's share of the work-life balance, they self-reported a higher proportion of life satisfaction than women. While women's concerns are related to continuing presence at the workplace or, conversely, having the opportunity to telework, men's self-reported satisfaction depends on their geographic region . Southern countries are both seriously affected by the pandemic and by more domestic division of work, reporting poor life satisfaction [25]. As Table 2 shows, Southern countries reveal the lowest life satisfaction, whereas Northern countries and Western countries have the highest rates. However, a general overview of the data discloses that keeping traditional roles entails serious risk for women; men from Southern countries display the lowest life satisfaction , which may point to pandemic worries and socio-economic factors, jointly with gender issues. Presenteeism and teleworking previous to the pandemic culture in different regions might also have raised the dissatisfaction of men in Southern countries. Accordingly, women appear more satisfied with teleworking when they have the opportunity with this work arrangement. The employment status of women is also relevant to learning more about differences between European regions. Women freelancers who started to work from home as a result of the pandemic in Northern and Southern European countries display higher satisfaction compared to male freelancers [8]. The pandemic would seem to have given them the opportunity to do remote work and be accepted by the labor market while taking advantage of organizing work-life time. Juggling domestic and labor responsibilities during the lockdown, women, in due course, felt tired and worried about work when they were not working and had difficulties concentrating because of family needs [15,18,19]. Women felt their well-being deteriorated, but surprisingly, men also indicated poor well-being, at least stronger in proportion than women who take on more care responsibilities. Although Eurofound [25] pointed out women's low resilience, we think that these findings display toughness to withstand adverse conditions working at home, more than men. Female low resilience would relate to socio-economic issues, but women would lead with a better work-life balance at home than men. With regard to men, the distance between happiness and life satisfaction suggests few endeavors to transform gender relations and a proportional feeling of quality of life. The quality-of-life assessment is key for undertaking minimal social changes. It suggests that it is very unlikely that gender transformation can be forced on men in the domestic arena. However, we frankly think that disseminating results-such as this work-may illuminate and raise men's awareness of the need to transform inequality at home to improve family life beyond the private habitus. Daily routines are rooted in the androcentric status quo, keeping the same roles, or with minimal changes, that reinforce men's roles having a privileged status. Social and technological changes such as public policy and normative measures are putting cracks into gender regimes [15,20], and now may be the time to show men how unequal the situation is and trigger their involvement in the household. Further research is needed to discover the determinants of well-being for male and female workers regarding work-life balance. Health status and poor labor conditions are also key variables to determine different responses. Teleworking and social acceleration are components of neoliberalism, and workers are increasingly embroiled in subtle and strong cultural variables, making us need to learn more about men's and women's awareness and resistances, depending on their lifestyles and socio-economic contexts. The heterogeneity of workers depending on labor activity, working hours, type of family, and household, among others, blurred our evaluation of the influence of telework on population groups. We need a more in-depth investigation into factors that separate groups of workers, including an eco-social framework of analysis that would contribute to explaining differences. --- Conclusions This research addresses the impact of male and female workers' well-being during the COVID-19 lockdown, where spending time at home is relevant to determine any social changes to the gender regime. Men appear to have greater life satisfaction than women. The source of worries for men is keeping safe at home, dependent on the general rates of health risk in their countries, while women's concerns depend on the availability of teleworking. Women workers present a work-centric orientation, although further analysis also proves they are in charge of family care and housekeeping. By regions, men self-reported high life satisfaction in almost all areas. Surprisingly, in Southern Europe, where all studies underline strong domestic division and work-life conflict, women's perception of life satisfaction during the lockdown is slightly higher than that of men. However, Southern men and women together display the lowest life satisfaction in Europe. Interestingly, women self-employed without employees in Northern and Southern Europe report a higher proportion of life satisfaction than men. Although evidence shows that men made few social changes regarding spending more time at home doing care and housekeeping, we want to know more about men's experiences at home. Findings report high difficulties concentrating on their jobs because of family and, correspondingly, men's minor involvement in work-life balance, but they "never" felt too tired after work to do housekeeping. Women self-reported that they were "too tired," had difficulties concentrating on their jobs, and worried about work when they were not working, but compared to men, they felt life satisfaction to a higher degree than would be expected. These findings point to men's high resistance to taking risks of spending more time doing care and housekeeping even during the pandemic and spending more time at home. Expected changes in gender regime seem remote, even though teleworking is widespread and equally affects men and women. We suggest twofold measures: providing men with incentives and messages about gender inequities at home; and social and economic support for women as the main players in work-life balance, even in special circumstances. --- Data Availability Statement: Restrictions apply to the availability of these data. Data was obtained from Eurofound with the permission of Eurofound through a data analysis request. ---

Lockdown during COVID-19 forced the emergence of a new scenario, with men and women teleworkers spending all their time at home. The purpose of this study is to address whether this situation has triggered a transformation in gender roles and self-reported well-being, comparing the responses of male and female respondents to the EUROFOUND April to July 2020 survey. The analysis addresses cultural differences across European regions related to diverse gender regimes, employment status, and the possibility of teleworking. It explores male and female well-being through life satisfaction, the distance between happiness and life satisfaction, and rates their feelings about work-life balance. Findings on life satisfaction display a low transformation of social roles, with women still worrying about work-life balance, while men were more affected by the health crisis. Men self-report high life satisfaction across Europe compared to women, although unexpectedly, female freelancers in Northern and Southern European had a higher life satisfaction ratio than men. Both men and women teleworkers reported difficulties with managing work-life balance at home, despite women handling core care and household tasks. These findings suggest that women would have received more support from men, as they worked harder and longer hours during the lockdown, despite their weak position in the labor market. This would seem to be a propitious setting for men to have assumed more responsibility at home, resulting in a more equal distribution of roles at home.

What did the researchers do and find? • We obtained information on socioeconomic factors throughout the life course of Norwegian adults and categorised them into low, middle, or high position. • We found that moderately frequent alcohol consumers had a lower risk of dying from cardiovascular disease than infrequent consumers, and that this was more pronounced among those with high position. Very frequent consumption was associated with increased risk of CVD mortality, but only among those with low socioeconomic position. • We also found that weekly binge drinkers had higher risk of dying from cardiovascular disease than current drinkers who did not binge drink the past year, but we could not elucidate whether the risk differed by life course socioeconomic position. --- What do these findings mean? • The study observed socioeconomic differences in risk estimates of CVD mortality associated with given alcohol consumption levels. It is unclear if this reflects differential --- Introduction Socioeconomic position is relevant to behaviours, exposures, and susceptibilities that may influence health [1], such as social support, financial resources, or the knowledge, awareness, and determination required to actively follow a healthy lifestyle or consult a physician if needed. There is an inverse socioeconomic gradient in the exposure to risk factors for cardiovascular diseases [2], which translates into a gradient in the risk of clinical CVD events [3,4]. The majority of heart attacks and strokes occur in late adulthood, but atherosclerosis development starts in childhood [5]. Socioeconomic disadvantage at different stages throughout the life course could therefore be relevant to risk factor exposure, atherosclerosis development, and the long-term risk of clinical cardiovascular events [6][7][8][9][10]. In contrast to tobacco smoking, which is more frequent among socioeconomically disadvantaged individuals and has a clear detrimental effect on health, the relationship between SEP, alcohol, and health is less clear. Disadvantaged groups tend to report less frequent alcohol consumption but experience more harm from a given level of alcohol exposure [11][12][13][14]. This is sometimes referred to as the alcohol harm paradox [15]. In terms of CVDs, associations between alcohol drinking patterns and CVD risk further complicate the situation. A drinking pattern characterised by more frequent consumption of low to moderate volumes is associated with a reduced risk in comparison to infrequent drinking or abstainers, while episodic heavy drinking, also called binge drinking, is associated with an increased risk [16]. Alcohol has multiple biological effects on the cardiovascular system, both harmful and potentially protective [17][18][19][20], and it has been suggested that differing dose-response relationships of these mechanisms may explain the overall J-shaped risk curve. Biological effects of alcohol should not differ by SEP, but the noncausal associations could do so if the lifestyles that accompany a drinking pattern differ according to SEP [21]. When consuming alcohol, for example, disadvantaged individuals may more frequently co-consume junk food or smoke cigarettes, while advantaged individuals may be more prone to combine drinking with advantageous health-related behaviours and characteristics [21]. These potential differences may be profound and captured by the measurement of important risk factors but may also be subtle and difficult to measure individually. The assessment of SEP throughout the life course could be an approach that encapsulates the effect of these potentially subtle differences over time. In this study, we investigated whether the diverging relationships between alcohol drinking patterns and CVD mortality differed by life course SEP. --- Methods --- Study population The Counties Studies [22], the Cohort of Norway [23], and the Age 40 Program [24] are three partly overlapping cohorts containing data from Norwegian population-based health surveys . Participants were recruited to the surveys through their personal identification --- --- Alcohol exposure The assessment of alcohol exposure differed between the source surveys, and we harmonised the data for use in the current study . Data identifying current and lifetime abstainers were harmonised into current abstainers for the main statistical analyses. Among current drinkers, alcohol consumption frequency was categorised into 'Infrequent', 'Once per month to once per week', '2-3 times per week', and '4-7 times per week'. In surveys in which beer, wine, and liquor consumption were assessed separately, we first recoded the reported ordinal frequency categories into days of alcohol consumption per month, then summed the days to reflect total alcohol consumption, and finally recoded the sum back into the ordinal categories for harmonisation. This approach assumes that each beverage type was consumed on different days of the month. Participants reporting to be an abstainer on one question and reporting drinking on another question were defined as drinkers. We defined a standard unit as 12.8 grams of pure alcohol, corresponding to a small bottle of beer , a glass of wine , or a small glass or shot of liquor . The frequency of heavy drinking episodes , which reflects the intake of high volumes, was categorised into 'Not last year', 'A few times', '1-3 times per month', and '!1 time per week'. The average amount of alcohol could be assessed and harmonised for a subsample. Three calculations were applied, depending on which questions were available in each survey. Two calculations combined the average number of units consumed per occasion with the drinking frequency reported either per month or in ordinal categories . The third calculation was based on the total number of units consumed of beer, wine, and liquor in the course of 2 weeks. --- Life course SEP CVDs tend to develop throughout the life course and manifest clinically in late adulthood. The manner in which risk factors and protective factors influence disease development may not be in unison; for example, there could be critical or sensitive time periods. A life course approach to epidemiology is one that takes this notion of time into account by acknowledging that measuring risk factors only once could be inadequate in order to assess the full impact they may have through the life course [25,26]. Previous studies have observed that CVD mortality is related to the number of occasions individuals have been exposed to socioeconomic disadvantage, measured by adding multiple indicators from different periods in the life course together in a cumulative manner [6,27]. We obtained a cumulative measure of life course SEP by combining indicators on household conditions from mandatory population and household censuses in 1960, 1970, and 1980 , household income from the census in 1990, and the highest level of obtained education ever recorded until 2011 . In contrast to the 1960 and 1970 censuses, which obtained almost complete population and household data, the census in 1980 did not pursue missing household data to the same extent. The household indicators have previously been observed to be independently associated with cause-specific mortality, as well as when combined into cumulative indexes [27,28]. A more detailed description of the role of the use of household indicators may be found here [1,25]. The household conditions, household income, and education provided a total of 20 indicators, which were scored and given equal weight by summing the scores to construct the cumulative index . A high score indicated disadvantage and low life course SEP. --- Covariates and outcome The health surveys provided self-reported data on current smoking, physical activity, diabetes, previous CVD , family history of coronary heart disease , objective measurements of blood pressure and heart rate, anthropometry, and biochemical nonfasting measurements of serum triglycerides, total cholesterol, and high-density lipoprotein cholesterol . The Norwegian Cause of Death Registry provided outcome data on causes of death using the ninth and tenth revision of the International Classification of Diseases . The primary outcome was CVD mortality . Three secondary outcomes were added in response to peer review, including death from ischemic heart disease , death from cerebrovascular diseases , and all-cause mortality. The registry is almost exclusively based on certificates filled out by on-site medical doctors, and in the few cases in which autopsies are performed, 32% of deaths are reclassified over major ICD-10 chapters [29]. --- Statistical analysis We described the study population according to categories of life course SEP as well as according to alcohol consumption frequency within categories of SEP. Continuous variables were presented as mean and categorical variables as counts . Analysis of variance and chi-squared tests assessed differences between the groups. In survival analyses, we followed participants prospectively until emigration , death from any cause, or December 31, 2014. Cox Proportional Hazard Models estimated hazard ratios and confidence intervals . Visual inspection of scaled Schoenfeld residuals against time did not indicate strong deviation from the assumption of proportional hazard. All analyses were conducted in R statistical software using R studio 1.0.44 [30], with additional use of the packages survival [31] and mice [32]. We did not impose a p-value cutoff to define statistical significance [33] nor apply survey weights. To evaluate whether the SEP index was relevant to the outcome, we estimated the risk of CVD mortality in a model with the SEP index, age, and gender. Potential mediators, such as smoking, were not included, in order to assess the total effect. The index was first modelled using a smoothed penalised spline, allowing for a visual presentation of the functional relationship with the outcome, and then as a continuous and categorical variable for a formal assessment. The aim was to assess if the relation between alcohol drinking patterns and the risk of CVD mortality differed by life course SEP. The hypothesis we tested statistically was whether SEP modified the effect of alcohol drinking patterns on the risk of CVD. We present HRs with 95% CI for alcohol consumption frequency and for the frequency of binge drinking episodes overall, in strata of SEP, and measures of effect modification on a multiplicative scale as HRs with 95% CI and p-values. Both exposures were modelled as ordinal categorical variables, with infrequent consumers and those who did not binge drink the last year as reference categories, respectively. Current abstainers were modelled separately as a dichotomous variable, with infrequent consumers as the reference category. Confounders of the relation between alcohol and CVD that were adjusted for included age, gender, current smoking, physical activity, body mass index , systolic blood pressure, heart rate, triglycerides, diabetes, history of CVD, and family history of CHD. In the subgroup with data on binge drinking, we adjusted the risk of CVD mortality according to alcohol consumption frequency for episodes of heavy drinking, and vice versa. Analyses were performed separately for total CVD, IHD, stroke, and all-cause mortality. Missing values were handled by list-wise deletion and totalled to 18.4%. We also performed missing value imputations of CVD risk factors and census data by chained equations among 245,336 eligible individuals with data on alcohol consumption . This reduced the amount of missing values to 3.5%. Alcohol exposure variables, CVD risk factors, census data, outcome data, and the SEP index were included in the imputation model and 10 data sets were generated. We then reanalysed the relationships with total CVD in each data set and report pooled HRs with 95% CIs. We performed 2 sensitivity analyses in response to peer review. In the subgroup with data on binge drinking, we reanalysed the relation between alcohol consumption frequency and the risk of CVD while excluding binge drinkers . In the subgroup with data on lifetime abstaining, we compared the risk of CVD when using lifetime abstainers and infrequent consumers as reference categories. We added analyses while performing planned statistical analyses, which were elaborated during peer review. Short-term experimental studies show a dose-response relationship between alcohol intake and levels of HDL-C [17]. Using ordinary least squares regression and models adjusted for age and sex, we regressed HDL-C on a continuous variable of drinking frequency . In a subsample, we also regressed HDL-C on a continuous variable of the average amount of alcohol consumed per day . Changes in HDL-C were compared to the dose-response relationship in a meta-analysis of experimental studies [17] to indicate if the main study variable was consistent with an increase in total alcohol consumption as judged by HDL-C and to indicate if the self-reported data were underreported. We also reanalysed the relationship between HDL-C and drinking frequency in strata of SEP to indicate if SEP could influence the ability to report consistently [34]. A formal test for a difference in slope was performed by including an interaction term between drinking frequency and SEP. --- Results --- --- Descriptive analyses Baseline characteristics differed according to life course SEP for all included variables . Participants with low SEP were on average older, more often female, had a higher prevalence of CVD risk factors, more previous diseases, and were more often a current abstainer or an infrequent consumer of alcohol. Participants with high SEP had the lowest prevalence of CVD risk factors, were more often frequent consumers of alcohol, and were more likely to binge drink within the subgroup for which these data were available. Estimates for middle SEP participants were mostly between the other strata. The distribution of covariates over categories of alcohol consumption frequency followed comparable patterns within the strata of life course SEP, but with different magnitudes . Notably, frequent consumers of alcohol were consistently more often also frequent bingers, but the percentage among the most frequent consumers who were also weekly bingers was 32.8% in the low, 19.1% in the middle, and 16.9% in the high SEP strata. --- Follow-up time and mortality The mean follow-up time in the study population was 16.6 years. In total, 25,950 participants died-8,435 from CVDs, including 3,837 from IHD and 1,972 from stroke. In the subgroup of current drinkers with additional data on heavy drinking episodes, 7,274 died during 15.4 years of follow-up, 2,284 from CVDs, including 1,028 from IHD and 553 from stroke. In the subgroup with additional data on lifetime abstaining, the number of CVD deaths in the course of 12.5 years was 2,166 . --- Life course SEP and mortality --- Alcohol consumption frequency, life course SEP, and mortality The risk of CVD mortality was lower among frequent drinkers than among infrequent consumers in imputed and complete case analyses , with even lower estimates and stronger associations when excluding or adjusting for binge drinking . There was no difference in risk between infrequent consumers and lifetime abstainers when they were used as reference categories in a subgroup was associated with a higher risk of CVD mortality in imputed and complete case analysis as well as a higher risk of IHD , and all-cause mortality (S14 --- Additional analyses The crude distributions of HDL-C according to categories of alcohol consumption frequency and life course SEP are presented . When adjusted for age and sex, the increase in HDL-C per increase in drinking frequency was 0.093 mmol/l and corresponded to an estimated mean increase of approximately 26.6 g ethanol/day when we compared it to the estimated dose-response relationship between alcohol and HDL-C in a meta-analysis of experimental studies [17]. The increases in the high, middle, and low strata of SEP were 0.095 , 0.090 , and 0.086 mmol/l, respectively . In a subsample, the change in HDL-C per increase in the amount of alcohol consumed per day was 0.009 mmol/l. This corresponded to a 0.113 mmol/l increase per unit of alcohol, which is in the upper range when compared to experimental studies in which 1-2 drinks/day corresponded to an estimated increase in HDL-C of 0.072 mmol/l [17]. --- Discussion --- Principal findings Among adult participants in Norwegian health surveys , alcohol consumption and episodic heavy drinking were more frequent among individuals with high SEP throughout their life course. Participants with low SEP were more likely to currently abstain or drink less frequently, but apart from that, they were more exposed to all other CVD risk factors. Moderately frequent drinking was associated with a lower risk of CVD, IHD, and all-cause mortality than infrequent drinking, and we observed that this association was more pronounced among participants with high SEP. Very frequent drinking was associated with a higher risk of CVD and all-cause mortality compared with infrequent drinking, but only among participants with low SEP. Frequent binge drinking among current drinkers was associated with a higher risk of CVD, IHD, and all-cause mortality compared with no binge drinking during the last year, but it was not possible to determine whether the risk differed by life course SEP. Because of few events, it was also difficult to make firm inferences regarding stroke mortality. --- Interpretation of findings The higher prevalence of current abstainers and infrequent consumers among those with low life course SEP is consistent with studies in other countries [35]. Alcohol taxes are particularly high in Norway, and differences in financial resources to purchase alcoholic beverages could contribute to this difference [36]. Interestingly, episodes of heavy drinking were somewhat more common among individuals with high SEP, illustrating the widespread acceptance of this behaviour in Norwegian society, even in the most health-conscious segment of the population. Another interesting observation was that individuals with low SEP were overrepresented in the most heavy drinking category when the frequencies of alcohol consumption and binge drinking were combined, a tendency that has also been observed previously in Europe [15]. The comparability to other populations in this regard strengthens external validity. We observed lower risk of CVD, IHD, stroke, and all-cause mortality among moderately frequent drinkers in the study population compared with infrequent drinkers, which is in agreement with the majority of similar studies [16,37]. However, evidence from Mendelian randomisation studies thus far do not support a protective effect of alcohol on CVD nor provide support for a causal effect of factors that were considered to mediate a protective effect of alcohol-in particular, HDL-C and fibrinogen-on CVD [19,[38][39][40][41][42]. It is therefore important to consider whether our findings may have been influenced by unmeasured confounders or misclassification of previous heavy drinkers [43]. In our study, we addressed the issue of reverse causality by choosing infrequent consumers over current abstainers as a reference category. We considered that small differences in alcohol consumption could not account for the difference in risk observed between these groups, unless moderate drinkers were misclassified as infrequent consumers because of underreporting [34,44]. Although we did not have data on lifetime abstainers for all participants, their risk of CVD did not differ from that of infrequent consumers in a subgroup analysis. We addressed the issue of confounding by adjusting for the uneven distribution of measured confounders, which did not strongly influence the associations. However, as measured confounders and thus, likely also unmeasured confounders, were distributed unevenly over categories of alcohol consumption within each stratum, there could clearly be residual confounding in the within-strata analyses as well. It is therefore unclear whether the findings reflect differential confounding of alcohol consumption with other exposures or differing effects of alcohol on health across socioeconomic groups. The stratified analyses and tests for effect modification suggested that the relationship between alcohol consumption frequency and the risk of CVD mortality differed by life course SEP. The association between moderately frequent consumption and lower risk of CVD was more pronounced for participants with high SEP than among participants with low and middle SEP. Very frequent drinkers in the middle and high strata of SEP had either lower or comparable risk of CVD, IHD, stroke, and all-cause mortality in comparison to infrequent consumers, while very frequent drinkers with low SEP had a higher risk of CVD and all-cause mortality. Alcohol is subjected to first-pass metabolism in the gastrointestinal system, and when alcohol is co-ingested with foods, metabolism by enzymes in the stomach is extended [45]. This reduces bioavailability of ingested alcohol overall and also delays and reduces peak blood alcohol concentration, which may attenuate the systemic toxic effects of alcohol [46]. If drinking were more often accompanied by meals in one stratum of SEP, such as those with high SEP, it could account for a lower risk among binge drinkers compared to those not binge drinking, but not when considering alcohol as a protective factor. Another possibility, as indicated previously, is that the differences in risk by life course SEP could have arisen or been influenced by confounders having different effects in each stratum, such as if alcohol consumption is accompanied by a different set of behaviours in each stratum. We observed higher CVD, IHD, and all-cause mortality among current drinkers who were frequent binge drinkers compared with current drinkers who did not binge drink during the last year, possibly mediated by increased blood pressure [19,47], cardiomyopathy, cardiac arrhythmias, and disturbances in blood electrolyte status [48,49]. A previous meta-analysis found binge drinking not to be associated with a higher risk of IHD in comparison to lifetime abstainers [50], and in that sense this finding sticks out. Findings were strong and consistent in the large middle stratum but less clear and less consistent in the smaller low and high SEP strata, which likely resulted in some inaccurate effect estimates and reduced precision when testing for effect modification. In light of the sample size and number of events, and the heterogeneity in the most extreme drinking categories, it is difficult to conclude with confidence that there is no socioeconomic difference in the relationship between binge drinking and CVD among adult Norwegians. --- Methodological considerations Multiple measurements of alcohol exposure over time is the best approach to account for variation in consumption [51], but this study was limited to a single self-reported measurement. Previous studies found a test-retest correlation for men and women of approximately 0.6 when using data on repeated measurements from the source population [52,53]. Furthermore, 52% of the men and 62% of the women reported consistently in a follow-up postal survey 10 years later, of which abstainers were the most consistent . Abstainers and heavy drinkers, however, appeared to be more prone to dropout than infrequent and light consumers. The large sample size also accounts for random variation. We used HDL-C as a biomarker of a change in the magnitude of total alcohol consumption, and after adjusting for differences in age and sex, we observed an increase in HDL-C of about 0.093 mmol/l for each categorical increase in alcohol consumption frequency. This value corresponds to an estimated mean increase of approximately 26.6 g ethanol/day if we compare it to the estimated dose-response relationship between alcohol and HDL-C from a meta-analysis of clinical trials [17] and substantiated that increasing consumption frequency accompanied increased amount of ethanol consumed. The increase in amount of HDL-C was comparable within all strata of SEP; thus, it seems unlikely that differential exposure misclassification can explain the differing relationships between alcohol consumption and CVD mortality in the different strata of life course SEP. The dose-response relationship of cardiovascular mortality with alcohol consumption seemed to nadir at a frequency of 2-3 times per week, or 40 g ethanol/day higher intake on average than infrequent consumers, which is comparable to overall estimates from previous studies when men and women are combined [37]. The increase in HDL-C per increase in the amount of alcohol consumed was higher in the current study than in short-term experimental studies. Although the dose-response relationship between alcohol intake and HDL-C might differ for short-term and long-term intakes, the comparison suggests that alcohol consumption may be underreported to some extent in the health surveys. Without information on previous alcohol intake or the cause of alcohol abstaining, we were unable to identify and exclude previous heavy drinkers. Only a few surveys distinguished between lifetime and current abstainers, which we combined with current abstainers in order to harmonise the data. As a result, findings involving abstainers have low generalisability. This also precluded the combination of infrequent drinkers and lifetime abstainers into a single comparison group, which has been suggested as the best alternative [51]. However, the sensitivity analysis comparing the use of lifetime abstainers and infrequent consumers did not indicate differences between these groups, which we consider a strength of our chosen reference category. Combined information on consumption frequency and volume were available for the subgroup with additional data on current episodes of heavy drinking. Apart from the overrepresentation of individuals with low SEP at the more extreme end of intake levels, there was a consistent increase in episodes of heavy drinking with increasing alcohol consumption frequency in all strata of life course SEP, suggesting that the main exposure variable, alcohol consumption frequency, differentiated individuals according to average alcohol intake. To reflect life course SEP, we used an index constructed from multiple indicators that, with the exception of education, we derived from census surveys performed decennially between 1960 and 1990. In order for all participants to have the possibility to obtain a full score, we imposed selection criteria regarding immigration, birth date, time of death, and census participation. This resulted in a clearly defined and homogenous sample, but at the expense of sample size. Previous studies have assessed the relationship between SEP and the risk of alcohol-related outcomes [54], and to various degrees, the mediating role of alcohol consumption [55]. Our study appears novel in the sense that it assesses the relationship between alcohol consumption and CVD mortality within strata of life course SEP, which appears to be very sparse or nonexistent in the current literature, regardless of how SEP is measured [55]. A possible reason could be the extensive sample size required to test for differences between groups and the registry linkages required to measure life course SEP, which highlights the major strengths of this study. For this reason and because of variation in alcohol consumption patterns, alcohol taxes, and socioeconomic inequalities between countries, it may be difficult to repeat the study in detail. The overall findings, however, should be available for replication in another population using similar study design, albeit with variation in the measurement of SEP or alcohol consumption. --- Conclusions In this observational study of Norwegian adults, we observed lower CVD risk among frequent consumers of alcohol compared with infrequent consumers and higher CVD risk among current drinkers who reported frequent episodes of binge drinking in comparison to current drinkers who did not binge drink during the past year. The lower risk of CVD mortality associated with frequent consumption appeared to be more profound among those with high SEP throughout their life course than among those with middle and low SEP. We also observed higher CVD risk among very frequent consumers compared with infrequent consumers, but only among participants with low SEP. It was more uncertain whether the association between binge drinking and CVD mortality differed by life course SEP. --- Committee, https://helseforskning.etikkom.no/ page/forside?_ikbLanguageCode=us, and from Statistics Norway, http://www.ssb.no/en/, as well as to ensure that the new member signs a confidentiality agreement with Statistics Norway. It is also possible to apply for data from NIPH, https://www.fhi.no/en/more/access-to-data/ elektronisk-soknadsskjema-for-datatilgang/, and Statistics Norway, http://www.ssb.no/en/omssb/ tjenester-og-verktoy/data-til-forskning, from the linked webpages. ---

Socioeconomically disadvantaged groups tend to experience more harm from the same level of exposure to alcohol as advantaged groups. Alcohol has multiple biological effects on the cardiovascular system, both potentially harmful and protective. We investigated whether the diverging relationships between alcohol drinking patterns and cardiovascular disease (CVD) mortality differed by life course socioeconomic position (SEP).From 3 cohorts (the Counties Studies, the Cohort of Norway, and the Age 40 Program, 1987Program, -2003) ) containing data from population-based cardiovascular health surveys in Norway, we included participants with self-reported information on alcohol consumption frequency (n = 207,394) and binge drinking episodes (!5 units per occasion, n = 32,616). We also used data from national registries obtained by linkage. Hazard ratio (HR) with 95% confidence intervals (CIs) for CVD mortality was estimated using Cox models, including alcohol, life course SEP, age, gender, smoking, physical activity, body mass index (BMI), systolic blood pressure, heart rate, triglycerides, diabetes, history of CVD, and family history of coronary heart disease (CHD). Analyses were performed in the overall sample and stratified by high, middle, and low strata of life course SEP. A total of 8,435 CVD deaths occurred during the mean 17 years of follow-up. Compared to infrequent consumption (<once/month), moderately frequent consumption (2-3 times per week) was associated with a lower risk of CVD mortality (HR = 0.78, 95% CI 0.72, 0.84) overall. HRs for the high, middle, and low strata of SEP were 0.66 (95% CI 0.58, 0.76), 0.87 (95% CI 0.78, 0.97), and 0.79 (95% CI 0.64, 0.98), respectively, compared with infrequent users in each stratum. HRs for effect modification were 1.30 (95% CI 1.10, 1.54, p = 0.002; middle versus high), 1.23 (95% CI 0.96, 1.58, p = 0.10; low versus high), and 0.96 (95% CI 0.76, 1.21, p = 0.73; low versus middle). In the group with data on binge

Polish discourse about family and school cooperation is steeped in the instrumentaltechnical perception of parental involvement in a child's education both at the theoretical inquiry level as well as at the everyday school practice level. This approach to parents' involvement is focused on ways of winning parents over to collaboration with the school in accordance with the school's priorities and goals. It manifests the normative and uncritical way of promoting cooperation in specified and limited forms . Because this way of thinking does not always meet the parents' expectations or needs, it results in tensions and frustrations on both sides. Correspondence concerning this article should be addressed to Marta Wiatr, e-mail: [email protected] This sense of incoherence is intensified by two socio-cultural currents. The one is the arising parental ideologies with "parental determinism" and "intensive parenting" at the forefront. The other are the debris of politico-economical processes of education decentralization and democratization emergent in the '80s and '90s such as formal parents' representations unitsparents' councils. These processes and phenomena compose a backdrop for parental activities. The educational initiative, undertaken by Polish middle-class parents and citizens of large cities, seems to embody the engagement ideal. After a closer look, this vision of engagement becomes vague. In this article, the parents' activity will be portrayed across the spectrum of instrumental involvement and participative engagement. I believe that a theory of reflexive modernization may help in understanding the parental commitment and understand shifts in the meaning of parental engagement. Phenomena such as flexibility, individualization, secularization, rationalization, alternative lifestyles, uncertainty, and etc., seem to have an enormous impact on individuals' willingness and readiness to accept the risk of shaping children's educational paths. Some parents use cognitive reflexivity to break from conventional ways of acting and to transform habits and practices linked to their own child's education. In the first sections, I will contrast the two concepts of parents' commitment to children's education: involvement and engagement. Against this backdrop, I will present the emergence of Polish quasi-schools and their broader socialpolitical context. This will be followed by a research project presentation complemented by examples of analyses indicating the emanation of the engagement category. In the last two sections, I will consider a concept of reflexive modernization by Beck and Giddens in examining the engagement category and understanding parents' educational decisions and activities such as reflexive modern practices and phenomena. In the end, I will bring a critical theoretical perspective to the concluding discussion. --- Engagement vs Involvement Although not all scholars make the distinction between involvement and engagement, I highlight the difference in modality between these two notions. In this context, involvement refers to activities and ways for a parent to interact with the school environment. It is considered to be a key to student's academic success and to school improvement . It portrays a school-driven way of home and school cooperation. School, being at the heart of activities leading to cooperation, is in a position to determine goals and assign tasks drawing from its priorities and needs . As long as parents adhere to the school's priorities, the family-school collaboration is considered to be appropriate. The teachers are experts who know best what is needed or required, and parents can take on the role of spectators or audience members. Parents' perspectives and diverse resources are not recognized by the school and thus are overlooked. As I mentioned earlier, the main concern is to win parents over and induce them to internalize the goals and tasks indicated by teachers . As long as the main purpose of winning parents over is improving students' achievement and strengthening school programs and curriculum, the leverage of such cooperation incorporates instrumental-technocratic purposes. Such parental involvement programs rest on encouraging the parents to be active in order to provide advantages for their own child. Therefore, it remains within the reach of the individualistic approach. Unlike involvement, engagement is deemed a process rather than a standalone activity or set of activities . It is defined as a dynamic, distributed and interactive process that parents go through to navigate barriers between home and school. Conjointly, it acknowledges that parents are not a homogenous group and thus the barriers are individually diverse due to parents' various social and cultural backgrounds. Engagement entails co-constructed, shared responsibility based on meaningful and situated participation nested in a system of comprehensive supports. It reaches beyond home and school activities towards diverse institutions operating in the neighborhood. Engagement is a parent-driven process and it involves their needs and priorities as well as their communities' priorities and resources . The parents are active participants and partners. The school takes the adjunctive and supportive role. Parents engage because they feel challenged to do something about what they feel is important. The purposes of their activity go beyond solely supporting their children in school and extend to collaboration to transform the local community . Thus, the process develops participative, emancipatory, and empowering potential. The spectrum between involvement and engagement seems to determine a course for Polish parents' educational initiative. The context of the initiative is shaped by the three phenomena and processes: decentralization and democratization of the Polish school system dating back to the '90s; parenting ideologies; and the sense of being trapped in 'involvement' mode. --- Decentralization, neoliberal parenting and parental "encapsulation" experience The emergence of the first quasi-schools incorporates a culmination of social-cultural processes and phenomena taking place throughout the last three decades in Polish society and its main structures including the educational system. Decentralization and democratization processes in the educational system , has tangled with the neo-liberal parenting ideologies, and with simultaneous parents' experiences of being omitted both by school staff at a school level and by politicians at the national reform level. It has led to the emergence of new parental practices linked to children's education. At the educational system level, the proceeding of decentralizations boiled down to handing over the state schools to local authorities . In addition, the possibility of creating and running schools by other organizations besides local authorities' units was allowed 1 . At the school level, the rules of co-managing the institution became regulated by establishing school stakeholders, including obligatory parents' councils and facultative school councils. They were ascribed certain limited competences 2 . At 1 In the early 80' within the framework of the emerging Solidarność social movement, the ideas of school autonomy arose in Poland. Significant changes came about in the '90s, after politicaleconomic transformation. Starting from 1990 the decentralization ideas were brought to the structure of national power. The Local Government Act dated to 8th of March 1990r. specified the tasks and obligations of the local authority which has been called to independently perform the tasks of public administration, and also equipped with material means enabling the implementation of its tasks. The Act of 7 September, on Education System indicated local authorities as a leading authority in organizing and carrying out educational tasks on the given territory. Besides the local authorities running public schools, the pronounced afore act expressed acceptance for other unitsindividuals, companies or associations for organizing and running non-public schools. It also provided the basis for non-public schools for applying for targeted subsidies from the state budget for scholarships and education of children, adolescents, and adults. The prior state monopoly for conducting universal education has been weakened. 2 Due to the Act of 7 September 1991 on Education System and the later Act of 14 December 2016, the Law on School Education, this level, the school's democratization seemed to open a space for parents' presence and voice inviting them to partake in a responsibility for children's education through various forms of parental involvement in their learning . In the mid-'90s, decentralizing tendencies and school-improvement concepts had started to weave with the new powerful discourses of accountability for education and students' performances, educational standardization and parameterization 3 . They seemed to have their roots in marketization which was strengthening at the time in Poland . The social-cultural shift that happened during the politico-economic transformation process in Poland opened the field for parents to search for their own life projects, including those related to parenthood and a child's development. Expansive neoliberal parenting ideologies such as parents' determinism ) and intensive parenting provided parents a new vision of their parental role. This also established visible shifts in the ways some parents started to construct their parental behavior. The parents received signals that the quality of their parenting determines the child's future happiness and wellbeing. Hence poor parenting was regarded as accountable for offspring's future failures. Coupled with the decentralization and democratization processes, neoliberal ideals of parenting brought about the parentocracy . The ideology of parentocracy allows parents freedom in their choice of their children's educational paths at the same time making them responsible for these choices. Parents, like customers in the free school market, choose schools or other educational settings, suitable to their needs and priorities. The schools selected by certain criteria are made accountable for fulfilling declarations made. The third essential phenomenon combined in the background of quasi-school emergence was a mass experience of the unimportance of parents' the parents' councils merely consult and evaluate school decisions. 3 Decentralization from the beginning was split between pedagogical-political and economic goals. In Polish reality, it soon became clear that the economic goals became the leading ones within the ongoing processes . This trend led to superficial participational and democratic changes in the school which resulted in making parents obligatory "partners" of the school. perspective both in a particular school and in a broader systemic perspective. The lack of space for parents' participation, other than the one limited to fulfilling the priorities determined by the school staff, was frustrating for many. It took the shape of encapsulation within parents' councils. Due to the limited competences of these councils, parents felt in fact cut off from the decisions being made in school. At the higher level, parents, who in public expressed their opinions and shared their ideas and felt manipulated and ignored by political parties. It became obvious that politicians have been using parental anger and determination solely to build advantages over their political adversaries . New institutional and political conditions created superficial participation, which pushed parents towards the position of clients or adherents, but not participants or partners. --- Emergence of Quasi-Schools In 2013 a group of Polish parents decided to take their children out of both private and public schools to create an educational environments for the children, in which they could develop and learn better, according to parents' opinion . I shall call these new educational environments -quasi-schools. Quasi -because these settings operate without school status and, therefore, they are not subject to pedagogical supervision of any educational authorities. This yields greater latitude in organizing the setting. They are not obliged to hire certified teachers, provide grades, and adhere to any sophisticated organizational and hygienic outlines. In 2013 four quasi-schools in two heavily populated Polish cities took off. Currently, dozens of such establishments operate in Poland and more are about to start. Their supporters draw from progressivism and modern psychological and neurological concepts of "brain learning" . Rarely do the quasi-schools' architects relate to democracy concepts or the broader social order. Quasi-schools operate based on the legal footing passed for homeschooling which is under certain circumstances legal in Poland. Parents, however, do not educate their children at home, but they organize smaller or bigger permanent "learning groups". Their concept is close to Meighan's flexi schooling where children can learn in different configurations and locations according to various schedules. In order to stay in homeschooling mode, children have to pass the yearly exams. How, when and where children learn for these exams is, however, not fully structured and hinge on the child. Failing the exams, however, entangles certain consequences for parents and subsequently for children. Parents lose the right to "homeschool" their child, and the children have to return to regular school. This kind of setting is not prohibited under Polish educational law. It is, however, also not explicitly outlined in it. Many would say, that determined parents devised a way within the educational law, to implement independence and participation. --- Research project The noticeable contemporary involvement of certain groups of parents in building educational environments has become a starting point for my considerations. This article draws upon a fragment of exploratory research conducted from 2014 which aimed to explore the essence and basis of exceptional parental involvement in child education in reference to parental learning ). The research questions concerned the parents' actions themselves such as their repertoire, forms and scope; the meanings given to these actions by parents; as well as wider ideologies in which parental activities find their justification. The qualitative study was based on constructivism and poststructuralism . This main theoretical framework was complemented by other selected theories and concepts. In this elaboration, the reflexive modernization approaches conceptualized by Beck and Giddens were referred to . The study employed the case study method with applied strategies of constructivist grounded theory and discourse analysis embedded in the sociology of knowledge . In-depth interviews were conducted with eleven mothers and fathers of preschool and schoolchildren involved in quasischools. The parents were representatives of the "new middle class" . They all grew up in socialist Poland and they are all denizens of heavily populated cities. A few parents were leaders of the first quasi-schools others had their children attending these establishments. Some were interested in emerging quasi-schools. All interviews were transcribed and analyzed from bottom-up. During the analyses, the engagement category arose. Its multidimensionality and diversity activated different paths of interpretation for "what was going on" within the described phenomenon. The parents' determination, readiness to assume risk and responsibility for a child's education seemed to fit in with the ideals of strength, resistance, participation and empowerment/ emancipation. After closer examination, the parents' actions appeared to reveal something more which calls upon a more detailed examination. Therefore, in what follows I shall present the emergence of the engagement category. I shall theoretically elaborate on it using the concept of reflexive modernization . I shall discuss the floating meanings of parents' engagement in a child education from different perspectives such as empowerment, engagement, participation on the one hand, and psychologization and disempowerment of the social realm on the other. --- Between risk, madness and responsibility The decision to withdraw a child from the educational system by parents is not an easy task. It is described by parents in terms of challenge, risk, uncertainty and fear but simultaneously as a duty, necessity or responsibility. This is how parents portray their decision: And I was afraid of this. That WRENCHING someone from a reality, which is, familiar, "marked with own pee", smells familiar, and recognized. [M05 34-34] …the anxiety we all feel, is: what if the child does not pass on core curriculum . So we're not forcing it to learn ... so, how are we supposed to make it learn? Like, say Physics. If it doesn't want to learn Physics? [F03 8-8] At the same time, parents feel that knowledge itself about the feasibility of other-thanmainstream education makes organizing it unavoidable: And suddenly it turned out that this is not a freak of nature, and nobody knows what it is doing, but it turned out that there are a lot of such places. So the credibility for me increased, that …this can be done. If it can be done, well, my children certainly will not stay put for 45 minutes. [F09 15-16] Parents report unwanted transformations the child undergo after having started attending systemic school. They refer to them as circumstances of their decision: ...Marta stopped showing any desire to do something else in her life. This search for her interests and passions. The school is focused on results, on the tests, and, and this is something that disturbed me a lot. [M04]. "... in the second grade, we already saw that there was a lot of apathy and sadness in a child who had been joyful and curious about the world. And who wanted to shut down, relax after school wanted to come back, preferably turn on a fairy tale, cut off from the world and just let the steam out" [M05 2-2] The risk of harm mentioned above, however, does result exclusively from the situation of leaving the child in the given educational system. It turns out that taking the child from the regular school entangles different types of risk and uncertainty. While the first perspective is connected with the risk of loss of a child's innate "virtues". The second one entitles some other problems. Some parents are afraid that this kind of solution might not last long. They worry that due to the child's potential failure at the yearly exam, the child would have to return to the regular school. Some of them are worried that the new setting will collapse due to the other parents' withdraw. The quasi-schools are fully dependent on another parents' support. Finally, some parents are afraid that their children will not take part in essential for future life generational experience: in the beginning, this rebel is really that, that you take out your children from a mainstream, and I do not know whether it would not hurt them, right? Will it not turn against them later on, right? Because at the end of the day you have to live in the system too, really. Well, not in the educational system but in a system that is a derivative of this system, right? And THIS was the biggest anxiety for me. The more that I had no outside support except for the parents -the lunatics, who went about it, right? [M07 30-30] Parents strive to balance their unobvious choices and the risk they carry out. The uncertainty of the choice is heard in two conflicting voices that represent two strong discourses. One is rooted in the belief in proven systemic education as leading to anticipated vocational success. The discourse builds on notions, such as, high-stake tests, academical achievements, effort and competitiveness: Maia's leading teacher was a person who competed with the 7-year-olds. Anyway, she used to say very often, "Ladies and gentlemen, we started the same textbook as seven-year-olds have," and then in second grade -for example: "Ladies and gentlemen, we are already in the fourth part of the textbook for the second grade, while the 7-year-oldsimagine that! -they have just reached the third part". [M05]. The second voice comes from the anxiety concerning the child's balanced and full development. It is saturated with such notions such as, emotional security, authenticity, creativity, inner-integrity, self-directives, autonomy, relationality, respect and responsibility: …this first grade was painful for her because she was brought up and used to the fact that she really made choices alone and did not have to adapt to the whole group. Which means, she could do it when she wanted to and when she chose this group. And when she was in a class, which she had to sit upright, she had to keep her hands on the bench, she had to ... do a lot of things she didn't understand: "why?". And asking questions that were cut, because they generally interfered the lesson. [M05 2-2] Two voices are combined and intertwined together in multiple ways. Sometimes they appear as a discussion parents have with other parents or other people. One father says: "I was sitting at the parents' council meeting, where it was presented in percentages and tables, which class, when, in what state, why, how much moved up, and what the headmaster would do in reference to it. So now, I am really broken that I didn't get up and didn't say that I wasn't interested and that I'd have talked about something else. But I was the only parent who didn't want to talk about it. After that, I talked to them about it and in general: "No, this is very important, school ranking, our children, high school diploma, something". People. [F09 65-65] Parents know from experience that doing things differently is perceived by others as taking away a chance for awaited child's success and happiness in the future. They report about other parents affirming and engaging in school achievement discourse: all children began to apply for various contests . I know from conversations with parents that these are the ambitions of the parents, not the children. They are. Because even a child who is really fascinated with, insects, is fascinated by insects and not by nature as such. So, this contest's thing is a pressure from parents. And among those children who I see, I do not see someone who would be fascinated by, e.g. a spelling competition. But parents know that there are additional points, that you can win something, so they push their children. [M04 9-9] Parents also hear other people, such as friends or relatives questioning their decision: "I work in an industry that should be quite liberated and have imagination and so on, and the people younger than me, I don't know, even 15 years younger, said: "What do you do? Why are you taking children from school? After all, the school doesn't bite. Each of us survived the school". [M07 30-30] Parents are aware that for many, it is foolish to abandon the proven solution even if it is not perfect. They hear that it is insanity, irresponsibility or fantasy. A father reports his discussion with his own father: …we had a really huge fight when it came to what I wanted to do with my child, what an evil-... I have such an email, such, such a compendium of unschooling, so "look at yourself, read, watch the videos". He got one sentence out of it and with one sentence he responded that this is not for everyone and so on, and it ended there. So, you know, it really evoked specific emotions..., a new kind of internal family conflict. [F03 200-200] Sometimes the competing discourse is expressed with one's own voice: it was difficult for me because I felt very strongly that I was right, but I couldn't find it, you know, somewhere like this confirmation in normal people, right? In those more like me than those completely there, those who have already left this system. [M07 30-30]. Those who are on the other side are no longer considered "normal people" even by the parents who want to join them. Stepping out of the proven solution, the parents require precise justifications to legitimize controversial choices. In the search for these justifications, they re-construct themselves using In fighting fears and uncertainty, parents are assisted by psychological and neurological books and parental guides. The parents also turn to experts, who are ready to take care about the process of parents' psychological transformation. These activities inscribe in the intensive parenting discourse , which is entangled with therapeutization and psychologization . A mother whose child is in quasi-school, explains her psychological work as a process that she undergoes together with other parents in this quasi-school: due to the fact that we went through the group process, and we also had a lot of psychological support for the whole time, and I think we are on a similar level, and on the same path -it seems to me -that we all are, despite all kinds of fears and doubts. [M05 34-34] Strengthened by experts and available readings, the parents become equipped to present their coherent justifications, in reference exclusively to the discourse of the child's emotional safety, authenticity, responsibility, innate curiosity and autonomy. By doing so, they construct themselves as exceptional parents and create their "local rationality". It makes their decisions and behaviors consistent, meaningful and legitimized. A father's account of a discussion he and his partner had is a good illustration of the ongoing negotiations, within which two discourses compete and "local rationalities" emerge: "She does not believe in this solution, that random people who aren't competent, who aren't educators, don't have pedagogical education, and, there is one psychologist, [they] will mentor and will deal with the education of our child, that it can't work that way . The belief that the child is able to start to be interested and learn by itself, was also impossible for her to pass. She didn't understand that the child will have a natural desire and interest to explore the world. …she remembered from the past that even after high school she didn't know what she wanted to do in life. "Then how such a small person would have known?" It turned out that we confuse the basic concepts, such as, initial acceptance, such optimism and that we would raise a child differently and that we would enable him in a natural way to explore the world, through experimentation, making mistakes, social integration, problem solving And because it is also cool that these kids like themselves, decide and set their own rights, regulations that they follow and that they solve conflicts there. It was cool, but, in a fact, these are one principle, and your internal beliefs, stereotypes, the pattern in which you were growing up, this is the second. This is probably the biggest barrier for all who have these, taking the children away. Because we are stripped of such flexibility of the idea that a really different system from the one we went through is possible to implement. And that it is effective or is just as good, if not better because the child gains additional strengths in the form that I didn't know, openness, way of seeking knowledge, and not only cramming and passing exams. [F03 2-2]. The father sets the ground for the new education, and accuses other parents and partially himself for being tied by old structures. He is eager to liberate himself; to emancipate from the old structures. What his partner considers important, such as, well-educated teachers and qualified pedagogues, is not important to him. While his partner believes that each child needs structure and strong leadership, he claims that children benefit from the lack of structure. In liberated learning, where emotions and cognitions come into play, the psychologist seems to be the right choice. His partner is afraid of the child's failure in terms of education. He values the child's freedom, autonomy, and selfconfidence. These concepts are handy for him to emancipate from the coercion of the old structures and language employed. He questions the mother's attitude as restricted by stereotypes, and fear-driven. Parents set up new local rationality as well as new rational, reflexive, well-informed and responsible subject. They argue that what others concern insane, is, in fact, rational and derives from parents' profound sense of responsibility, broad comprehension, and scientific knowledge. They challenge -in their opinion -outdated dogmas and structures. They revise assumptions about the child's development, its developmental needs and about being parents. It is illustrated by a mother's utterance: "when Olga was small, I started to develop myself. I did some interpersonal trainings there. I did it to be a good mother". [M02 67-67] Some parents feel that they need to transform into good parents. Later on the same mother's relates: "I went through a personal transformation, because I changed, my husband changed, my surroundings changed, and so our environment changed a lot, through motherhood" [M02 69-69]. Parents, however, do not see it yet, that constructing a "better" parent as well as changing the "local rationality" and language justifying it, they become captured and restricted by the other discourse, and new structurespsychological ones . They shape their vision of a child and its development in reference to psychological knowledge. The child seems to be a susceptible and fragile creature, vulnerable and helpless vis-à-vis the austere mainstream school. Parents' vision of the school is evaluated as harmful for the child and unfriendly to parents' participation. They do not believe in the success of the inside transformation of the given educational system. One mother describes it as a fight with windmills. [M02 81-81] Parents' motivation derives from their concern for their child's well-being and emotional safety. It is oriented to improve present conditions for its development, as well as to protect the child's future. There is a point when an active subject emerges in the form of an engaged parent who is ready to take the floor, and ready to engage multiple individual resources to protect a child. At this point, parents' determination seems to include important transformative traits of engagement and participation which potentially might result in transforming social structures and social relations. The "go-getter" parents search for a space to act and participate. Soon they discover that none of these formulas are envisaged in the present school. Instead, they are offered to follow the school-driven and school-managed collaboration models passed down in stiff traditional school structures. Once activated, parents challenge the school system and dare to look for better possibilities. --- Reflexive Modernization Era and Reflexive Subjects Categories such as reflexivity and rationality of the subject, its permanent exposure to uncertain and controversial choices, perpetual struggle for their justifications, acceptance of individual responsibility for own decisions, as well as high skills in mobilizing available resources, have led my theoretical explorations to the theory of reflexive modernization offered by the three authors: Ulrich Beck, Anthony Giddens and Scott Lash . Although their perspectives were not thoroughly convergent, the authors offered three explanations of contemporary social changes. Aware and familiar with critical revisions of this approach I claim that it remains a valuable framework for a more profound understanding of the parental activity. Its indisputable value consists of its possible extensions with such concepts as Bourdieu's habitus, Rose's politics of life, Foucault's governmentality, Althusser's interpellation or Furedi's therapeutic culture and psychologization. The main tenet of Beck's and Gidden's approaches is that reflexive modernization is a process of modernity alternations. The reflexive character of these transformation means that the changes occur not as a result of political planning, based on some diagnosis, but they occur surreptitious and unplanned. They are 'side effects' of modernity . Beck claims that the reflexive modernity process consists of methodically questioning the modernity and its faith in the power of science and technology. Challenges of the modern world result from the discovery that the gain from techno-economic progress has its dark side in the form of uncertainty and risk . As Beck constates, we now all live in the 'risk society' or as Giddens calls it in the situation of "manufactured uncertainty" . New "mega risks" have various social consequences, both structural and individual. It causes modernity self-confrontation and questioning traditional structures as well as profoundly influences the social environment of individuals and transposes into the fabrics of personal existence ). The awareness is growing that institutions designed to manage the risks are the ones responsible for its production. It entails the process of undermining traditional structures and institutions and results in the dismantling of conventional norms and patterns in which individual and collective decisions used to be rooted. Owing to the liberation from conventional norms and ties, the individual becomes free to make their own decision and to construct their own life as one pleases or how one is able to make their choices. From now on, individuals are to choose and decide about "education, profession, job, place of residence, spouse, number of children" . Doing so they assume the role of makers of their own meaningful biography, beyond the conventional patterns and they become a reflexive do-ityourself project ). --- Towards Subpolitical Engagement and Beyond The newly gained freedom is an ambiguous nature as it sets individual insecurity. Free choices are unavoidably risky and unpredictable. Do-it-yourself-biography becomes a risky project and the individual life becomes life at one's own risk. Consequently, people are forced to live paradoxes. The conditions of a risk society exert pressure on individuals to make choices and to take the individual responsibility for potential individual failure. The liberated individual is not entirely released. As the old structures dismantle, the new coercive structures emerge . The 'free' individual becomes entangled in a new network of norms and patterns of behavior which are mediated by the labor market, the educational system, the professional regime, the mass communication, the overwhelming consumer markets, experts' discourses, and etc. Emerging structures offer diverse systems of knowledge, which are often incoherent or even contradicting. It means knowledge is exposed to questions. Its truth claims are limited to a certain context . The experience of conflict about what and how and under what circumstances people know results in the emergence of multiple local rationalities where individual decisions and choices find their justification. The human existence in such conditions requires developed reflexivity from the acting subject -the aptitude to calculate, collect and analyze data, and to stay informed . Multiple rationalities open ground for political struggles which, due to the collapse of the traditional institutions, are debated far from roundtables and administrative levels. According to Beck, these multiple rationalities with their truth claims and priorities, spawn the field of subpolitics . Subpolitics, denotes the reflexive modern way of articulating and realizing priorities of groups, their interests, and needs at stake, beyond the world of formal politics. According to Beck, a gap after political involvement of the first modernity is filled in by microlevel and subpolitical society governed from below. Subolitics are the spaces in which agents, reflexive individuals negotiate their local rationalities, and doing so engage in the redesign of the social . The subpolitical activity resembles the individual cooperationalism but is however regarded by Beck as a "boon" and an opportunity for democratic processes . The way parents engaged in quasi-schools create their own rationality and justification of their doings seems to inscribe in the subpolitical process typical for reflexive modernity. They find common spaces for co-operational, subpolitical engagement in order to fulfill their own vision of education and support for children's development. The lineage and dynamics of the quasi-school phenomenon could be shortly portrayed on a continuum. The starting point is the introduction of superficial parental participation in school. It manifests in the school-driven parental involvement as well as in the encapsulating of parents' councils. The next stage emerges when under the pressure of the parental ideologies, some parents, heavily focus on the protection of their child's fragile psyche and transforming their parental role. The final point is the shift toward the parents-driven engagement. It happens however beyond conventional structures and proven solutions of the traditional school system. Parents find a way to put their perspective and intentions "in motion". They are able to afford it because of the multiple resources available. On the one hand, parents dislocate from the system in which their roles are static or superficial and their needs for high-quality parenting are disregarded. On the other hand, creating new, better educational environments, parents dismantle traditional institutions. They act in a subpolitical way. Parents organize the new educational environments freely, independently, from scratch and according to their own ideas. The new educational settings are composed of different available "blocks" like labs, lectures, workshops, crafts, and parents' other resources focused around the child's developmental needs. Being so, quasi-schools inscribe in the mode of other "do-it-yourself" projects. Next to the "do-ityourself" biography or "do-ityourself" family , the "do-it-yourself" school is born. The parental initiative proceeds along the continuum from being offered school driven involvement with encapsulating experiences to parents' driven engagement. The main tenet here is that the child's wellbeing and happiness are the main triggers of parental role and commitment. This great potential, however, "runs out" in an individual power that is untranslatable to social empowerment. Subpolitical co-operationalism, in this case, loses the sight of community needs, neighborhood's common good and the common social vision. In subpolitical engagement, parents are the individuals interested precisely in the education and development of their own children. Their drive is not embedded in their neighborhood but in mostly psychological developmental needs. --- Conclusion Feeling both obliged to act in a responsible and reflexive way for the sake of vulnerable children and constrained by the existing school regulations and practices, the parents decide to disassociate themselves from the unreactive structure and assemble their own do-it-yourself school. By releasing themselves from the constraints of the school system, parents do stop engaging in the improvement or modification of the old structures. Hence, those stay untouched. Abandoning systemic school, parents take over the control and the responsibility for their own children's education and bring about the individualization and privatization of the incumbent public sphere. Their energy and resources are directed to their kids only, with no broader reflection on the social realm and public engagement. Parents' perspective dwells in individual liberalism and their choices are a manifestation of their prior empowerment. The logic of reflexive modernization makes the parents feel not only ready but eligible and empowered to take responsibility for their children's happiness and well-being formulated in a therapeutic and psychologic terms. The fact that some parents are well-equipped to individually benefit from the reflexive modernization poses a question about the winners and losers of the contemporary processes typical to reflexive modernization. The do-it-yourself modus operandi seems to be more adjusted to selected groups of contemporary society. In this case, the reflexive modernization logic empowers the empowered parents in their do-it-yourself life. It, however, simultaneously disempowers the unprivileged, who, due to socioeconomical structure, do not have the required resources to live at one's own risk. It might result in individualizing social problems. In the process, the transformative concepts such as engagement, empowerment, participation in reflexive modernization mode remodel themselves and start serving the framework of the strong neoliberal discourses.

This paper lays down a discussion on the contribution of a group of parents who decide to abandon public education and create for their children new educational environmentshere called quasi-schools. It focuses on the engagement category captured during analysis and its shifting meanings. The theoretical elaboration draws on exploratory qualitative research that started in 2014 which aimed to explore the essence and basis of the phenomenon of exceptional parental activity. The research questions concerned the parents' actions themselves and the meanings given to these actions by the parents, as well as wider ideologies in which parental activities find their justification. Aside from the theoretical framework composed of social constructivism and poststructuralism, in this elaboration, the concept of reflexive modernization by Beck and Giddens was employed. The research utilized the case study method with applied strategies of constructivist grounded theory (Charmaz) and discourse analysis embedded in the sociology of knowledge (Keller). Interviews with e1even parents involved in quasi-schools were conducted and analyzed from bottom-up. The analyses allow us to reconstruct the multifaceted meanings of parental engagement in late modernity as corresponding with ideals of strength, participation, democracy and empowerment on the one hand and endangerment on the other hand.

Introduction Hong Kong's Umbrella Movement was a mass protest in the form of 79 days of occupation between 26 September and 15 December 2014 in three key locations, blocking main arterial roads through the city. 1 The movement was part of a wider struggle for democracy and genuine universal suffrage in Hong Kong. It developed from the 'Occupy Central' campaign against the decision of the Standing Committee of the National People's Congress not to allow a fully democratic election of the Hong Kong Chief Executive 2017. The eventual occupation was far larger, more extensive and long-lasting than had been anticipated. It was triggered by a student strike and demonstration against the NPCSC ruling, resulting in an occupation of the main roads in the Admiralty district, close to the government offices, which quickly spread to Causeway Bay and Mongkok. The police initially responded with tear gas and pepper spray, against which the demonstrators protected themselves with umbrellas -hence the 'Umbrella Movement'. For the protesters, the occupation was an extremely intense experience. For others, whether sympathetic or not, it was inescapable: it was a central preoccupation of the population at large, dominated the news daily and disrupted the usual rhythm of life in Hong Kong -as well as the traffic. The campaign for democracy and associated political turmoil has since continued up to and beyond the Legislative Council elections in July 2016 and the Chief Executive election in 2017. Political opinion remains sharply divided, opening up deep rifts in Hong Kong Society which affect both public and private life. This paper is concerned with the relationship between participation or non-participation in the democracy movement and personal life, particularly the ways individuals 'do family' in the context of political turbulence. We are interested in both the politics of the personal and the personal consequences of politics, Venturing into the overlapping of the public and the private opens up a space for us to explore how campaigning for democratization of society might promote aspirations for the democratization of family relationships and how intimacy, which is traditionally considered as private and apolitical in Hong Kong, might become a site of political engagement. We focus particularly on the ways in which Hong Kong Chinese family practices, which still tend to emphasize harmony and hierarchy , might be brought into question as a result of political events. For participants in the Umbrella Movement, we argue, the struggle for democratic freedom became enmeshed with family tensions and was integral to how they made sense of themselves in facing up to authoritarian families and government. Whereas Chan and Ng found that family disagreement acts as a deterrent to political participation, our participants' stories suggest that when the political stakes are high, this may not always be so. We begin by situating our project in relation to relevant literature on social movements and families and relationships, before outlining our project and its methodology. We then go on to analyse how family relationships figure in the narratives of both non-participants and participants in the Umbrella movements and how participants, most of whom found themselves at odds with their families, sought to renegotiate family practices. In many of these narratives, individuals constructed reflexive accounts of self-transformation or resistance to change . --- Politics and family life The commitment to the politics of the personal, which was central to second wave feminism, led many activists to seek alternatives to oppressive family relationships in their personal lives . Yet within mainstream social movement literature there is relatively little work exploring the personal consequences of social movements relative to that on their political impact, as has been noted in influential overviews of the field Personal relationships also feature more prominently in social movement research on how mobilization happens . Within this body of literature, only that on that on the individual biographical consequences of political contention has a bearing on our concerns. Much of the pioneering work in this small field was concerned with New Left Activism in the USA, including civil rights activism, in the 1960s and 1970s . Among the most influential is McAdam's study of participants in the 1964 Mississippi Freedom Summer. These activists formed new relational ties that in turn sustained longer term political commitment. As McAdam notes in a commentary on these early studies, they suggest 'a powerful and enduring effect of participation on the later lives of the activists', despite differences in samples and methodologies . These effects include continued left-wing political commitments, careers in teaching or the helping professions and a greater likelihood of remaining single. More recent work summarised by Guigni covers a wider range of movements and includes those involved in lower level, more routine forms of participation that the 'high risk' activism featured in some of the earlier studies, but with similar findings. Much of this work is based on large scale surveys, which Guigni sees as a distinct advantage over earlier studies based on smaller samples. Such surveys do indicate some significant consequences of activism over individuals' life courses, including long term: for example that they are more likely to cohabit, have no children, be unemployed or work in particular occupations, but they tell us little or nothing about how activists make sense of the impact of political participation on their lives. They also, in focusing on life course effects, say little about the ways in which activism might affect individuals' existing close personal relationships during and in the immediate aftermath of episodes of intense activism. This is what we address here. The approaches we have found most productive are those that consider the conditions under which and processes whereby social movement participants experience a break with their past selves and world views. One such is Yang's study of the Red Guards during the Cultural Revolution. He argues that social movements can have 'liminal' effects, involving detachment from the everyday social world, 'when much of what has been bound by social structure is "liberated", and the transgression of norms and conventions becomes possible,' . Thus, during the period when the Red Guards were travelling away from home to 'exchange revolutionary experience' they underwent both an emotional and cognitive 'self-awakening' . Social movements, Yang concludes, can provide a setting in which 'freedom, egalitarianism, communion, and creativity' are valorised and reflexive self-transformation can occur. In a rather similar vein Debra King explores how activists come to distance themselves from and challenge hegemonic constructions of the world. For King's Australian activists there was no physical separation from their everyday place in the world as was the case of the Red Guards, but a cognitive and emotional one achieved through developing self-reflection and emotional reflexivity. In more recent unpublished work , 2 she posits the idea of an epistemic break resulting from the experience of confrontation and moral shock, prompting the emotional reflexivity that deepens commitment to activism. Here reflexivity is both emotional and cognitive, with an emotional response to injustice -shock, anger, outrage -leading activists to question themselves and seek a new world view. For both Yang and King, then, there is some form of separation or distancing from quotidian reality facilitating a reflexive remaking of the self. Whereas both Yang and King are interested in the movements themselves, we apply these ideas, particularly King's conceptualisation of epistemic breaks and emotional reflexivity, to understand how the self-remaking provoked by the umbrella movement is played out in the context of individuals' relationships with their families. This was particularly the case for activists who were part of the Umbrella Movement, who experienced the liminal space of weeks on the streets and the epistemic break that this engendered -not only through negative experiences of confrontations with the police but the positive experiences to which Yang draws attention -freedom, egalitarianism, communalism and creativity, apprehended both cognitively and emotionally. What the umbrella movement -and the wider democracy movement -have been fighting for is not only democracy but greater autonomy and freedom from control by China. One consequence of this resistance to political authoritarianism is a parallel resistance to domestic authoritarianism which our study reveals. There has been a long tradition of feminist critique of inequalities within families . Chinese families, like those of other Confucian cultures, have traditionally been marked by pronounced gender and generational hierarchies based on a strong collectivist orientation and emphasis on harmony . These characteristics are still evident in many Hong Kong Chinese families and continue to impose restrictions on junior members. Moreover, hierarchically imposed harmony is also consonant with the emphasis on harmony in the Chinese Party State's propaganda promoting a 'harmonious society' , to which the family is seen as central. While paying attention to inequalities, it is equally important to take account of the actual practices of family life -the ways in which families are sustained and may change through everyday practices. In this respect we take inspiration from David Morgan's work on family practices and Jamieson's concept of 'practices of intimacy'. In focusing on interpersonal relations and the everyday ways in which individuals enact family relationships, this approach allows for individual agency while not ignoring issues of power and inequality. As Jamieson notes, practices of intimacy are not always egalitarian. We can therefore consider the possibility of 'intimate injustices' that shape the experience of the individuals when facing conflicts and political disagreements within their families. Our aim in this paper, then, is to consider how the experience of the Umbrella Movement informs understandings of family practices and practices of intimacy within families, in particular in relation to challenges to hierarchical harmony in Chinese families and the wider political culture. --- The Study This was not a pre-planned study but was occasioned by extraordinary and unexpected circumstances. Our methodology evolved as we sought ways to understand the impact of the umbrella Movement on Hong Kong citizens' lives, informed by our existing interest in personal and family life. The data we present here derive from three sources: two sets of interviews with five paired men, who were strangers to each other, conducted before and after the occupation; a mixed gender focus group with five men and six women and another with five young women activists. All participants were Hong Kong Chinese apart from one mainland woman. We recruited through personal networks, by far the most productive approach in East Asian contexts. It fits 'somewhat more naturally with Confucian mores and expectations than attempting to recruit unknown individuals who lie outside networks' . Recruiting through known and trusted intermediaries helps to induce people to participate and to build rapport; in Chinese societies this is related to the importance of personal connections, guanxi . This strategy also enabled us to ensure that our small sample was as diverse as possible. The men interviewed and the members of the mixed focus group came from a variety of class backgrounds and a range of occupations, ages, sexualities and political orientations. The women activists group was more homogenous, comprising women aged from their early twenties to early thirties. There is, therefore, an imbalance in the sample in terms of gender and political orientations: whereas all the women were involved in the movement to some degree, only four of the men were. The interviews with pairs of men were begun initially for another purpose -exploring men's understandings of their intimate relationships as part of a larger study on Hong Kong men . It was these interviews that alerted us to the relevance of the political situation for personal life. The first three pairs were interviewed in July 2014, before the Umbrella Movement occurred, but when Occupy Central was active. Significantly these men referred often to the political situation in accounting for and justifying their conduct in intimate relationships. When the occupation occurred the project was suspended due to the disruption it caused. In April 2015 we resumed interviewing, involving two more pairs of men and conducting second interviews with all five pairs. At this stage the occupation became a central issue. For the men it was something they could not help but talk about: although only four had actively participated, it had affected all their lives in some way. At the same time we, as researchers, had become interested in the consequences of movement participation for our core area of concern -personal and intimate life. We therefore proceeded from the men's political preoccupations and probed further into the meanings of the Umbrella Movement for them, whether and how it affected their everyday lives and relationships and how they made sense of this. In May 2015 , as we wished to enquire further into the personal consequences of the Umbrella Movement, we conducted the mixedgender focus group. Some of these men and women had been on the streets throughout the occupation, some had supported it in other ways; others were bystanders, among whom attitudes to the movement varied from sympathetic to hostile. This made for some heated and often emotional exchanges, but also produced very self-reflexive accounts of the impact of the movement on their lives. The discussion in the focus group emerged from responses to our first question: do you think the Umbrella Movement was a success or a failure? Surprisingly, most perceived it as a success, despite it not achieving its political goals. The movement's supporters did not evaluate success in terms of constitutional changes but focused on personal transformation, increased social awareness, and renegotiated family relationships. Those who had not participated in or supported the movement also talked about its wider impact on everyday life. We subsequently asked them to elaborate on these points, eliciting narratives about their personal lives and relationships. In August 2015 we conducted the focus group with young women activists, having been alerted to particular problems they faced during the occupation. They were asked about their experience of the occupation itself, which raised issues about gender relations within the movement, and about the impact their involvement had had on their relationships with others, both inside and outside the movement. In interviews and focus groups we sought to create dialogical spaces that fostered interaction between participants, encouraging them to exchange views and experiences across their differences. The resultant data were recorded and transcribed in the original Cantonese. The original interview language was used for analysis, which we see as essential to preserving nuances in meaning: how participants talk about their practices of intimacy, how they locate themselves in relation to others and what this reveals about how they make sense of themselves and their social circumstances. Later we undertook careful translations into English of selected data extracts for publication purposes. Data analysis was conducted in an inductive, open and flexible manner to facilitate the development of alternative understandings of people's family and political participation. Full verbatim transcriptions were coded line-by-line, while themes that emerged from data were consolidated to form the basis of our analysis. The narratives presented in this paper are considered as both topic and resource . As resource they tell us something about Hong Kong family life, but they are also reflexively constructed stories of self and were produced as situated performances of self . These stories are also situated in time: told shortly after the event and told both to us, as researchers, and to co-participants. Narratives of the past are always constructed from the standpoint of the present ; they are products of the sense made of events at the moment they are told, including the temporal distance from the events they recount, and are shaped by the audience to which they are told . This approach to our participants' accounts necessitated paying close attention to how men and women reflexively made sense of their lives and relationships. Reflexivity as we understand it is relational: it is not simply an internal conversation with oneself , but is fundamentally social, a capacity developed through an interplay between self-other relations and internal dialogue. Following Mead we see it as foundational to a social self that is never fixed but can change with time and circumstance. Mead's theorization allows for differing degrees of reflexivity , from the basic ability to orient ourselves to and interact with others to more complex, self-conscious self-examination. It is novel situations that are likely to produce such heightened reflexivity . For many of our participants, the novel experience of the Umbrella Movement created an epistemic break that led to a reflexive rethinking of themselves and their relations with others. This process was not only cognitive but also emotional, in keeping with Holmes' view that that the 'reflexive self is formed by emotional relations to others' . --- Non-Participants in the Movement: Preserving the Existing Order Among participants in our study, orientations to the Umbrella Movement tended to coincide with attitudes to family life. Both those who had participated in the occupation and those who did not made explicit reference to the generational and gendered hierarchies that shape Hong Kong Chinese familial culture, but the bystanders were more accepting of the status quo. These were all men, which may be a coincidence and an effect of our sampling, but may also reflect men's vested interestsalthough there were also men who were very active in the movement. The non-participating men were not immune to the effects of the movement. As one, Hei, said 'we were all in the whirlpool whether we agreed with it or not.' Hei was among the bystanders who were sympathetic to the aims of the movement and drew attention to the parallel hierarchies of family and state: Around ten years ago, I did not challenge what my father chose for me. To a certain extent it was because I trusted him. I thought it was okay to let him choose. Now I have grown up, I think I cannot let the government choose for me. Unfortunately, the government also dares not to let us choose for ourselves. Although Hei raised the possibility of challenging hierarchical authority, he was hesitant to do so in practice, considering it too risky. He also explicitly defended patriarchal authority -his own -in the family, still wanting to make decisions for his children and not expecting his wife to disagree with him. He described himself as 'the boss in the house' and as 'not democratic but loyal'. The men who did not participate in the occupation often claimed that they sought to preserve the stability of Hong Kong. Some were critical of China's autocratic rule over Hong Kong, but they remained ambivalent about challenging it. They frequently employed a familial idiom, common in Hong Kong when referring to the Beijing government, the ultimate powerful master, as 'Ah Yae' , paternal grandfather, did not see himself as having a right to a political opinion. Whereas Hing, like some nonparticipants in Hensby's study, dis-identified with movement activists -they are 'not rational' -Shmily is critical of them but includes himself among those lacking in political wisdom. These men, while ultimately supporting or capitulating to the existing order, nonetheless demonstrate reflexivity and a relational sense of self. Ultimately, they are willing to accept the status quo and the authority of grandfather in Beijing. To do otherwise would pose a risk to the lives they have struggled to build for themselves and their families. --- The movement participants: Finding a Voice The movement activists also spoke of families in which the authority of patriarchs went unchallenged, but for them the movement had produced an epistemic break, giving them both a new perspective on, and an impetus to question, that authority. Many of them told stories of personal transformation that enabled them to challenge or at least passively resist the constraints families imposed upon them, which contrasted with the friendship, solidarity and new perspectives on the world that they experienced in the occupation. These participants often referred to the family patriarchthe father or, in one instance, an elder brother as autocratic or 'the boss'. In a particularly extreme case, Keung told of how he had been bullied and beaten up by his elder brother. He said 'Since my parents' death my big brother has been the boss. No-one dares to say no to him.' Yet he had found the courage in the context of the occupation to take to the streets in defiance of his family. In the focus group he contested Hei's claim that loyalty could compensate for a lack of domestic democracy: [My elder brother] is very loyal to me. He treats me as his little brother. But when I fight for my right to speak, he would never listen. This is a challenge to him so he treats me violently…It is important that he listens to me, know what I want, or just listen to what I have to say, but he never listens. Activists' narratives often included a description of the ways in which potential disagreements within families had been masked by silence until the Umbrella Movement rendered silence untenable. Lydia, for example recounted how her personal values were not important for her parents, 'they do not want to accept the idea that society needs to respect diversity'. She explained how such differences had been dealt with in her family: Our family is not good at handling differences in opinions. Our general pattern is that, someone says something, if you do not agree then you keep silent. This is our way of managing our family problems. Our silence is our solution… But in this event when you cannot be silent, this reveals the underlying problem. You have to solve the problem… Actually this is not limited to democracy. This is the change that is needed in the modes of human relationships. Maintaining harmony for the collective benefit of the family is important in Chinese culture; children are expected to take responsibility for the wellbeing of seniors in the family, which includes not bringing shame to them and ensuring they maintain face . Yet, as Lydia indicates, this has broken down in the context of the Umbrella Movement, which suggested to her that human relationships needed to change. The epistemic break that facilitated such insights functions in narratives such as this as a turning point, or what Denzin calls an 'epiphany', on which the narrative hinges. Wing also presented an account that turned on 'the event', leading her to question her family's imposed, silencing consensus. Basically we do not quarrel at home. My father is like a boss. My mother just listens to my father. Then the whole family would be silent. Only since this event I have discovered my parents are not always right.' Wing later contrasted her experience of the movement with her previous strategies for avoiding conflict over her lesbianism, by never coming out to her parents: 'if we do not touch on this topic [sexuality] there would not be conflicts'. The occupation, however, was not a topic that could be avoided, it dominated the media and daily life so that 'you couldn't get away from it.' Silence could be used by subordinates, as it had been for Wing, as a means of making space for themselves without damaging familial harmony. For both Lydia and Wing, however, this had become less viable. Disparate political stances towards the Umbrella Movement exposed underlying tensions in these families and silence has been replaced by a struggle to find a voice against the authoritarian familial consensus. A number of other participants also spoke about how the movement had provided them with an impetus for self-reflection leading them to resist authoritarian elders and develop independent thinking. It was not only subordinates within families who told stories of self-transformation and changing family relationships. One of the most striking examples we have is from Thomas, who represented himself as a reformed patriarch. He had spent 60 days in the occupation during which he said that he 'saw a lot of beautiful and virtuous things.' The epistemic break he experienced during the movement was one he narrates as being powerfully emotional. This 'emotional reflexivity' was also evident in his telling of the story, in which he disassociates his present self from his past self -the autocratic boss at work and at home. I opened a factory in the Mainland in 2003. I missed the growing up of my [elder] daughter in those ten years... I remember writing her letters during the Occupation. That was the first time I apologized to her. I was also that kind of autocratic person especially because I am an elder son. I decided most things, big or small. I told her I was wrong…The biggest effect of the whole movement on me is that I became softer. I thought I was a really tough person in the past. The time I spent crying during the occupy movement was ten times that of the past 50 years… Thomas's story suggests that the heightened reflexivity experienced through the occupation had created the potential for greater familial equality and intimacy than had previously existed. It was because he was the position of the patriarch that greater family democracy became possible -he could choose to relinquish his authority. The only other instance of family solidarity being enhanced through the movement was one in which there was no immediate patriarchal authority: a single mother, Venus, who told us that her participation in the occupation had strengthened her bonds with her teenage daughters: In most cases, however, the activists found themselves at odds with their families. Having become critical of the silence imposed by hierarchical harmony, they had to find strategies to cope with intra-familial political differences and opposition to their participation in the occupation. Unlike Thomas, they were subordinates in their families with little power to effect change. --- Intrafamilial conflict and negotiation Movement participants created a range of new practices of intimacy to deal with the tension and conflicts made manifest by the umbrella movement. For some this created actual or potential rifts, for others it meant retreating back into silence to sustain harmonious familialism to maintain a safe space for themselves within the family. In one case tensions rose to such a pitch that it created a major rift in the family. Chrystal, a young woman activist moved out of the family home because her 'stiff-necked' mother refused to compromise -and so did she. She felt that she had no option but to move out and 18 months later it had not been possible for her to return. This drastic strategy incurred significant personal costs -without a home she was reliant on sleeping on others' sofas and has been cut off from her entire family. She told us that she had recently wanted to buy a toy for her little brother, then realised she no longer knew what he would like. Peggy had considered splitting up from her husband because he did not support the movement: 'I nearly wanted to initiate a divorce. I told my husband: "I would never have married you if I'd known your political stance." I was actually very emotional.' Peggy's story, however, did not end in divorce: her refusal to remain silent and decision to challenge her husband's political views had a better outcome than might have been expected. Although she did not manage to change his attitudes to the movement he nonetheless tolerated her involvement and provided practical support of a kind that enabled her, despite arguments, to continue: I thanked him for taking care of the family. I always had sick leave and did not go to work [during the occupation]. We have a mortgage and he took care of all the financial burden so that I could be free to go there [to the occupation] anytime. Sometimes I did not go home to sleep, but he did not make any complaint. I came back the next day at 6am and we even had an argument. He then went to work and I went to sleep. I thank him for being so accommodating. . Peggy and her husband thus arrived at some kind of accommodation, adjusting their family practices to enable her to spend nights at the occupation. Others were less successful. Gin encountered emotional blackmail across the barricades -literally. Her brother is a policeman and they came face to face on opposing sides in the early days of the occupation. She described how her brother, instead of ordering her around as he usually did, begged her to leave: He said he did not want to arrest me. Also he needed to live his life. He loves what he does. He wants to advance.… Then he started crying. I understood and I wanted to preserve his dignity. I said I understood… Because his wife was just pregnant, they were going to be parents. Gin did succumb to this entreaty, but it did not end there. Like most of the participants in our study, she spoke of 'avoiding confrontation' as fundamental to family life, but she found a way to try to establish a dialogue. Gin's account offers insights into many participants' willingness to compromise. Caught between her political convictions and her emotional ties to her brother, she tried to stand up for her opinions while demonstrating her care for her brother, his wife and the baby they were expecting. I told him he did not need to be afraid. His sister is not doing bad things; she is just fighting for things for you and your children that he cannot fight for in his position…He may not even believe in this…I asked him to believe in Hong Kong Both Lydia and Gin sought a non-confrontational exchange through which opposing voices could be heard. While this did not produce consensus, a degree of understanding, and the dialogical space necessary for it, could be maintained. This suggests a small, but possibly significant, shift in practices of intimacy within their families. Others were less successful in achieving this, but expressed hope for respect and validation from their families. Keung, who had a history of mental health problems and was bullied not only by his patriarchal elder brother but also by his sister, provides an example. My sister always thought I was susceptible to others that I would rush to Lung Wo Road, occupy the road and sleep in the tent. She always called me to check my whereabouts. I think it was quite autocratic to ask me not to take to the streets. Actually her words were hurtful. Sometimes I did not want to fight against her and upset her. Therefore I said yes unthinkingly but actually I would take to the streets. Whereas Keung coped with family pressure through subterfuge, Wing said that she had learnt to stand up and speak for herself. Of her father she said: 'I know I actually can refuse to comply with his rules'. Yet she would also have liked to say 'I love you' to him, something very uncommon in Chinese families, indicating her hope that their relationship could contain their differences. Similarly Peggy expressed her continued commitment to making her marriage work despite her political disagreements with her husband. For these activists, struggling to have their voices heard within the family ran parallel to campaign to have their voice heard by the central government. Their claims for democracy in the family were modest: they wanted to be able to air their opinions and have them respected, to develop a negotiated intimacy that preserved family ties rather than requiring but no longer required total subservience. --- Concluding Discussion These aspirations for renegotiated intimacy are consonant with the emphasis on doing, variability and fluidity in the conceptualisation of family and intimate practices. This approach, however, also focuses on the everyday, the routine, the habitual . Considering what happens when everyday routines are disrupted by external events, as in the case of the umbrella Movement, is therefore of significance to the further development of this approach. The accounts of our activist participants tell of two forms of such disruption. The first is practicalthe reshaping of domestic routines necessitated by spending time on the streets. The second, which could interconnect with the first, is the questioning of what was previously taken for granted as 'normal' family life. The novel situation of the occupation and the issues this raised for their families engendered new forms of reflexive sense-remaking, particularly when the avoidance of conflict, so central to maintaining hierarchical harmony, was longer tenable. Thus not only were family practices unsettled, but those practices themselves became a site of contention. This reappraisal of family practices was a product of the critical and emotional reflexivity developed during the occupation, and the epistemic break that facilitated it, which also underpinned the stories of self-transformation told by many activists. These narratives portrayed a move away from acceptance of the status quo within family and society towards a new view of both. Thus Thomas represents himself as a reformed patriarch while others, such as Wing, Gin and Keung, told of forging a new sense of independence from authoritarian families. These narratives suggest that Hong Kong politics has, for some, created spaces for envisaging new ways of doing gender and family, aspiring to family practices through which they might achieve 'negotiable intimacy' , rather than authoritarian consensus. A politicized, reflexive interpretation of family practices not only challenges them but facilitates a more positive re-imagination of them. As Andrews argues, 'political narratives engage the imagination, not only in constructing stories about the past and the present, but in helping to articulate a vision of an alternative world' . Those of our participants who had not been involved in the movement expressed no desire to change family life. For them security lay in preserving the status quo and in some cases, such as that of Hei, their own patriarchal privilege. Those among them who, unlike Hei, opposed the movement insisted on the need to preserve hierarchy and harmony and saw the Umbrella Movement as socially divisive, harming Hong Kong's economy and risking the ire of 'grandpa' in Beijing. Thus their political non-participation was related to their ideals about personal life. The accounts of the impact of the umbrella movement we received from our participants were produced at a particular time, relatively soon after the end of the occupation. Had we talked to them earlier or later would have heard different stories. Given that narratives are always reconstructions of past events from the standpoint of the present, distance from the past and the narrators' current circumstances matter. We also have no way of knowing what the longer term impact of Hong Kong's divisive politics on these individuals' lives might be. If Thomas were to slip back into his old patriarchal ways or if Chrystal were reconciled with her family they might frame their memories of the movement differently. Our data, then, cannot tell us anything of the long term consequences of political upheaval on family lives, nor are we drawing causal inferences from them. These narratives do, however, begin to illuminate the circumstances under which family hierarchies might or might not be challenged and ways in which social movements may or may not contribute to reflexive self-transformation. They also raise issues of the complex intersections between gender, class and sexuality and their bearing on political choices, but not in a systematic manner. There are middle class men who felt that the movement was a threat to their comfortable life, but also working class bus and taxi drivers who also saw it as undermining what little they had. Some subordinate members of families in terms of gender and generation rebelled, but one patriarchal businessman reformed himself. An educated young lesbian told of how the difficulty of coming out motivated her to take action to change society, but the older gay man living on social security wanted nothing to do with politics. Thus orientations to the movement and perceptions of family relationships are not just about class or gender or age or sexuality. We need to consider intersections between multiple social locations as well as biographical experience before we can make sense of the choices of different individuals, which suggests the need for further research on these issues. The strength of family ties in Hong Kong is certainly a key issue. In Hong Kong's high cost, low welfare society survival depends on the mutual material support family members provide for each other . But family bonds are as much emotional as instrumental. Family opposition to involvement in the movement might well have been motivated by care and worryfear that a child, brother or spouse might come to harm, immediately or in the future, given the repressive potential of the Chinese state. Moreover, those who participated in the occupation against the wishes of their families did not stop caring about them: love, care and conflict are intertwined in the accounts we have gathered. When politics divided families, when the imposed silence of the hierarchical harmony of Chinese families failed to contain the struggle for freedom associated with the Umbrella Movement, some decided to speak out and find their voices in the family and society. Nonetheless most of the activists juggled family connectedness and disagreement by strategizing to display affection and care, and to maintain dialogue, while not abandoning their own political commitments. In most cases there was, therefore, an element of compromise with the hierarchical family structure. --- Endnotes: 1 It was ultimately cleared away by the police. 2 We cite this work with King's permission. 3 The yellow ribbon is the symbol of the pro-democracy camp. --- thereby referencing the traditional patriarchal and patrilineal characteristics of Chinese families. Tom , for example, suggested tactfully dealing with 'grandpa' by being 'humble' and 'harmless', because the real power is in the hands of grandpa, not his small grandson, Hong Kong. From such idioms we can sense the deeprooted patriarchal belief that people have no other choice than to comply with patriarchal authority. 'Happiness before rights' and 'not knowing enough to be involved in politics' were tropes often employed by non-participants to justify the preservation of hierarchy within and outside the family. Hei fell into the first camp, feeling that he had too much to lose to take action against Beijing's power -and would also risk his family's well-being for which he, as head of the family, was responsible. Other movement non-participants talked a great deal about being happy with their lives and saw little benefit in rocking the boat and offending 'Ah Yae'. Hing, who was opposed to the occupation, thought that Hong Kong people should be content with what they had rather than asking for 'one person one vote' and focusing too much on 'rights': '… I think Hong Kong people ask [for] too muchthe Umbrella Movement people are not rational' . He complained that the occupation had disrupted his family life as he spent so long trapped in traffic jams caused by blocked roads that he hardly had any 'quality time' to spend with his wife. Like Hing, Shmily, a gay man in his seventies, said that people should appreciate the government more: 'If you don't ask too much, you will be happier' he said. The idea that those in authority know better was also evident in Shmily's account. He said his family had taught him not to take a public stand and care about public issues and that he had not even taken to the streets in his youth when 1.5 million Hong Kong people demonstrated against the 1989 Tiananmen Square Massacre. He went on: 'In my opinion as a gay man, those who do not possess political wisdom should not participate in politics…You do not know anything. Then why do you speak about it?' Shmily thus represents himself as both a subordinate in the family and accepting of the authority of his elders and the government, as well as lacking the knowledge to take action -which he then extends to others: the ignorant and subordinate are not qualified to challenge the powerful. Unlike Hing, he did not express strong opposition to the movement; rather he

While social movement researchers have investigated how personal relationships and emotional attachments are implicated in activism, less attention has been given to the ways in which activism affects personal lives. This paper addresses this issue, drawing on interviews and focus groups with the Hong Kong's Umbrella Movement's active participants, bystanders and opponents to explore its consequences for family life. While those who were not involved in the movement articulated an acceptance of hierarchical family structures and their imposed silences, movement activists saw their experience of the occupation as enabling them to find a voice within their families. The Umbrella Movement, we suggest, has opened up a space for the reflexive exploration of personal life and raised the possibility of modifying Hong Kong family practices.

Introduction Since 1980 the United States has accepted more than 3.1 million refugees . Among refugees resettled in the US in 2019, 50.9% were of reproductive ages and 49% were female . The fertility of foreign-born US residents is about one third higher than their native-born counterparts. According to the Annie E Casey Foundation 23% of US births in 2019 were to foreign-born persons . Foreign-born persons in the US experience challenges with language, cultural differences, stigma, poverty, and access to healthcare. Refugees have often fled from their homes with little or no preparation, many spending months or years in a resettlement camp, where they may have been exposed to violence, disease, and the deaths of family members including children. They may speak little or no English upon arrival, and some lack literacy in their native language. The "healthy migrant effect" with regard to pregnancy is an unexpected finding that foreign-born women have healthier birth outcomes . Studies identifying the HME have found that the infants of refugees have fewer low birth weight births, lower premature births, and fewer infant deaths than infants born to US mothers . Most of the studies on the HME compare the births of US born and foreign-born persons, without distinguishing between refugees and other immigrants. Wanigaratne et al. examined the rates of severe neonatal morbidities among Canadian infants with regard to mothers' natality; SNM of infants of refugees to Canada , immigrants , and Canadian-born mothers differed modestly, with the infants of refugees and immigrants having fewer SNMs. A study that addressed parental natality found that foreign-born mothers and foreign-born fathers were each independently associated with fewer low birth weight births, compared with US born parents . Research on the protective factors involved in the HME is less robust than the studies on incidence, prevalence out outcomes cited above. Among the studies that address protective factors resulting in the HME are those that cite migrants' lower tobacco use , better oral health of migrants , and lower alcohol use . Also, migrants who arrive in the United States are likely to be healthy than similar individuals from their country of origin who did not have the health or stamina to make what is frequently an arduous journey. --- Research Question The study was designed to review the birth outcomes for infants of refugee, foreign-born, and US born women to use in program planning, education of obstetrical residents and other clinicians, and health education to refugees. --- Methods Setting Syracuse, a city in Central New York, has the third highest receipt of refugees of cities in the United States. There are two very active refugee resettlement agencies, who have welcomed new Americans to the city beginning with the mid-1970s arrival of Southeast Asians following the Vietnam war . The 2020 US census found 13% of the city population to be foreign born, although the census does not distinguish between voluntary migrants and refugees. Local healthcare providers have extensive experience caring for refugees and immigrants. Local healthcare providers follow the federal Office for Civil Rights mandated under Title VI legislation, to offer Limited English Proficient services to patients whose care provider does not speak their primary language . But even accurate translation of the patient-provider communication may fall short of full understanding. --- Research Design and Participants The study used secondary data of all births in Syracuse for 3 years , from the New York State Perinatal Data System , which compiles and documents prenatal data and birth outcomes from all birth hospitals in New York State. The births included in this study are all births to mothers residing in the city of Syracuse, NY, who gave birth in the city of Syracuse, NY. The SPDS is a de-identified population-based birth registry that captures pregnancy and birth information for quality improvement. The 3 years used for analysis 2017, 2018, 2019 were chosen in order to avoid potential bias resulting from the COVID-19 pandemic. The SPDS operates under New York State rules and regulations as follows: "All hospitals and freestanding birthing centers that provide perinatal health care services shall participate in the statewide perinatal data system. All live births shall be entered into the state perinatal data system" https:// regs. health. ny. gov/ conte nt/ secti on-40022-state wide-perin atal-data-system. We consulted with the staff of a local refugee resettlement agency, who advised that individuals from the DRC and Somalia were the most recently resettled during the 3-year period of the study . Mothers' natality was separated into three groups: US born, mothers, those born in the Democratic Republic of the Congo, and Somalia , and foreign-born women from other countries. As documented in the SPDS, the "other foreign-born mothers" were from 98 different countries. The rank order of their country origins were: Bahamas, Bhutan, Cuba, Kenya, Yemen, Sudan, China, Thailand, Nepal, Iraq, Vietnam, Syria, India, and Greece. The birth database does not distinguish between refugees and voluntary migrants; the group of "other foreign-born" women may include both refugees and voluntary migrants. Women from the DRC and Somalia may also include some voluntary migrants. --- Measures The variables documented in the SPDS are based on the women's disclosure during pre-natal care , depression during pregnancy, intentions of pregnancy, maternal race) coded as Black only, white only, Asian only, and other). Other variables are extracted from the antenatal and hospital delivery records. Self-reported tobacco use, alcohol intake, and illicit substance use were coded as dichotomous variables, participation in WIC , diabetes , gestational diabetes , employment , Medicaid . All data, including those about substance use are from the SPDS. Among births to single mothers in Syracuse-area birth hospitals, the baby's father is given the opportunity to sign the Declaration of Paternity on the morning after the birth. If the father does not sign, it is an indirect indication of less paternal emotional and financial support at the time of the birth . --- Data Analysis We compared births in three groups categorized by location of the mothers' births: US born, refugees from the Democratic Republic of the Congo and Somalia, and other foreign-born women from other countries. The main outcome variable in this analysis is low birth weight births, defined as births of less than 2500 g. Variables were compared with ANOVA for continuous variables and chi-square or Fisher's exact tests for discrete and dichotomous variables, as appropriate. In order provide sufficient data to understand the differences among the three groups in prenatal risk factors and birth outcomes, we include information about the labor and delivery characteristics, birth weight, gestational age, NICU admissions, and Down's Syndrome diagnoses. Multivariable logistic regression analysis was used to estimate risk of having a low birth weight birth among refugees and other foreign-born mothers relative to US born mothers. We examined maternal age, maternal race, maternal education, insurance status, maternal employment status, tobacco use, alcohol use, illicit drug use, participation in WIC, previous birth, pregnancy intention, and paternal acknowledgement as potential confounding variables. Variables were determined to be confounders if they were independently associated with country of birth, delivery of a low birthweight infant, and changed the odds ratio for country of birth and delivery of a low birthweight infant by at least 10 percent. The final multivariable logistic regression model includes all variables meeting the confounding definition and included maternal race, maternal education, insurance status, maternal employment status, tobacco use, and illicit drug use. --- Results There was a total of 5998 births in Syracuse in the 3 years , which were divided into three categories: births to US born , refugees from the DRC and Somalia , and other foreign-born women . As presented in Table 1, US born mothers had twice the percentage of teen births compared to births of other foreign-born mothers and almost 4 times as many as refugee mothers. Maternal race differed greatly in the three groups, reflecting their countries of origin. The refugees had over twice the proportion of non-high school graduates as the two other groups. All three groups had experienced substantial poverty, as measured by Medicaid insurance status, with nearly all of the refugees having Medicaid coverage. Multiple births were not statistically different among the three groups. None of the refugees used tobacco during the pregnancy, compared with 2.2% of the other foreign-born women, and nearly a quarter of the US born women. Very few women in any of the groups reported using alcohol during the pregnancy. More than one in five US born women reported illicit substance use during their pregnancy compared to 1.4% of other foreign-born women, and zero refugee women. Most women in all three groups had given birth previously. Among US born and other foreign-born women just over 10% were primiparous, compared with just over 5% of the refugees. Among the US born women 45.1% reported that their pregnancy was unintended , compared with 28.2% of the other foreign-born women, and 6.2% of the refugees. Self-reported depression during pregnancy was Table 2 shows that US born women had slightly over twice the prevalence of pre-pregnancy hypertension, compared with other foreign-born or the refugees. US born women also had twice the prevalence of pregnancy-related hypertension of the refugees, and over twice as much as the other foreign-born. Gestational diabetes, however, was similar in US born women and refugees, but almost twice as high among the other foreign-born. As presented in Table 3, the refugees had significantly fewer cesarean Sects. . Fewer refugees needed medical induction of their labors, maternal analgesia, or epidural/ spinal anesthesia. As presented in Table 4, the refugees had lower incidence of infants with LBW , and other foreign-born mothers , but higher macrosomia . Among full term infants, refugees and other infants born to foreign-born mothers had nearly 50% lower LBW than infants born to US mothers. Premature births also differed among the three groups, with 3.2% of infants of refugee mothers being premature, compared with 5.7% of infants of other foreign-born mothers and 10% of infants of US born mothers. NICU admissions were lower among infants of refugee mothers , compared with infants of US born mothers and infants of other foreignborn mothers . An exception to the generally more positive birth outcomes for refugee and foreign-born women is that the full-term infants of refugee mothers had much higher macrosomia than infant of US born or Number missing on counseled about during pregnancy variables varied. Chi-square or Fisher's exact tests when cells < 5. ANOVA used for continuous variables Table 5 presents the adjusted odds ratios of LBW among singleton births of the three groups, controlling for maternal race, Medicaid insurance as a proxy for poverty, and tobacco use. We examined age as a potential confounder by comparing logistic regression models with and without age included. Including age in the model did not change the effect estimates , so we did not retain age in our final model. As the table shows compared to infants born to US mothers , infants born to refugees had 58% lower odds of LBW and the odds for infants born to other foreign-born mothers were 41% lower. --- Discussion This 3-year secondary data analysis of birth outcomes in Syracuse supports the healthy migrant effect in that foreignborn women had fewer low birth weight births, premature births, and cesarean sections than their US born counterparts. Contributing to the paradoxical nature of those findings, prior to this birth more refugee women were identified as "high risk," compared to other foreign-born women , and US born women . Refugee mothers on average had less education, more poverty as indicated by Medicaid insurance coverage, and higher rates of employment. Yet, recent refugees from the DRC and Somalia had better outcomes on measures listed above than foreign-born women who had resided in the US for a longer time or US born women. Foreign-born Syracusans have higher fertility than US born residents, as documented by the crude birth rate of foreign-born Syracuse residents , compared to US born .1 This higher fertility is likely attributable to the foreign-born population being younger, with more individuals in the reproductive age group. Another factor is that many of the foreign-born mothers come from countries and cultures where fertility rates are higher than the U.S. Refugees from the DRC and Somalia have better birth outcomes than US born women or foreign-born women who arrived earlier. This may result from lower rates of high-risk behaviors that are linked to poor birth outcomes. Refugee mothers were significantly less likely to be teenagers. No refugee mothers reported use of tobacco, alcohol, or illicit substances during pregnancy. Refugee mothers appeared to be more likely to have support from the infant's father. Refugees had far fewer teen births, and no refugee women used tobacco, took illicit drugs, or consumed alcohol. In fewer than 10% of refugee and other foreign-born births the father did not sign the DP, compared with 26.5% of US born births. Previous research documented that when the father did not sign the DP, the rate of post neonatal mortality was 3.8 times that of infants in which the father was either married to the woman or he signed the DP . Intended pregnancy also differed greatly among the three groups, with nearly half of US born women reporting that their pregnancy was unintended, compared with 28.2% of the other foreign-born women, and 6.2% of the refugee women. Participation in WIC also differed greatly, with over 90% of the refugees, compared to 67% each of the US born and other foreign-born women. WIC participation during pregnancy was previously found to be associated with fewer LBW births . --- Limitations A limitation of the study is that the operationalization of refugees could exclude refugees from other countries, particularly those who arrived prior to 2017. Poverty and partner support are both examined by proxy , which introduces a limitation. An additional limitation is that the SPDS secondary database uses self-report for illicit substance and tobacco use, which is not as precise as would be using more objective biological measures. --- Data Availability The study used secondary data from the NY State Perinatal Data System. Code Availability Statistical analysis was done using SPSS. --- --- Author Contributions GG: identified the research questions and conducted the literature review. MF: conducted the quantitative analysis. SDL: assisted with the quantitative analysis and wrote the review of results. MS: advised on the obstetrical issues involved in the analysis and suggested variables and measures to include. CS: advised on the obstetrical issues involved in the analysis and suggested variables and measures to include. RAR: participated study design and in writing the manuscript. AS: responsible for obtaining IRB approval and for the research design. Also participated in the analysis. Funding Not applicable. --- --- Ethical Approval This study was approved by the Institutional Review Board of SUNY Upstate Medical University and was designated exempt. --- Consent to Participate Not applicable. --- Consent for Publication Not applicable. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author or other rightsholder; author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Objectives In Syracuse, NY among 5998 births in a 3-year period (2017)(2018)(2019), 24% were to foreign-born women, among whom nearly 5% were refugees from the Democratic Republic of the Congo and Somalia. The impetus for the study was to identify potential risk factors and birth outcomes of refugee women, other foreign-born women, and US born women to inform care. Methods This study reviewed 3 years of births (2017-2019) in a secondary database of births in Syracuse, New York. Data reviewed included maternal demographics, natality, behavioral risk factors (e.g., drug use, tobacco use), employment, health insurance, and education.In a logistic regression model controlling for race, education, insurance status, employment status, tobacco use and illicit drug use, compared to US born mothers, refugees (OR 0.45, 95% CI 0.24-0.83) and other foreign born (OR 0.63, 95% CI 0.47-0.85) had significantly fewer low birth weight births.The results of this study supported the "healthy migrant effect," a concept that refugees have fewer low birth weight (LBW) births, premature births, and cesarean section deliveries than US born women. This study adds to the literature on refugee births and the healthy migrant effect.The United States is a major receiving country for refugees. Because refugees have often experienced serious trauma and limited healthcare prior to resettlement, analyses of their birth experiences and pregnancy outcomes would help to guide care during pregnancy and post-partum.

If I can criticise on one aspect of the paper, but this is my personal opinion, I would be cautious with including governance and policy directly into the model instead of running --- Response to Reviewer Comments My comments are only minor and hopefully help improve the argument presented the paper. 1. While I agree that internal government structures are not well captured in ABMs , and while the paper discusses the need to improve them. I would have appreciated a little bit more discussion on how this might be done . # Response: We added the following section to approach your comment: Line 635-658: "Sketching the phases of a research project can help to operationalise the ideas discussed above as part of such an ABM development cycle. While using the example of international food trade, the first step could involve mapping relevant actors across different scales and levels, e.g. relevant ministries such as foreign affairs and trade for the decisions made in regards to international agreements and agriculture to capture changing policies that affect agricultural practises and crops grown; multi-national firms as they are especially relevant as price makers in the food sector, due to their big role in agricultural technology development to improve farming practices, and their influence on policies through lobbying; farming communities and associations as they represent the primary sector, receive and implement policies and the same time lobby governments. This would be followed by scoping interviews with representatives from the key actors to identify what dynamics they influence, and how they interact with other stakeholders. Further interviews could be undertaken with actors that have been identified as relevant by the first round of interviews and were not involved. Part of the interviews could involve questions aimed at mapping both actors and relationships between them. The second stage of the project could involve scenario-based workshops with key members of relevant stakeholder groups, where they will be presented with a future scenario , and their responses to different checkpoints captured in the scenario . Once this information has been collected and collated, the development of a meta-model for the behaviour of these actors could start by generating a general framework of responses for each actor based on their reactions to prompts or be informed by relevant theories from cognitive and behavioural sciences. Follow-up interviews could be organised with key stakeholders to fill the gaps or clarify specific reactions and therefore finalise the behavioural meta-model for the different actors." 2. Similarly, the role of stakeholders is discussed but little is given with respect to who specifically might the stakeholders and decision makers and why they should care about such models. # Response: We added and revised several paragraphs following your comment. Line 109-113: "Understanding the roles of multiple stakeholders such as political actors, resource users, citizens or agencies who may have direct or indirect influences and interest in decisions making is integral for understanding SESs across scale. The core proposition is that in a world that is increasingly connected and multi-scale, solutions that support policy design and decision making must be as well." Line 395-397: " input from domain experts and/or stakeholders ". 3. On another point, I was wondering if models have to consider many issues will they not become overly burdensome or complicated and what can we do to stop this from happing? In the sense I would hate to see a paper like: Lee Jr, D. B. . Requiem for large-scale models. Journal of the American Institute of planners, 39, 163-178. Appearing after ABM took the advice from this current paper. # Response: Thank you for this comment. We approached this with an additional paragraph which was influenced by comments of the retrospective paper of Lee . In this paper, Lee acknowledged that some of the stated criticisms of Lee et al. had been overcome by increased computing power during that time. Moreover, in our view, most of Lee's critique is based in misunderstanding of computation, and thus we further emphasised on the aspects of "ensuring transparency and reproducibility of model code and clarity of model purpose" in the added paragraph. Line 715-722: "The core proposition of this paper is that in a world that is increasingly recognized as being connected and multi-scale, solutions must be as well. This might lead to complex and intricate models, but perhaps the complexity of the real world requires use to embrace this in our modelling efforts. While large scale modelling has received much criticism in the past [151], most of these issues could be addressed by increasing computing power [152], and can be further overcome by ensuring transparency and reproducibility of model code and clarity of model purpose." 4. With respect to the comment above, something else that comes to mind. Specifically, if models need to be larger in the sense of numbers of agents, networks , processes etc. Will it be possible to run such models? The authors note "while increasing computing power enables us to simulate systems of interest in ever greater detail," but to develop a truly large scale model incorporating many of the elements discussed in this paper seems impossible without some sort of distributed / cloud architecture behind the simulation models . --- # Response: We expanded and revised the following paragraphs to acknowledge your comment. Line 507-513: "This may further require distributed, parallel computing systems, or server-/cloudbased network architecture to meet the high computational demands needed to complete simulations in a reasonable time as is quite common in climate change and hydrological modelling applications to date. On the other hand, it is not only computational power that might restrict model size; usability and user understanding which might 'self-restrict' the size of the model as well [67]. Line 578-590: " While available computing power enables us to simulate such cross-scale interactions in ever greater details, this can only be made possible using modular modelling structures such as are available in NetLogo, but more importantly will require some sort of larger-scale distributed computing systems rather than a single desktop or laptop. Where model run-time is longrunning but acceptable, then cloud-based approaches using platforms such as Microsoft Azure© or Amazon AWS© might be sufficient to allow for the multiple model runs needed for parameter space exploration or what-if scenario generation. Where models need to be accelerated even in single runs more traditional high performance computing architectures can be exploited with frameworks such as RepastHPC, which provides the ability to scale to very large numbers of agents in both gridded and networked configurations [135]. Some of the associated technical difficulties in dealing with this kind of large model in languages like Java are covered in [96]." Line 681-682: " and the development of models that can run fast enough in an individual or parallel-sense." 5. Also the authors write that to succeed "effort needs to involve multiple research groups across the globe, taking a multiplicity of approaches, but focused on producing models with substantial improvements in their capacity to simulate the socio-economic components of SESs," While I agree with this, I would also add that modelers should also consider sharing their models something that is still not commonly done in the ABM community and also in the SES community. # Response: We expanded and revised the following paragraph to acknowledge your comment. Line 350-364: "In agriculture, linking models of disease spread and mitigation procedures is accepted practice, as e.g. in the work of [89] that integrates a simplified individual-level model of the spread of potato late blight , in a landscape-level model of farmer's crop choice and management. First, the natural system was modelled. Then, farmer practices were added, both in the model and in interactive sessions with farmers [90]. Similarly in [78] an individual ecosystem model for tree growth provided a dynamic landscape for farmers to both harvest trees and clear land for crop growth. The modification of the soil permeability then fed a hydrological model for simulation of the subsequent change in the profile of flooding. Coupling of these models was achieved through access to the source code for each sub-model and re-writing them to form a common framework in which the space and timescales could be matched to the smallest appropriate for the whole model set. However, feedbacks from the environmental modification into farmer behaviour or forest dynamics from the altered pattern of flooding, and the potential effects of this downstream of the model catchment, either in terms of other residents, or on policy for forest conservation or flood management were not accounted for, despite a nominal model run time of hundreds of years." Line 457-459: "The software PCRaster can be drawn as an example that allows for dynamic and spatial-explicit modelling of SESs further allowing error propagation techniques such as Monte Carlo or Kalman Filter techniques." 7. With respect to new direction "Conceptual and methodological directions" could one not argue that many modelers are already aware that there are interactions between systems i.e. the notion of near decomposability , in the sense in the short term, the systems can be considered independent but in the long term they impact each other. I bring this up as I believe one needs to balance model complexity to the research at hand. --- # Response: We expanded and revised the following paragraph to acknowledge your comment. Line 566-571: "Many modellers are well aware that there are cross-scale interactions between systems which can considered independent but in the long term impact each other see also [12,13,15,16,19]. Hence the overall aim will be to balance model complexity and the simulated interactions between systems cross-scale to derive outputs that are meaningful and help to derive implications for decision making and policy design [133,134]." --- Introduction --- 73 The social-ecological systems concept describes the tight coupling of human and may not be easy to understand. How much these dynamics need to be incorporated in a given 172 model will depend on model purpose, but the complexities of variation across scale need to be 173 considered. For example, while social networks are commonplace in many agent models, 174 people will typically belong to multiple networks with different physical and social reach. The 175 interaction between these networks is likely to be of as much importance for some phenomena 176 as the networks on their own . 177 The anthropocentric nature of the ecosystem service concept has further re-focused attention 363 The implication we draw is that the last of the these three modes discussed above is really a 364 requirement rather than an option: since the systems modelled are complex and the relative --- Acknowledgements --- stakeholders in the process, decisions, and validation of model design; or it may be later in the 707 process, in using the results of a model to open up discussions with stakeholders, and/or even 708 using the model live to explore connections between assumptions, scenarios, and outcomes 709 [150143]. 710 711 --- Concluding remarks 712 The issues we have discussed here emphasize the need for ABM of SESs to include the 713 feedbacks that are implied by the presence of both multiple time and spatial scales. The core 714 proposition of this paper is that in a world that is increasingly recognized as being connected 715 and multi-scale, solutions must be as well. This might lead to complex and intricate models, Dr. Melvin Lippe is working as a senior scientist at the Thuenen Institute of International Forestry and Forest Economics in Hamburg, Germany. His research interest lies in the assessment of drivers and the multi-temporal patterns of land cover/land use change in tropical environments. He uses participatory approaches, GIS-and remote sensing techniques, as well as dynamic, spatiallyexplicit landscape models and spatial-role play games to explain the social-ecological complexity of landscapes as tools for decision making and policy advice. Dr. Mike Bithell is currently working as the Assistant Director of Research in Computing at the Department of Geography, University of Cambridge. His research interests lie in numerical modelling of spatially distributed systems, including fluid flow, atmospheric physics, climate and its interaction with ecosystems, changes in land use and socio-economic processes. Dr. Nick Gotts is trained in psychology and artificial intelligence. Until 1996, his research was mainly concerned with qualitative spatial representation and reasoning in humans, other animals, and computers. Since then he has concentrated on complex systems dynamics, particularly on agentbased social simulation, in the areas of land use change, energy demand, and social and informational networks. Since July 2012 he has been an independent researcher, particularly interested in ways of comparing agent-based models, and in the relationships between analysis and simulation in the study of complex systems. He is a visiting fellow at Manchester Metropolitan University. Dr. Davide Natalini is Research Fellow at the Global Sustainability Institute, Anglia Ruskin University in Cambridge, UK, where he co-leads the Global Resilience and Risk research theme. Davide is an interdisciplinary environmental social scientist and studies complex social-ecological systems, systemic risk, and sustainable development, with specific focus on environmental security and conflict. He implements both quantitative and qualitative methods in his projects. Dr. Peter Barbrook-Johnson is a UKRI Innovation Fellow and Research Fellow at the University of Surrey, UK. His research interests are at the intersection of environmental policy, social science, and complexity science, where he specialises in the use of agent-based modelling and participatory systems mapping of social and policy systems. Dr. Carlo Giupponi is a Professor at the Department of Economics of Ca' Foscari University of Venice and Dean of the Venice International University, Italy. His research activity focuses on sustainability science, valuation methods, the integrated assessment and management of natural resources, with main focus on water and agriculture, local to regional scales, and the interactions with global change drivers . Dr. Mareen Hallier is currently working a research associate at the Institute of Mathematics, Brandenburg University of Technology Cottbus-Senftenberg, Germany. Her research focusses on the statistical analysis of agent-based simulations of social-ecological systems. This includes for example the construction of models of reduced complexity as well as the identification of system regimes and hierarchical pathways of change between such regimes. Dr. Gert Jan Hofstede is Professor of Artificial Sociality at Wageningen University, Netherlands. His long-standing aim is teasing out the generic dynamics of human social behaviour. He co-authored the best-selling 'Cultures and Organizations: software of the mind". Gert Jan uses simulation gaming

Many thanks for your constructive comments, questions and feedbacks which we approached below. Our response represents the collective action of all fourteen co-authors with their various backgrounds and expertise in simulating social-ecological systems (SESs) using agent-based, spatiallyexplicit, or other forms of qualitative, and quantitative modelling. Our response is structured into: general comments, reviewer #1 and reviewer #2. We subsequently numbered the reviewer comments in bold, and responded to it individually as indicated by # Response. With sincere regards on behalf of all co-authors, The acknowledgement section was expanded due to several new funding agency requirements.  Line numbering was added to ease the review process.  Some sentences were additionally revised based on our internal editorial discussions:

Introduction Health promotion is a trend that emerged in the early 1980s, which has come to represent a unifying concept for those who recognize the need to change their ways and conditions of life in order to promote health. In this way, it represents a mediating strategy between people and their environments, synthesizing personal choice and social responsibility in health to create a healthier future. The basic resources for health are income, shelter and food [1]. The Ottawa Charter is considered the founding document of health promotion and was presented at the end of the First International Conference on Health Promotion, held in Ottawa, Canada, in 1986. It is a letter of intent that seeks to contribute to health policies in all countries, equally and universally [2]. Thus, it refers to a social right, inherent to the condition of citizenship, requiring public policies to guarantee the population's access to it. --- Materials and Methods --- Study Design This is a case study that used mixed methods-qualitative and quantitative-in order to obtain broader perspectives to achieve the proposed objectives and a closer view of the reality of the gymnastics modality as a municipal public policy. The Portuguese version of the World Health Organization Quality of Life-Abbreviated Version [7] questionnaire was used for all participants, and the focus group technique was used in two classes which were selected using the criterion of social classes in inverse situations. --- Research Subjects The participants were 239 women aged between thirty-five and seventy-four years old, who were students of the SEL's gymnastics classes and were randomly selected according to availability. Two classes were selected using the socioeconomic criterion of social classes in inverse situations to carry out one focus group session; then, twenty students belonging to the two classes were randomly selected. Inclusion criteria were as follows: students of an SEL gymnastics group, who were willing to participate in this study, completed the WHOQOL-bref questionnaire, signed the informed consent form, participated in the focus group, belonged to one of the two classes selected and attended the program assiduously . Exclusion criteria were as follows: students who did not sign the free informed consent form or who did not answer the entire questionnaire. The study flowchart is shown in Figure 1. tions and that the practice of gymnastics provides an improvement in physical abilities and biological indicators, directly reflecting the quality of life of the students, as well as a moment of leisure, which leads to long-term adherence. --- Materials and Methods --- Study Design This is a case study that used mixed methods-qualitative and quantitative-in order to obtain broader perspectives to achieve the proposed objectives and a closer view of the reality of the gymnastics modality as a municipal public policy. The Portuguese version of the World Health Organization Quality of Life-Abbreviated Version [7] questionnaire was used for all participants, and the focus group technique was used in two classes which were selected using the criterion of social classes in inverse situations. --- Research Subjects The participants were 239 women aged between thirty-five and seventy-four years old, who were students of the SEL's gymnastics classes and were randomly selected according to availability. Two classes were selected using the socioeconomic criterion of social classes in inverse situations to carry out one focus group session; then, twenty students belonging to the two classes were randomly selected. Inclusion criteria were as follows: students of an SEL gymnastics group, who were willing to participate in this study, completed the WHOQOL-bref questionnaire, signed the informed consent form, participated in the focus group, belonged to one of the two classes selected and attended the program assiduously . Exclusion criteria were as follows: students who did not sign the free informed consent form or who did not answer the entire questionnaire. The study flowchart is shown in figure 1. --- Quantitative Instrument To assess quality of life, the Portuguese version of the World Health Organization Quality of Life-Abbreviated Version [7] questionnaire was used, which consists of 26 questions, that is, 2 general questions on quality of life and 24 on the following domains: physical, psychological, social relations and environment. According to the sample size calculation, the minimum size required was 204 participants. For the calculation, the total population of 800 students who currently participate in the gym- --- Quantitative Instrument To assess quality of life, the Portuguese version of the World Health Organization Quality of Life-Abbreviated Version [7] questionnaire was used, which consists of 26 questions, that is, 2 general questions on quality of life and 24 on the following domains: physical, psychological, social relations and environment. According to the sample size calculation, the minimum size required was 204 participants. For the calculation, the total population of 800 students who currently participate in the gymnastics program was used, considering a margin of error of 5% and a confidence level of 90%. An increase of 20% in the minimum size was calculated, so 244 questionnaires were answered; however, 3 questionnaires were excluded before tabulation, as the students did not sign the free informed consent form, and 2 were tabulated but invalidated by the tool because the students failed to answer a whole page of the questionnaire. Thus, according to the size of the final sample of valid responses, 239 current students of the gymnastics program, aged between thirty-five and seventy-four years old, participated in the research. The data collection process was completed in 5 weeks, and the students were invited to participate in the research at the time of the gymnastics class. --- Qualitative Instrument The focus group technique is a qualitative method that employs interviews on a specific topic among a small group of people, usually homogeneous, called a focus group. In this technique, the researcher can gather information about several individuals in a single session [8]. Regarding the number of participants in the focus groups, the literature recommends between 6 and 15 participants [9,10]. Table 1 presents the background information of the focus group participants. In this study, one focus group session was carried out in two classes which were selected using the criterion of social classes in inverse situations. This criterion is a socioeconomic criterion that refers to two distinct groups with clear class divisions. This criterion was chosen since one group lived in a more privileged region; therefore, this division represented-if not an "inverse" situation-at least a "diverse" situation in terms of income, housing, mobility, education, access to health care, etc. We chose to carry out the focus group in only two gym classes, instead of all of the classes, since all of the classes were similar and conducted in the same way, as was the case in a previous study [11] . It was verified that the answers were similar in the different groups, meaning it would have been unnecessary to carry out the collection in all of the classes; therefore, this form of selection was chosen to also understand whether economic and social differences influenced the responses. Twenty students, aged between forty-four and sixty-five, who were randomly selected according to availability, participated in this stage. A pre-established script elaborated by the authors was used as a guiding scheme, with the main information to facilitate the progress of data collection, as well as practical guidelines for carrying out the focus groups. The questions used were divided into three sections: personal history of the students; gymnastics program; and general questions. The questions on the students' personal history related to the following: age and time of participation in the gymnastics group; practice of physical activity in childhood and adolescence; and current physical activity status. The questions on the gymnastics program related to the following: reasons for adhesion and adherence ; meaning for the students; and strengths and weaknesses. The questions on general issues related to the following: quality of life; sport and leisure facilities; and autonomy in performing physical activity. The sessions were recorded and later transcribed literally, ensuring the reliability of the data, and the students' names were replaced by numbers. A battery-operated Sony Recorder, model ICD-PX240, was used, and the audio was transferred directly to the computer via a universal serial bus cable. --- Data Analysis For the statistical analysis of the WHOQOL-bref questionnaire, the tool provided by Pedroso et al. [12], in their study entitled "Calculation of scores and descriptive statistics of the WHOQOL-bref using Microsoft Excel", was used. Even though the WHOQOL instruments have been widely used and disseminated, the use of SPSS software to calculate the results is a limiting factor, since it is a paid software. In this sense, aiming to eliminate this limitation, the authors built a tool using Microsoft Excel software to perform the calculation of the scores and descriptive statistics of the WHOQOL-bref questionnaire. It proved to be a reliable tool, which was tested in different versions of Microsoft Excel, and the results returned were compared with those returned by SPSS. The results obtained were identical to those of the syntax proposed by the WHO for SPSS software, in all tested versions of Microsoft Excel [12]. Regarding the analysis of the results of the focus groups, a content analysis technique was used, following Bardin [13], which presents the following criteria for organizing an analysis: pre-analysis, exploration of the material and treatment of the results. In this way, a first reading of the transcripts was performed. In sequence, the themes that appeared in the transcripts of the dialogues were initially categorized and ordered, followed by a specific analysis of each theme, re-reading and scoring the subthemes. The audios transcribed by the researcher are available in Appendix C of the dissertation [6]. --- Ethical Issues All participants were informed about the research procedures and gave their informed consent for inclusion before participating in this study. This study was conducted in accordance with the Declaration of Helsinki and the Regulatory Guidelines and Norms for Research Involving Human Beings of the National Health Council, and the protocol was approved by the Ethics Committee for Research in Human Beings at the USP and registered as protocol CAAE: 64651517.6.0000.5390. --- Gym Classes Gymnastics groups began in the mid-1980s and have existed uninterruptedly for about thirty years. To enroll, the student must be a female resident of the city of Valinhos aged between eighteen and fifty-nine years old and present a medical certificate authorizing them to practice physical activity. For people under eighteen, various sports are offered, and for those over sixty, there is a specific program for seniors, but students who turn sixty while attending the classes can continue in the class. Classes take place twice a week, lasting one hour each, at specific public spots in the city, with the intention being that the student takes the class in their own neighborhood or in a neighboring one but always close to home, preventing the need for a means of transport. This means that only the teacher travels to the location, allowing greater adherence and lower absenteeism. The classes include localized, functional and aerobic exercises, circuit training, and stretching and flexibility exercises. There are also thematic and commemorative classes, walks, open classes with all groups, a June party with square dancing, an end-of-year get-together with a secret friend game and excursions. --- Results Initially, in this section, the results of the analysis of the WHOQOL-bref questionnaire will be presented, with the aim of characterizing the quality of life of gymnastics students and what quality of life means for them. Next, the results of the analysis of the two focus groups will be presented, and each subtopic corresponds to one of the main subjects approached in its realization: physical activity practices in childhood and adolescence; current physical activity practices; reasons for joining the gymnastics program; reasons for adherence to the gymnastics program; meaning of participation in the gymnastics group; strengths and weaknesses of the gymnastics program; public sports and leisure facilities used for physical activity; physical activities performed autonomously. --- Characterization of the Gymnastics Students' Quality of Life Table 2 presents the descriptive statistics of the WHOQOL-bref questionnaire for all of the gymnastics students, referring to the tabulation results of all valid questionnaires. Below are, separately, the results of the gymnastics students from Neighborhood A and Neighborhood B, with Neighborhood B being socioeconomically more favored than Neighborhood A. In the tables, the scales are presented in scores that vary from 0 to 20, and on graphs from 0 to 100. The analysis of Table 2 demonstrates that the greatest amplitude in the responses occurred in the social relations domain, with the smallest amplitude in the psychological domain. The highest averages occurred in the physical and QoL self-assessment domains , while the lowest occurred in the environment domain . The general average for gymnastics students was 15.23. It can also be observed that, in descending order of the scores, the sequence was: physical, QoL self-assessment, social relations, psychological and environment. Table 3 presents the descriptive statistics of the WHOQOL-bref questionnaire for gymnastics students from Neighborhood A. It can be seen that the greatest amplitude in the responses occurred in the environment domain, with the smallest amplitude in the physical domain. The highest averages occurred in the QoL self-assessment and physical domains , while the lowest average occurred in the environment domain . It can also be observed that, in descending order of the scores, the sequence was: QoL self-assessment, physical, social relations, psychological and environment, similar to the general scores of the students. However, the overall average for these students was 14.84, slightly below the overall average for the students presented above. Table 4 presents the descriptive statistics of the WHOQOL-bref questionnaire for gymnastics students from Neighborhood B. It can be observed that the greatest amplitude in the answers occurred in the social relations domain, with the smallest amplitude in the psychological domain, similar to the general results of the students. The highest averages occurred in the physical and QoL self-assessment domains , while the lowest occurred in the environment domain , similar to the general results of the students with only the values of the scores changing. It can also be observed that, in descending order of the scores, the sequence was: physical, QoL self-assessment, psychological, social relations and environment. However, the overall average for these students was 15.51, slightly above the overall score presented above. According to the analyses presented, in the three cases, the domain with the highest score was the physical domain, while the domain with the lowest score was the environment domain. Table 5 presents a comparison of the averages of the WHOQOL-bref descriptive statistics among all the subjects, Neighborhood A subjects and Neighborhood B subjects. It can be observed that the average of all research subjects was 15.23, while that of Neighborhood A was slightly lower and that of Neighborhood B was slightly higher, which also had the highest average for QoL self-assessment. This indicates that the socioeconomically favored group obtained an average quality of life above the overall average of the groups, while Neighborhood A obtained a lower average than the overall average. This demonstrates that, although the difference in means is small between the groups, socioeconomic differences are also reflected in the individuals' quality of life, which relates to the social determinants of health. These are the non-medical factors that influence health outcomes, the social conditions in which people are born, grow, work, live and age, and the broader set of forces and systems that shape the conditions of daily life. These forces and systems include policies and economic systems, development agendas, social norms, social policies and political systems. They have an important influence on health inequities, which are the unfair and avoidable differences in health status observed within and between countries. In this sense, health and disease follow a social gradient: the lower the socioeconomic position, the worse the health [14]. Regarding the meaning of quality of life for the gymnastics students, in Neighborhood A, the main answers were good relationship with family members and people around them and the practice of physical activity, but there were also four mentions for the practice of physical exercise, as well as having a good relationship with the people around them, including the importance of a good family coexistence, to the quality of life. There were also two citations for leisure being important for quality of life, having good physical health and healthy aging, as well as citations linking the concept with good nutrition; however, they did not mention the term healthy eating. This term relates to the condition of being financially capable of having a good diet, in the sense of having purchasing power to buy healthy and balanced food. There was also only one answer for the relationship between quality of life and financial situation, such as having a job and income that makes it possible to buy a home, the right to housing, healthy habits, access to education and independence in the activities of daily living. As for Neighborhood B, it can be seen that the main responses were healthy eating , physical activity and being in good health . Regarding healthy eating, one of the students also mentioned the possibility of eating fruits and vegetables, in the financial sense, and there was only one citation for the following: feeling good about oneself, praying, meditating, the population's access to fundamental rights, such as health and education, having independence in activities of daily living, leisure, everything that is good for the individual and having the will to do everything . The interaction between living conditions and health has been widely indicated throughout history [15], and the study of individuals' quality of life has become a prominent topic in contemporary society; however, research involving this subject should consider that this is a complex topic, because it involves objective and subjective aspects, conditions and lifestyles, as well as multidimensional factors [16]. As observed in the responses, it is a broad concept, and as it also has an individual parameter, it can vary from one individual to another, which explains some of the different responses within and between the groups. However, even with different answers, both groups essentially agreed on the importance of physical activity to the quality of life, as well as healthy eating, good health and leisure activities. --- Physical Activity Practices in Childhood and Adolescence In this section, the physical activity practices in childhood and adolescence of the gymnastics students from Neighborhood A and Neighborhood B are presented in the Table 6. --- Invented games Hurricane , jumping across the river with a pole, floating in floodwaters inside a basin with another person pushing, bean straw hut and bonfire. Jumping in the river with a bamboo pole and going down the hills sitting on cardboard. --- Physical activities related to nature Swimming in a river, horseback riding, climbing trees and fishing with a sieve. Climbing fruit trees. --- Other physical activity practices Physical activities related to festivities: participating in ball , June parties, carnivals, samba and matinees. Invented toys: playing with clay, making little animals by putting sticks in chayote and making dolls out of green corn on the cob. Labor activities: drawing water from the well and carrying firewood. Walking to and from work. Walking to and from school. Walking to and from school. Playinghouse. Riding a bicycle. --- Swing . Swimming in the club's pool. Rodeo . Dancing. In Neighborhood A, the students showed excitement when telling their life stories, but some expressed a certain sadness when remembering what they had lived through in childhood, especially in relation to the home environment; at various times, they identified or sympathized with each other's stories. Several students reported that they almost did not play because their father was very strict, they were beaten often, they started working very early, they got married very young and they had children when they were still teenagers. The issue of access to study was also a striking point in this neighborhood, highlighted by the impossibility of continuing their education due to the need to work imposed on the students by their families. In the responses of several students, it was mentioned that they only studied until the fourth year of elementary school. It is also important to point out that, in some cases, the same game was cited by some students as a school game-generally more like a playground game than school physical education, as several students did not take this subject, and some studied for only a few years-and by others as a game played in the street. Work activities, such as drawing water from the well and carrying firewood and walking to and from school were also cited as physical activity, in addition to participation in festivities mainly due to the practice of dancing on these occasions; however, some students reported that they participated hidden from their fathers. They also mentioned that it was rare for someone to have a bicycle at that time, so those who had one would share it with their siblings, cousins or friends. Regarding Neighborhood B, the students also showed excitement when telling their memories of childhood and adolescence, but with slightly more excitement and happiness, with some reporting that their childhood and adolescence were good. Unlike in the previous group, none of the students mentioned that their childhood was sad, that they would rather not talk about that period, that they almost did not play because they started working too early, that they had to drop out of school because of work or that they got married very young. According to their responses, these students also had a greater opportunity to study, and several reported that they participated in physical education at school, that the teachers were demanding, as it was during the military period, and that the physical exercises were so rigorous that some students even felt sick in the classes. The issue of male chauvinism in the ban on playing soccer was also mentioned, but to a lesser extent, and in this case, it happened in the school physical education program, as it was during the military period. However, Student 17 reported that she played soccer in a team that represented her city, as well as playing in championships outside her municipality about forty-five to fifty years ago, which was not common at the time, demonstrating her belonging to a socioeconomically privileged group. Student 14 reported that she had been a member of a club since she was eight years old, and that she used the pool quite often, as it was close to her residence. Additionally, she also mentioned that she used to ride a bike frequently. In Neighborhood B, the students also commented that it was difficult for someone to own a bicycle, as bicycles were very expensive at the time, but they did not comment on sharing a bicycle. Likewise, some students carried out work activities in the fields, but did not report them as physical activity. Furthermore, activities related to festivities were not mentioned, but some toy constructions were mentioned, which did not occur in the previous group, as well as school physical education during primary school. --- Current Physical Activity Practices Regarding current physical activity practices, in addition to gymnastics classes, in Neighborhood A, the most common practices were walking and weight training , but there were also two mentions for playing with grandchildren, Zumba, walking around town, water aerobics, dancing at senior dances and housework, and just one mention for swimming, child care, Lian Gong, walking a lot during work, running and walking intervals, and riding a roller cart. An interesting finding is that housework and taking care of children were mentioned as physical activity, due to the physical fatigue they cause. Regarding Neighborhood B, the most common practices were walking and ballroom dancing , but there was also a citation for localized gym-nastics and stretching following a cassette tape, running, climbing trees, riding a bicycle, bodybuilding, Zumba and dance fitness. Thus, the main physical activity performed by the students in both groups, with the exception of gymnastics, was walking. --- Reasons for Joining the Gymnastics Program Regarding the reasons for joining the gymnastics program, in Neighborhood A, the main reasons were medical advice and invitation from members . Two students mentioned that they signed up for both reasons together, and one student did not respond. Student 7 reported that, in addition to her friend's invitation, she wanted to improve her daily life activities, and Student 11 reported that she likes to perform physical activity. Student 9 mentioned that she returned to the group after a period of absence, because she loves the teacher so much; this affection was a remarkable observation during the focus groups. Regarding the students from Neighborhood B, the main reason for joining was invitation from members . Only Student 18 started the class due to medical advice. Student 16 tried bodybuilding but did not like it. Student 14 aimed to improve her physical and psychological health. Student 13 was one of the citizens responsible for starting gym classes in the neighborhood in 1996, which demonstrates the social control of public sport and leisure policies in the city of Valinhos. --- Reasons for Adherence to the Gymnastics Program Concerning the reasons for permanence, analyzing the individual responses from Neighborhood A, it was possible to verify that the reasons most cited by the students were the enjoymentof participating in the group , having fun during classes and the possibility of leaving home . There were also three answers for taking care of one's health, feeling less physical pain and making friends. It should be noted that there were no results related to aesthetics or weight loss. Most responses referred mainly to the gym class as a sociability space, that is, a moment of meeting, exchanges or social contributions between friends inthe group. This was the main reason for the students' permanence in the gymnastics group in Neighborhood A. The answers related to this idea were the enjoyment of participating in the group, having a laugh or having fun during classes, the possibility of leaving home, making friends, being well received by the teacher and the students in the class, feeling less ashamed to speak in public after participating in the group, having a moment of relaxation, enjoying the festivities during classes, leaving the routine at home, taking time to meet friends, the possibility of having friendships, enjoying the friendship of the teacher and the students, the possibility of talking to other people and learning new things. Some answers related to health, both physical and psychological, were also presented, and one student reported that the place is easily accessible, as it is close to her home. The responses related to physical health were feeling less physical pain, feeling less cramps during sleep, feeling less physical fatigue and indisposition in everyday life, improving conditioning and disposition in daily activities, especially domestic ones, maintaining the amplitude of movements of the joints and postural correction. With regard to psychological health, the responses were feeling good about participating in classes, it is good for the mind and soul, feeling good about participating in the group, stress control and the possibility of working on emotions. The students also showed great affection for the teacher. With regard to Neighborhood B, the reasons for adherence most cited by the students were liking the teacher , enjoying participating in the classes , liking the group and liking the friendships . There were also three answers for providing well-being and enjoying trips and tours, and two answers for feeling less physical pain. In this group, the students also showed great affection for the teacher. It is important to point out that in this group, there were also no results related to aesthetics or weight loss, and, as in Neighborhood A, most of the responses referred mainly to the gym class as a sociability space. This was also the main reason for the adherence of the students to the gymnastics group in Neighborhood B, and the answers related to this idea were that it is already part of life, liking the group, the friendships, the trips, the tours, the festivities during the classes and the moments of exchange. Some responses related to health were also presented, such as the possibility of taking care of the body, health and well-being. With regard to physical health, it was mentioned that the students felt less pain in their body, and in terms of psychological health, an improvement in depression and well-being was provided by the classes. To finalize this question, it is important to report that in both focus groups, the main reason for the adherence of the students was the understanding of the gym class as a sociability space, transforming the participation in the classes into a moment of meeting and exchanges , as well as the possibility of leaving home, making new friends, talking to different people, having fun and belonging to a group. For some, this represents the only possibility since most of the students are housewives. In the background, they also aim to take care of their physical health-mainly to feel less body pain-and psychological health. In addition, it can be seen that participation in gym classes represents a moment of leisure, and this is also an important health promotion strategy. --- Meaning of Participation in the Gymnastics Group Regarding what it means for students to participate in the gymnastics group, analyzing the individual responses from Neighborhood A, it was possible to verify that the main responses were the group means family , and the group means everything in the life of the student . There were also two answers for the following: possibility of leaving home, making friends, meeting friends, having fun and taking care of one's health. Additionally, there was just one answer for feeling better physically and psychologically, a moment of relaxation, the possibility of participating in a group, having friends, feeling less shy about interacting with other people, happiness at being able to talk to other people, a moment of laughing and learning new things. Regarding Neighborhood B, the main answer was that participation in the gymnastics group means a moment of exchanges and sociability among the students, which helps in promoting health, especially psychological health. For example, Student 20 reported that their depressive condition was improved by participating in the gymnastics group. There was also only one answer for the possibility of meeting different people and liking them, a moment of meeting the other students, relieving stress and occupying one's mind. The group means a lot because of the friends and everything that happens in the classes, such as trips, and because it improves their health and provides a family, becoming a part of the student's life . For the students, there is nothing better; the group is very good because the students and the teacher are nice, and it provides the possibility to explore one's "inner self", and improves depression, emotions and life itself. To finish this point, it can be said that for the gymnastics students from Neighborhood A, the group represents a family, and for the students from Neighborhood B, the group offers a moment of sociability. In both groups, the classes positively influence quality of life, mainly through the psychological and social benefits provided by the practice of grouporiented physical activity, which is directly related to the reasons for their permanence in the gymnastics group. In view of this, in addition to physical health, it is possible to state that the motivations for the practice of physical activities involve the daily interactions made possible by new collegial relationships and even friendships established throughout the activities, the search for well-being and for new knowledge, and the perspective of improving the quality of life through established bonds and shared experiences [17]. --- Strengths and Weaknesses of the Gymnastics Program With regard to the strengths of the gymnastics modality, as a municipal public policy of the SEL, the main positive point reported by the students from Neighborhood A was that the gymnastics class was addressed to them, which they appreciated, considering it good, active and appealing, along with the well-being provided by its practice. There were also two answers for the friendships among the students, and one answer for seeing results, an easily accessible place because it is close to home, the teacher's dedication and studies, the possibility of healthy aging and changing one's mind when participating in the classes. With regard to the gymnastics group in Neighborhood B, the main positive point reported by the students was the teacher, where they praised her professionalism, her dedication, the level of demand and the intensity of the class. Two students also mentioned the friendships among the group. As for weaknesses, the main issue addressed by the students from Neighborhood A was the desire for more class days per week. They also mentioned the lack of a substitute duringthe teacher's vacation period. Meanwhile, three students stated that there were no weaknesses. In Neighborhood B, the main answer was that there were no weak points. However, three students mentioned that they would like more days of class per week, and another three students reported that the weak point was when they themselves missed class. --- Public Sport and Leisure Facilities Used for Physical Activity With regard to public sport and leisure facilities, the students from Neighborhood A agreed that there is good will on the part of the municipal administration and that there are many good classes taking place but that the city needs more public facilities for autonomous practice, including spaces for walking and bicycle lanes. In addition, it was reported that there are public spaces that are underutilized for classes, such as sports courts that need repairs. In Neighborhood B, it was verified that the public facility most used by the students is the Worker's Leisure Center , a park that has an indoor and outdoor walking track, with mileage tracking, facilitating the practice of walking, as well as a runway. It also has an indoor bicycle lane, an outdoor gym and a functional outdoor gym for the elderly, and it is close to their homes. Of the eight participants in the focus group, five declared that they use it frequently. One student mentioned Taquaral Lagoon Park in Campinas , and another reported that she sometimes goes to the 500 Years Square where it is possible to walk and use the outdoor gym. They also mentioned Washington Luiz Square and two other outdoor gyms although they rarely attend them, preferring the CLT due to its structure and greater proximity. As for bicycle lanes, the students commented that they would be interesting for the city, but it is necessary to first make drivers aware of respect for cyclists, and it may also not be advisable throughout the municipality as the city has many hills which would make commuting by bicycle difficult in these locations. In addition, they reported the existence of only one public swimming pool in the city, which entails a long waiting list and great difficulty in obtaining a vacancy; however, they mentioned the existence of a project to build a public swimming pool in their neighborhood. Finally, the facility most used by Neighborhood B is the CLT, which guarantees not only the practice of physical activity but also a moment of leisure in a green area within the city. Urban parks, as important leisure spaces, have their current configuration due to the intense changes that cities experienced from the 19th century onwards, with the processes of industrialization and urbanization. They contain elements of the countryside, appearing as refuges in the city and enabling the urban society to escape from the hardships of the industrial city [5]. --- Physical Activities Performed Autonomously Regarding the practice of physical activities performed autonomously by the students, that is, not guided, in Neighborhood A, the most mentioned activity was stretching , followed by postural correction and sit-ups . There was only one citation for teaching gym class exercises to an adult son and doing home gymnastics with a granddaughter. As for Neighborhood B, the main citation was postural correction , followed by stretching , using the CLT outdoor gym and running . In this way, in both neighborhoods, the main physical activities carried out autonomously by the students were stretching and posture correction in everyday life, both in the practice of physical activities and in carrying out housework and work activities, with the aim of feeling less pain in the spine. --- Discussion This study's main objective was to understand the reasons for students' adhesion and adherence to gymnastics groups, in order to understand why this public policy has been taking place uninterruptedly for over thirty years and to characterize the students' quality of life. The secondary objectives were to identify what this participation means for them, to understand the strengths and weaknesses of the gymnastics program, to identify what quality of life means for the students, and whether they use other sport and leisure facilities or carry out other guided or autonomous physical activities. When assessing the quality of life of the SEL gymnastics students, it can be seen that the socioeconomically favored group obtained an average quality of life above the general average of all the groups, while Neighborhood A obtained a lower average, demonstrating that social and economic differences can also be reflected in quality of life. It was also possible to verify the difference in the meaning of quality of life between the two classes and between students of the same class. This agrees with the WHO, which defines quality of life as an individual's perception of their place in life, in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns. It is an extensive and complex concept, which encompasses physical and psychological health, the level of independence, social relationships, personal beliefs and relationships with the characteristics of the environment. This definition reveals criteria which state that quality of life refers to a subjective assessment, with positive and negative dimensions, and that it is rooted in a cultural, social and environmental context [18]. Likewise, it reflects the perception among individuals that their needs are being met, or that opportunities to achieve happiness and self-fulfillment are being denied, regardless of their physical health status or social and economic conditions [18]. Additionally, as such, it can be said that the concept of quality of life is subjective, complex and multidimensional [9], as observed in the students' responses in the focus groups. During the realization of the focus groups, it was also possible to observe these socioeconomic differences between the two studied groups, in the students' answers and mainly in their memories of childhood and adolescence, both in the reports of what family life was like, and in access to study and physical activity and leisure practices. However, despite these differences reflecting opportunities to experience physical activity from childhood to the present moment, including access to public sport and leisure facilities close to the residence, the reasons for joining and adhering to the gymnastics group were similar in both neighborhoods. Additionally, when carrying out the two focus groups, the need for the participants to talk about life episodes that were not foreseen in the script was observed; therefore, through the present study, it was possible to give voice to these women who felt welcomed by the researchers. Regarding the reasons for joining, it can be said that in both groups, the students joined primarily due to the indication of practicing students, whether they were relatives, friends or neighbors, and due to medical indications. Afterwards, it was possible to understand that, when starting to attend gym classes, their initial objectives were contemplated and they were faced with a new reality, that of belonging to a group and of perceiving the gym class as a space of sociability. For the students, the group became a moment of encounters, interpersonal relationships and social support, which led to low turnover in the classes and the students' permanence in the group for many years, regardless of their social class. These findings coincide with other studies, showing that, currently, the reasons for the practice of physical activity, such as sport, isthat it is part of the social life of individuals and, among the countless biological benefits provided, is increasingly sought after for the sensation of pleasure and well-being it provides. Health, as a concept, is also based on human and social sciences [19]. However, the complexity of health is undeniable, regardless of the perspective from which it is approached [20]. A study carried out by Meurer, Benedetti and Mazo showed that health, pleasure and sociability are the main reasons for staying in a physical activity program [21]. In the same sense, Conceição et al. [22] analyzed the reasons for joining and permanence in a physical activity program for 112 elderly people. The most cited reasons for joining the physical activity program were medical indications; to maintain/improve health; and the need to do physical activity . With regard to the reasons that led to the elderly participants' permanence in the physical activity program, the following stood out: feeling good by regularly practicing physical exercise; their health improvement/maintenance; and the relationships/social coexistence that the program makes possible. However, in one point, their study differs from the present study, because the general indication from friends, neighbors and/or family was one of the least cited reasons for admission, contrary to our finding. However, in a study by Cardoso et al. , it was concluded that the main reason for joining the physical exercise program was the relationship aspect, that is, invitations from friends and/or family members [23]. Therefore, a new study would be necessary to understand specifically why the indications of students already participating in the program have such an influence on the adhesion of new students. We believe that this may be related to what it means for the students to participate in their gymnastics group. With regard to the meaning of participating in the gymnastics group, the results demonstrated that for the students in the Neighborhood A gymnastics group, the group represents a family for them, and for the gymnastics students of Neighborhood B, the group represents a moment of sociability. In both cases, it is worth concluding that the group has a positive influence on the quality of life, through the psychological and social benefits provided by the practice of group-oriented physical activity. In the same way, it was verified that these meanings are directly related to the reasons for their adherence in the gymnastics groups. As presented by Kim, Lee and Kim , several existing studies have examined the effects of exercise. Among them, it can be said that group exercise programs improve social interaction, and exercise improves cognitive functions and depressive symptoms [24]. According to Silva et al. , among the differences in the quality of life of people who practice physical activities compared to those who do not, there are not only physical health aspects but also psychological and cognitive aspects [25]. According to Conceição et al. , coexistence and social relationships stood out as relevant factors for continuity in the physical activity program [22]-in other words, sociability. For Mathias et al. , in the practice of physical activity, the sociability dimension concerns the opportunity to live with friends and obtain new friends [26].Another point observed during the students' responses was that gym classes also mean a moment of leisure. In this sense, leisure is presented as a human need which the students satisfy through their participation in classes. Combined with the fact that this moment entails benefits in the quality of life, it can be said that leisure is also configured as a health promotion strategy. Physical activity, as a phenomenon oriented towards the subject's achievement, seems to be a domain in which society has gained greater awareness, being associated with what is favorable and consistent with the highest aspects of the individual's quality of life, that is, their well-being. In contemporary society, there is also a very marked connection between physical activity and leisure, or a culture of the body in leisure activities, which cannot be dissociated from the concept of health or from the set of references of quality of life and well-being associated with it [27]. As presented by Batista, Ribeiro and Nunes Junior , when a practice is chosen, experiences that are pleasant and more suitable for the use of free time are chosen. In addition, the choice of a certain experience not only achieves biological benefits, such as physical conditioning and regulation and strengthening of metabolic systems, but also contributes to other social determinants, such as sociability, balanced use of time and access to culture, which will certainly trigger results in maintaining health as a whole [28]. In addition, with the aim of expanding the offer of physical and leisure activities in the city, it is important to mention the possibility of associating oriented programs with the stimulation of autonomous physical activity, through programs that encourage these practices, as well as the improvement of existing sport and leisure facilities and the availability of new facilities, in different locations in the municipality, in order to expand the reach of the population, which will allow the expansion of health promotion actions in the city and greater democratization of spaces and practices. Together, there must be a constant evaluation of municipal public policies, so that they are rethought, adapted or replaced, when necessary, with a focus on the best possible service for citizens in projects and programs of the various secretariats, considering the quality of life as a broad and multifaceted concept, which also demands intersectional actions and interdisciplinary projects. Another important point that cannot be forgotten is that the practice of physical activity also improves mental health, which relates to a significant current need of humanity, mainly due to the increase in or worsening of cases of mental disorders caused by the COVID-19 pandemic [29][30][31]. In particular, in developing countries such as Brazil, public health systems are overwhelmed, as reported by Bastos et al. [32], and patients are often on a long waiting list for health care. Therefore, it is important that other integrative and complementary practices, such as physical activity, be associated both in the prevention and treatment of comorbidities, as an alternative to relieve the public system and guarantee greater access to health for the population since public policies are essential for social rights to be contemplated, hence the importance of studying and discussing public policies on physical activity and their relationship with the expansion of health promotion actions through active behavior. The limitations of this study are as follows: the sample consisted only of women; a control group was not used to compare the results of the WHOQOL-bref questionnaire, since the main focus of this study was to analyze the reasons for joining and adherence to the gymnastics program; a comparison of these motives between the most recent students and those who had been participating the longest was also not carried out; this study also presented many objectives to be achieved, since the intention was to understand all the reasons that have led to the gymnastics program existing for so long, even though it is a municipal public policy, which is not very common in Brazil. However, despite these limitations, we believe that this article presents relevant data to serve as an example for the creation of future public policies, mainly by demonstrating that the practice of physical activity not only promotes benefits in physical health but also represents an important strategy for the expansion of health promotion actions. --- Conclusions This study proved that the practice of gymnastics as a municipal public policy led to an improvement in the physical and mental health of the students, as well as positively influencing several facets of their quality of life, which further emphasizes the importance of public policies as catalysts for health behavior and guarantees of constitutional social rights, such as access to health, sport and leisure. These policies need to be inclusive, democratic and designed according to the needs of those served; therefore, it is essential to understand the reasons for adhesion and adherence to a program, in order to improve the program or use it as an example for the creation and elaboration of future policies or programs. In addition to reducing sedentary behavior and improving biological indicators, which are extremely important for health, mainly because they prevent chronic non-communicable diseases, which overload health systems, physical activity also provides social and psychological benefits, which positively influence the quality of life of the practitioners. This also enables physical activity to be characterized as a health care strategy, from a biopsychosocial perspective, which is a topic that needs to be addressed in undergraduate courses, continuing education courses and postgraduate courses in the field of physical education. --- Data Availability Statement: Not applicable. ---

Background: The present study analyzed a gymnastics program offered by the Department of Sports and Leisure at a Brazilian city hall, representing an example of an effective and consolidated public policy. Main objective: The objective of this study was to understand the reasons for enrollment/joining (adhesion) and permanence (adherence) among female students in gymnastics groups in order to understand why this public policy has been taking place uninterruptedly for over thirty years and to characterize the quality of life of female gymnastics students. Methods: This is a case study that used mixed methods. As a quantitative method, the Portuguese version of the World Health Organization Quality of Life-Abbreviated Version (WHOQOL-bref) questionnaire was used. As a qualitative method, the focus group technique was used. Thus, 239 women aged between thirty-five and seventy-four years old, who were students of the gymnastics program, participated in this research by answering the WHOQOL-bref questionnaire. Two classes were selected using the socioeconomic criterion of social classes in inverse situations to carry out one focus group session; then, twenty students belonging to the two classes were randomly selected. Results: It was verified that the public policy studied has led to an improvement in the quality of life of the students, with the benefits extending beyond physical health, and that in both groups adhesion occurred mainly due to the indication of students who already practiced, or due to medical indications. It was also found that, in both groups, the main reason for long-term adherence was the understanding of the gymnastics class as a space for sociability and a moment of leisure. Conclusion: Physical activity is an important health promotion strategy. In addition to its biological benefits, which are necessary for the prevention of chronic non-communicable diseases, it also improves health and quality of life through social and psychological benefits, characterizing it as an effective health care strategy from a biopsychosocial perspective.