Datasets:

mimiklee

/

masterthesis_clean_text_20k

like 0

Community based participatory research has commonly been utilized in low income communities and communities of color to address the complex health issues they face . The Federal Agency for Healthcare Research and Quality defines CBPR as "a collaborative … approach designed to ensure and establish structures for participation by communities affected by the issue being studied, representatives of organizations, and researchers in all aspects of the research process to improve health and well-being through taking action…" . A core attribute of CBPR is the academic-community partnership in which both collaborate in an equal partnership to identify and address community health issues and to share knowledge and expertise with one another to further action and change within a community . This partnership must cross social, economic, political and cultural barriers to reach goals and become sustainable . CBPR has been commended as a process that facilitates community capacity building in terms of health promotion . Poole indicated that "the purpose of capacity building is to foster conditions that strengthen the characteristics of communities that enable them to plan, develop, implement, and maintain effective community programs." Capacity building encourages community members to address their own health issues and allows health practitioners and organizations to assist in the process , thus promoting sustainable solutions through community empowerment and leading to improved health outcomes . Evaluation and measurement of capacity building within community projects has posed challenges to researchers which include, but are not limited to, a lack of valid and reliable instrumentation, ever evolving community capacity constructs, and length of time required for community capacity building . However, evaluation and measurement of capacity are critical in order to better define constructs of and track progress toward building community capacity and consequent health outcomes, as well as to understand elements and processes that contribute to capacity building. Various community capacity frameworks have been described which include a number of common constructs, among them resources, leadership, asking why, communication with community members, partnerships, and social or external networks . Qualitative methods, including semi-structured interviews and focus groups with academic representatives, community leaders and members, in addition to the case study approach, have been extensively used to assess community capacity . A few studies have employed a mixed methods approach to evaluate community capacity, an approach that is considered to offer promise in better understanding its role in health promotion . --- Background --- H.U.B City Steps is a CBPR walking intervention involving a collaboration among a public university, a local city government, and community members in a southern small city. A primary aim of this intervention is to lower blood pressure in participants who are drawn primarily from the African-American population of the community. Elements of the intervention include empowerment, culturally tailored health education, and motivational interviewing to identify and track progress toward personal health goals. As a CBPR project, a primary aim is to build community capacity to promote health through physical activity and nutrition. Aspects of the project that foster community capacity and therefore attainment of this aim include: 1) inclusion of staff and faculty with a vested interest in the community; 2) involvement of project staff, a community advisory board , local city government, local agencies, and project participants in program development, implementation, evaluation, and overall guidance of project activities; and 3) development of participants' skills and knowledge of health promotion to attain individual and community change. These are accomplished via core activities including 1) recruiting and training community walking coaches a) to serve as role models and champions who will motivate their fellow residents to engage in physical activity and choose better eating habits and b) to establish walking groups which will provide social support to program participants for healthy lifestyle changes; 2) conducting monthly nutrition education and physical activity sessions focused on making more health-conscious choices; 3) providing health screenings for program participants; 4) using motivational interviewing to encourage participants to identify lifestyle change priorities; and 5) promoting local community walking tracks to increase their visibility and use by residents. The H.U.B. City Steps intervention also aims to assess community capacity for health promotion through the life of this CBPR project. The first step in this process was to engage the CAB in a collaborative process to identify and define a framework for assessing community capacity, through workshops conducted by research staff during the first project year. Having adopted the community capacity framework of Lempa et al. and its constructs, project staff guided the CAB in developing a working definition of each construct, from a community perspective. The CAB also developed a project mission statement and action plan for each community capacity construct. The H.U.B. City Steps community capacity framework and constructs then formed the basis for assessing perceptions of community capacity over the life of the project. This paper describes perceptions early in the life of this CBPR community intervention, after intervention planning and walking coach recruitment and training but before intervention kick-off, across three groups of project stakeholders, using a mixed methods design. --- Methods This assessment included two components, described in more detail below. The quantitative component assessed community capacity perceptions using a questionnaire adapted from an instrument developed and empirically tested by Lempa and colleagues with participants from community initiatives across the U.S. The qualitative component involved focus group type listening sessions in which responses to the quantitative questionnaire were elaborated further. --- --- Community Capacity Questionnaire Community capacity scales developed by Lempa and colleagues were utilized, asking representatives of each stakeholder group to rate the H.U.B. City Steps collaboration and its performance relative to selected community capacity constructs from a total of eight: leadership, resources, external networking, visibility and recognition, personnel sustainability, ability and commitment to organize action, communication with community members, and relationships with influential others. Item response was an 11-point Likert scale, with 0 denoting strongly disagree and 10 denoting strongly agree. Lempa and colleagues' 44-item, six-factor scale for leaders was administered to the RS and CAB, and included items on leadership , resources , external networking , visibility/ recognition , ability and commitment to organize action , and personnel sustainability . The 38-item, five-factor scale for non-leaders was administered to the WC and included items on leadership , resources , ability and commitment to organize action , relationships with influential others , and communication with community members . Three scales -leadership, resources, and ability and commitment to organize action-termed universal constructs, were common to all groups. Questionnaires were distributed to RS through office mailboxes, to CAB members at a regular meeting, and to walking coaches at a training meeting. Respondents were asked to complete the anonymous questionnaires at their convenience and return to research staff. For analysis purposes, a construct score was calculated for each construct by averaging all items under the construct and transforming to a 10 point scale. Total community capacity scores were calculated by averaging the construct scores . SPSS 17.0 was used for all analyses. Total community capacity scores were compared between RS and CAB groups using an independent t-test and across groups for the three universal constructs using an ANOVA. MANOVA was used to examine differences: 1) between the RS and CAB for the six constructs in the leader scale and 2) among the three groups for the three universal constructs. For the MANOVAs, the Pillai's trace test was interpreted to examine model significance due to the negative skew of the data. This test is robust to such violations of multivariate analysis assumptions . Games-Howell post-hoc tests were utilized to examine differences between groups due to unequal group sizes . --- Listening Sessions Separate listening sessions were conducted with volunteers from the RS, CAB, and WC groups. Questions were developed based on the results of the quantitative analysis. The listening session question guide included a total of 27 questions divided across 8 constructs. Special attention was given to the constructs for which there were significant quantitative differences between groups. The same listening session questions were used across all three groups to elicit more detailed information about why the differences existed and how to address differences. Listening sessions were audio recorded and transcribed verbatim. Themes elicited from the transcripts were used to develop a qualitative analysis codebook, which contained descriptions or definitions for each of the codes. The codebook was developed by a team of two researchers, one project staff, and one CAB member . The qualitative analysis team determined meaning and context using the eight constructs to help guide the interpretation. This method was utilized in order to increase internal consistency of code assignment. Once there was agreement across team members on context and meaning for each construct, quotes were extracted from the transcripts to provide examples for the interpretation. The team compared interpretations across groups and identified any notable differences between groups. --- Results --- Community Capacity Questionnaire Table 1 indicates the number of stakeholders from each group whose data were used in the analyses. Three CAB and three WC questionnaires were excluded from the analysis due to missing data on more than one construct. --- Researchers and Staff and Community Advisory Board: Six constructs Table 2 indicates the mean total score and construct scores for the RS and the CAB. The CAB scored total community capacity slightly higher than the RS . The CAB scored five of the six constructs higher than the RS. Visibility/recognition was scored lowest and ability and commitment to organize highest by both the RS and the CAB. Based on MANOVA, the model examining the differences in construct scores between the RS and the CAB was significant at F = 5.65, p = .03. As shown in Table 2, perception of leadership and personnel sustainability differed between the RS and the CAB, with the CAB scoring these constructs higher than the RS. Walking Coaches: Five constructs-The total community capacity score for the WC calculated from the five constructs rated by this group was 8.70. Among the constructs, communication with community members had the lowest score ; leadership, resources, and ability and commitment to organize all reached a mean score of 9 out of 10. Relationships with influential others was scored as 8.34±1.93 out of 10. --- RS, CAB and WC: Three universal constructs-The total community capacity score for the three universal constructs was 8.27±0.93 for the RS, 8.36±1.20 for the CAB, and 8.76±0.76 for the WC . MANOVA analyses indicated a significant difference across the groups for the three universal constructs at F = 3.60, p = .04. Post hoc analyses indicated differences between the RS and the CAB and the RS and the WC for leadership, with the RS scoring leadership significantly lower than the CAB and the WC. --- Listening Sessions Two listening sessions were conducted with the RS , one with the CAB , and one with the WC . Notable quotes for each construct from all groups are listed in Table 4. Data from the listening sessions on project leadership was examined closely because the CAB and the WC scored this construct significantly higher than the RS in the quantitative assessment. Of the nine questions asked concerning project leadership, one was "How do you think we can make the leadership more visible to people in our community?". There were several similar themes identified in the data analysis, including project members could attend community events outside of the project to build relationships with other community people and organizations and they could make people aware of the project activities through networks such as faith-based, community organizations, sororities, and fraternities . Another theme viewed by project staff as a capacity building opportunity was that the community does not participate and understand the research aspect of the project . Another question asked related to the leadership construct was "In what ways or areas could the leadership increase their follow through on commitments that have been made?". The CAB felt that the leadership was doing a great job, although the RS expressed reservations. One theme elicited from the RS reflected the need for more collaboration in the action plan . Like the WC, the RS felt that the community may not understand the research aspect of the project . The CAB also expressed concerns in the area of leadership. One question was "In what ways or areas does the leadership need to show more compassion?". This elicited the theme that project members are not flexible by allowing the community to guide the project . Conversely, another theme was that staff and community share openly and staff gives clear direction . When asked "In what ways or areas could the leadership increase their follow-through on commitments that have been made?", walking coaches viewed themselves as the leadership indicating that project and community members working with the project may not motivate others in the community to adopt or maintain healthy physical lifestyle because they themselves did not display sufficient motivation . The qualitative data focused on personnel sustainability was also examined closely since the CAB scored this construct significantly higher than the RS in the quantitative assessment. The first question asked concerning this construct was "If H.U.B. City Steps leaders left today, what do you think would happen with the project?". Themes elicited for the RS and the CAB included such things as all involved agree and understand the project ; all people are present that are needed to carry out the goals and objective of the project ; training for community members to take on leadership roles and be effective in those roles ; community and staff input in action plan ; and personnel preparation for sustaining project . Conversely the RS expressed the sentiment that project and community members would not continue to actively participate over the length of the project , perhaps explaining the quantitative difference between the RS and CAB. A similar question asked "If H.U.B. City Steps leaders left today, what would need to happen for the goals of the project to be achieved?". Themes elicited from the RS and the CAB suggested opportunities for building capacity. The CAB noted availability of such things as money, people, community groups and organizations to carry out the goals and objectives of the project , and that the project allows individuals to take on leadership roles . Themes elicited from the RS reflected different perceptions of needs or limitations, and included such things as all people are not present that are needed to carry out the goals and objectives of the project and the project is not connected with all needed community groups and organizations to help carry out the goals and objectives . As previously stated, of the five constructs assessed quantitatively for the WC, four were rated 8 out of 10 or above. The fifth construct, communication with community members, had the lowest score among the constructs. The elicited themes related to this construct were that the community does not participate and understand the research aspect of the project and that information diffusion to the community may not be as effective as it could be . --- Discussion and Conclusion The assessment of perceptions of community capacity for health promotion associated with the H.U.B. City Steps project described here was a baseline assessment before implementation of the walking intervention phase, but after considerable planning and training of walking coaches had been completed. This mixed methods assessment will be implemented annually to monitor changes in perceptions of capacity over the life of the project and to identify areas that need additional targeting for capacity building. The results suggest that the project is strong in its ability and commitment to organize; however, communication with community members and visibility/recognition of the project needs to improve. We are hopeful that this construct will be scored higher in the next round of data collection because we have taken suggestions from the listening sessions on the best ways to communicate and also better ways to make the project more visible. The project is naturally becoming more visible the longer it is in the community. In the quantitative analysis, leadership was viewed significantly lower among researchers and staff compared to the other two groups. The researchers and staff also perceived personnel sustainability significantly lower when compared to the CAB members. Neither of these differences was apparent in the qualitative analysis. This may have occurred because the researchers and staff may not have felt as comfortable sharing their thoughts in the listening sessions because they were not anonymous as was the case with the selfadministered questionnaire. Alternatively, this group may not have felt as confident about its own leadership role and performance on the project prior to the implementation of the intervention. As stated previously, there have been few quantitative instruments measuring community capacity that have been developed and empirically tested. Community capacity is an abstract concept and difficult to measure quantitatively, therefore it seems helpful to use a mixed methods approach in order to obtain more comprehensive information as a basis for evaluation and future action. A limitation of this study is the small sample size, therefore decreasing the power to detect differences across groups. Thus, additional areas that needed to be addressed to increase the capacity may have been missed. In this mixed methods assessment, the quantitative assessment allowed objective identification of specific areas where there were differences in group perceptions. The qualitative component then allowed further assessment of the reasons for these differences and ways to address them. This assessment has been a useful evaluation tool for this project in that it provides indications of strengths that are in place as well as weaknesses that need to be addressed to assure adequate community capacity for sustainability of project aims and accomplishments. Developing, identifying, and utilizing methods to evaluate community capacity is an important adjunct to CBPR due to its importance in promoting sustainable community change. CAB: "A coalition between the university and the community in helping the underserved." WC: "I think its just going to take time …It takes time to build relationships… the proof is in the pudding." Visibility/ Recognition RS: "People have started bringing their teenagers to learn because the word has gotten out that it is educational so they bring friends and family." CAB: "And so the program is kind of saying we are taking these average Joes in the community and taking these average Joe coaches… and we're doing this. I think creating awareness that people can be responsible for their own health. The first step and how they can involve their friends and other communities." WC: "Like I said its word of mouth cause when people see you out walking and with the t-shirt on… they are curious and want to know what kind of program is this and how can they be a part of it." Ability and Commitment to Organize Action RS: "Originally as the CAB we were supposed to have them review the procedures and other things more thoroughly and then we were supposed to teach them more of the research aspects of things. I'm not sure how much follow through has transpired with that and I think that is something we need to strive for." CAB:"Or if we didn't want to lose Hub City Steps… since its such as wonderful program, we could incorporate and become our own 501c3 and board of directors and make sure every year we rotate on and then the information won't just evaporate and we could just put it into place and keep rolling!!" WC: "Realistically if the project leadership left today you will not have that much cause some people are motivated by what we get… incentives and so forth but I do think a high percentage of the people will move on… 'cause I'm not turning back." Personnel Sustainability RS: "I think it's always a good thing no matter who is doing it… whether it be us more on the research end or the community end just making sure we reaffirm we want the feedback so we can improve things for participants, coaches and community." CAB: "Keeping the program interactive…Most people are working so and taking time out of their busy schedules so making it interactive so it doesn't feel like work." WC: "And another thing is like the thing… people just knowing that you've lost weight that opens the door for you being able to talk about it you know. You know saying I've been walking, 'cause I find myself saying that. People say you know it looks like you've been losing weight and I say 'yeah I'm in a group'" Communication with Community Members RS: "I know one of the goals was to help the community to understand research. That's one of the goals of CBPR and then…sustainability. So, I think that's something that needs a little follow up… maybe showing them some of the reports we're coming out with so they understand some of our procedural aspects…" CAB: "Cause I was at a housing meeting and when I brought up the information and I had the little shoes and letting them know and she was saying how she was turned down but it made her more aware of her health cause she was a diabetic… but she wasn't taking care of herself… and her saying just because I couldn't participate in this program, I started anyways… that's a testimony." WC "And the t.v. has been good and the media but the main thing is sharing the success stories. One of the things I mentioned to the project's community coordinator and I hope she can do and --- Community Capacity Constructs Notable Quotes I recognize the idea of confidentiality in giving those numbers. But at our next assessment, if she could provide us as a walking coach with stats not by name. If participant one cholesterol went down overall or if blood pressure went down, so there I could go and say this is my team and we lost overall this amount of pounds…if we could get this type of data similar and we could share this type of data. Relationship with Influential others RS: " …and with the mayor, I think it's been a big deal having him involved because he reaches way outside out community. You know he took some of our pedometers to meet with Michelle Obama." CAB: "Bonding with the city was a huge move. The biggest thing was partnering with the city." WC: "I say that one avenue would be the churches… if you could get the churches involved, get the preachers involved. They follow them quite well and I think that would be a way to say basically be healthy, this is how we're going to serve the Lord.

H.U.B. City Steps is a five year community-based participatory research walking intervention designed to help lower blood pressure in a majority African-American population in southern Mississippi via community collaboration and capacity building, increased walking, culturally tailored health education sessions, and motivational interviewing. Building community capacity for physical activity is a key component of this intervention. Qualitative and quantitative methods have been used assess how project stakeholders perceive the community capacity building efforts of the project. This paper illustrates the baseline results of this mixed methods approach from the perspective of three groups of stakeholders: Project researchers and staff (RS); Community Advisory Board (CAB) and intervention walking coaches (WC). Eight constructs were examined including leadership, resources, external networking, visibility and recognition, personnel sustainability, ability and commitment to organize, communication with community members, and relationships with influential others. Quantitative results indicated significant differences among stakeholder groups for project leadership and personnel sustainability. Qualitative perspectives provided an opportunity to examine possible reasons for these differences. Overall findings provide direction related to improving intervention outcomes and sustainability.

I. INTRODUCTION Language is a double-edged sword. Its use depends on the motive of the users . In the disruptive era, the other side of the coin is technology. Technology is an integral part of culture and its development; it cannot be separated from the society that owns it . Since language is the social mirror, language cannot be separated from technology, as in information, digital, and internet technology . Furthermore, it should be emphasized that the close relationship between language and technology has great implications for the development of the study of language fields under the umbrella of linguistics . Of course, the greatest impact is on the study of meaning or pragmatic meanings in verbal communication, where language is the means of communication . The study of meaning focuses on pragmatics . It continues to shift from the most conventional pragmatics, called systemic pragmatics or semantico-pragmatics due to its systemic and semantic dimensions, to the general pragmatics, to the culture-specific pragmatics, and is now developing into a new branch of pragmatics called cyberpragmatics . The latest pragmatics is the most closely related to technology, especially digital-based technology and internet-based technology. Furthermore, it should be emphasized that the study of a speaker's meaning is bound to context, be it social-societal, cultural, or situational dimensions . The three types of context in pragmatics play a simultaneously important role in determining the speaker's meaning. Ignoring contexts in interpreting the speaker's meaning will result in misunderstanding, fuzziness, and ambiguity . However, each type of context in pragmatics plays a significant role in determining what to mean. The social-societal context plays an important role in determining the meaning of societal pragmatics introduced by . Cultural context is very important in determining the speaker's meaning from the perspective of cultural pragmatics, both in the sense of intercultural pragmatics, cross-cultural pragmatics, and culture-specific pragmatics proposed by Blum-Kulka and Halliday . The situational context plays an important role in determining meaning in the general pragmatics perspective and subsequent pragmatic developers . Thus, in the various types of pragmatics, it is clear that there is evidence of meaning triadicities, instead of meaning disabilities as the focal point in semantics . This article should contribute to pragmatic studies both in and outside of Indonesia, which have not been explored fully. The development of pragmatic studies has lagged behind other fast-growing and ever-increasing research in linguistic and nonlinguistic phenomena. The development of digital technology and internet technology impacts the increasing use of language in social media, which inevitably results in various hoaxes and hate speech. The use of language in social media is not accommodated well in various research in language studies. Consequently, the issues of meaning in verbal communication are not elaborately researched. The ensuing consequence is the rampant cases of confusion, fuzziness, and misunderstanding in social media communication. Now is the right momentum for the study of pragmatics to evolve from general pragmatics to virtual or cyberpragmatics. Similarly, the study of contexts in pragmatics must evolve from conventional external contexts to virtual external contexts. This short paper focuses only on the socio-societal context from the cyberpragmatic perspective. The dimensions of the social-societal context to be described in this paper deal with the nature of changes, namely the changes to the elements of context. In certain cases, the description of the changes will be more exhausted to the aspects of each element of the social-societal context. As mentioned earlier, the changes are the results of the development of information technology, digital technology, and internet technology. It is important to describe the evidence and aspects of the element changes so that the interpretation of digital and virtual dimension speech is easier and clearer. Based on the description, the research problem is formulated as: "What is the evidence of the social-societal context element changes found in cyberpragmatics?" In agreement with the formulated problem, this article's objective is to describe the evidence of the social-societal context element changes in cyberpragmatics. --- II. THEORETICAL REVIEWS In this article, two basic theories are discussed as a framework of reference, namely the theory related to the changes from systemic pragmatics to cyberpragmatics , and the theory linked to the changes from conventional external context to virtual external context . Regarding the first theory, it is worth noting that the study of the speaker's meaning is not independent of the semantic meaning . The pragmatic study is still oriented to two branches, namely semantics, and pragmatics. The systemic semantic meaning study is known as the systemic pragmatics era or semantico-pragmatics . Thus, the study of meaning is still dependent on the study of semantic meaning, which is dyadic. The rigor of the study of meaning emerged coincidentally with the interdisciplinary linguistic sciences such as ecolinguistics . Ecolinguistics was further developed into metaphorical ecolinguistics and naturalistic ecolinguistics . Both branches of ecolinguistics emerged to respond to formalists' perspectives in linguistics who view language as being homogeneous. In its development, the sociology of language evolved into sociolinguistics, focusing on language variations. Contexts in sociolinguistics are needed to identify language variations, while contexts in ecolinguistics are needed to identify the role of the environment as metaphorical and naturalistic . Thus, from the functionalists' perspective, the entity of contexts is seen as important as its existence. Functionalism in linguistics assumes that language is heterogeneous. Language is never singular and homogeneous as the formalists believe . The study of meaning is understood along that line. Originally, the study of meaning focuses on the dyadic meaning of a word. Recently, the study of meaning has focused on the triadic meaning of a word. The triadic meaning is derived from the involvement of and dependence on contexts because language is never independent of the context . Although the study of meaning has shifted from the pure semantic meaning to the study of the speaker's meaning, in its early development, the functionalists had their footings on two sides, namely the semantic side and the pragmatic side. The next stage is the development of general pragmatics. The development of pragmatic rules is linked to pragmatic parameters, pragmatic principles, Cooperative Principles, politeness principles, and other linguistic principles based on Western languages and cultures . The study of language in non-European and non-American countries has not been widely explored. Thus, the pragmatic rules formulated based on the study of European and American languages are followed, borrowed, and applied blindly to the study of Mediterranean and Asian languages. Of course, the rules are not always applicable to all. The principle of linguistic politeness and its maxims, for example, cannot always be implemented in Asian society and cultures . Similarly, the cooperative principle and its maxims are not always suitable for non-Western communities and cultures. Thus, the development of general pragmatics, which led to the study of meaning based on external linguistic factors, was admittedly phenomenal in its heyday. However, it is no longer appropriate to apply it to the local languages and cultures. Meanwhile, local cultures possess a wealth of cultural wisdom and extraordinary cultural values . Hence, the rigor to study meaning based on a specific local culture began. This development inspired cultural pragmatics, such as intercultural, cross-cultural, and culture-specific pragmatics. Therefore, pragmatics evolved into intercultural pragmatics, cross-cultural pragmatics, and culture-specific pragmatics . From systemic pragmatics to cultural pragmatics and its variations, pragmatics' development confirms the pragmatic studies' gap as described earlier. However, along with pragmatic development, globalization is also crucial in affecting all aspects of social life. The industrial revolution and the emergence of globalization gave birth to the glocalization perspectives proposed by Graddol . In turn, they led to remarkable developments in information technology. Language is increasingly complex as it involves aspects of information technology and digital devices. The development of increasingly complex matters is driven by the internet of things perspective of life . Language is inseparable from the technology-ridden reality. The original and simple function of a language to develop common sense and a vehicle for strengthening interpersonal cooperation seems to have been distorted. The growing use of language for different genres and purposes on social media has inevitably created hoaxes, hate speech, and other language function abuse . A new perspective in pragmatics, called cyberpragmatics, which is virtual, emerged in 2011 . Interpreting language cannot be independent of the virtual contexts as an inevitable impact of digital and internet technology manifested in various social media contexts such as Facebook, Instagram, Twitter, etc. With the advent of cyberpragmatics, the basis and concern of sociolinguistic studies have shifted from the social community to the virtual community . In the past, the similarity of the verbal repertoire between individuals creates a speech community. Now, the parameter has changed. The speech community has shifted to a virtual community. The virtual community is not based on the similarity of verbal repertoire or the similarity of language competence . The virtual community is characterized by the common interest in certain public figures, making the news in the digital media and the internet . Therefore, the netizens or the virtual community may not know each other, do not speak the same language, do not share the same cultural roots, and live thousands of kilometers away. Despite the long distance across different time zones, digital technology makes it possible for them to form a virtual community. Communication and interaction can be built and established well. They can even go beyond the social community in the past. This has led the researcher to examine the evidence of social-societal context elements and their aspects . It guides researchers to examine and observe more carefully and deeply the evidence of changes in social-societal context elements and their aspects. The preliminary study context is divisible into three, namely the social-societal context, cultural context, and situational context. The social-societal context is based on society . The social context has a horizontal dimension, while the societal context is vertical. Social contexts are intertwined with social distance and social equality, while societal contexts are intertwined with social status and social ladder. The social and societal contexts are very crucial to determine the speaker's meaning from the perspective of sociopragmatics and pragmalinguistics, proposed by Leech , and the societal aspects, proposed by Mey . The social contexts have been identified, while the local context is elaborated by Rahardi . Mey further explains the social context, and the local context is developed further by Rahardi . The social-societal context is also intertwined with the cultural context. This context refers to the various cultural dimensions shared in society, including cultural knowledge, cultural behavior, cultural artifacts, and the cultural wisdom held by society. The nature of cultural context has been discussed by Blum-Kulka , Halliday , Rahardi , Suszczynska , and Wierzbicka . The last type of context is situational context. Verschueren has specifically featured the situational context in Pragmatics. The general pragmatic studies typically address this general situational context. Subsequently, the virtual external context is currently being studied by Rahardi . This virtual external context will be very useful to study pragmatics from the perspective of cyberpragmatics . In Rahardi's observation, the conventional external context has changed the context's elements and aspects. The result of the change is what Rahardi refers to as a virtual external context. The evidence of the changes is described in his articles, such as the change in the speaker and addressee, and other interlocutors. In the conventional external context, the three elements of context are usually seen from the dimensions of age, gender, and background . However, from the perspective of cyberpragmatics, the elements to consider in the interlocutors are aspects of their life, life philosophy, social circle, daily life, life mission, etc. The aspects attached to the interlocutors are very complex. Information on the interlocutors' gender and age alone will only inform the language variations. The pragmatic meaning of the speech will be correctly interpreted by taking into account the complex aspects mentioned earlier. Consider the other contextual elements, such as the channel of speech. The channels used to speak and communicate in the present are very different from those in the past. People needed to use "amplifiers" to hear their voices clearly by people from a long distance. For example, in remote villages, devices of pasimology, such as the sound of slit bamboo drums, bells, etc., are still used to signal incidents and summon villagers . At present, such a device is ubiquitous and within a hand's reach, as in the "digital gadget." With digital devices such as smartphones, people can spread the news worldwide in just a matter of seconds by simply clicking or tapping the "share" button. On the one hand, the speed of information transmitted with the "digital gadget" device is an advantage . News travel in a fast, time-efficient, and effective manner. On the other hand, information speed is at the expense of language use, which is used carelessly. This may lead to long-term irrevocable damage to the language. It is so easy for religious groups, delinquent juveniles, and anarchist groups to plan and organize a crime or terrorist attack, and in just a matter of seconds, they orchestrate a series of attacks to destroy their targets. In this paper, pragmatic theory shifts from systemic pragmatics to the cyberpragmatics; similarly, the theory of context changes from conventional external contexts to virtual external contexts. These contexts are used as a reference framework in this article. --- III. METHODS This research is descriptive-qualitative type. The source of this research's locational data is a variety of social media available on smartphones. The substantive data sources are texts on social media in which there are objects and data of this research. Furthermore, the object of this research is the evidence of a shift in the socio-societal context. This research's data are excerpts of speech containing the object of research, which proves the shift in the social-societal context . Therefore, it is obvious that the object of research, and its context, is the manifestation of the data in this study. The data were collected using the observation method, which is equipped with note-taking techniques and recording techniques. The recording and recording results are then transcribed for identification, selection, classification, and data typification . Furthermore, the data were analyzed by applying the contextual analysis method. This contextual analysis method is applied by basing and linking the context. This contextual analysis method is also called the equivalent analysis method, especially the extra lingual equivalent analysis. The contextual analysis method or the extra lingual analysis method is applied because this research's perspective is the cyberpragmatics perspective with a virtual external context as the main determinant of its meaning . Before applying the analysis, data that has been classified and typified properly is triangulated with the expert and consulted on relevant theories. The purpose of implementing data triangulation is to ensure the validity of the analyzed data. After the triangulation process is complete and the data analysis process is done properly, the final step is to present the results of this study's analysis. Presentation of the analysis results is carried out by applying informal methods . --- IV. RESULTS In this section, two things are presented, namely research findings and discussions of the research findings. The researcher deliberately separates the research results from the discussion with the intention that these two things are clear and have a deep description. The following shows the presentation of the two parts presented one by one. Elements of social-societal contexts change following the development of digital and internet technology. In addition to the context elements' changes, aspects of each social-societal context element are likely to change. The context elements are based on Dell Hymes' speaking grid, a classificatory grid as a descriptive framework for the ethnography of communication. Speaking is the mnemonic device to describe the setting, participants, ends, act sequence, keys, instrumentalities, norms, and genre. In this research, the setting can be differentiated into two, namely, the setting of place and the setting of time. The participants' element can be seen as two kinds, namely speakers as participants and hearers as participants. The element of objectives or ends can also be differentiated into two, namely the single and multiple ends. In addition, the element of norms can be divided into two, namely the norms of interaction and interpretation. The following table shows evidence of the changes. Norms if Interpretation √ Genres √ --- V. DISCUSSIONS Table 1 shows that the changes occur in the following grids: participants, setting, ends, and instrumentalities. Changes in the setting element occurred in the aspects of venue, time, and atmosphere. Changes in the participant element occurred in the aspects of perception of gender, age, and social status. Changes in the end element occurred in the aspects of monodimensional goal manifestation and multidimensional goal manifestation. Changes in the instrumentalities element occurred in the aspects of the kind of tools and the range of errand communication. Each piece of evidence of changes will be discussed in detail. The element of participants can be seen as two kinds, namely speakers as participants and hearers as participants. The element of settings can be differentiated into two, namely, the setting of place and time. The element of objectives or ends can also be differentiated into two, namely the monodimensional and multidimensional ends. Each of the above elements will be discussed in detail below. --- A. Evidence of Context Change in the Participant's Element The speaker and the addressee play very important roles in speaking and communication. With the speaker and the addressee, the other participants are the third party or audience involved in a conversation. In a real-life conversation, the presence of an audience determines the linguistic codes being used. The presence of a third party who has no --- THEORY AND PRACTICE IN LANGUAGE STUDIES business in the subject matter may cause the speaker and the addressee inconvenience. The third party can be anyone who intentionally appears on the scene and wants to hear the conversation or any passersby who accidentally appear on the scene and overhear what is being said . The third-party could be any individual or a group of individuals. The age and gender of the speech participants were relevant to determining linguistic codes in the past. For instance, talking to children is different from talking to adults or the elderly. Similarly, the linguistic codes used to talk to men are different from those used to talk to women. In the study of language variations and the study of meaning, these elements are considered important. From the perspective of cyberpragmatics, the aspects of participants are not the same as the ones presented earlier. When digital and internet technology is crucial to determine meaning in communication, the aspects discussed earlier do not play a significant role. Such aspects are not needed in a virtual conversation on social media whose participants have different cultural and social backgrounds . When we send a message, share news, and upload information and images on social media, should such aspects be considered? The answer is no. Anyone can share anything on social media in an instant, and it will be available online for anyone to see, regardless of their background, social interest, socio-political background, social circles, etc. In other words, there has been a change in the boundaries of a region. Communication has become completely 'borderless' from one person to another. The aspects of age and gender are not important to be discussed in a virtual conversation. The comment sections in social media posts may involve people from different gender and age groups. The linguistic codes and manifestations are not determined by aspects of participants as identified earlier. This proves that in the virtual external context, the elements of participants have changed. Understanding the speaker's meaning by considering the change of the elements of the participants in cyberpragmatics will not result in the interpretation of the true meaning . Thus, the pragmatics' focus should be on different aspects from the earlier pragmatics because the aspects and context elements are different from those in the past. Table 2 showed the social-societal context changes in the element of participants. --- B. Evidence of Context Change in the Settings Element One of the most important social-societal elements to be considered in interpreting the meaning of an utterance is the setting. Setting refers to the situation where an utterance is spoken. However, the setting is limited to the situation and the physical and social setting in which the utterance occurs . In a specific social setting, such as a family gathering in a Javanese community, people tend to speak formally and politely. The language used considers acceptable manners and etiquette to ensure that everyone feels valued and respected by others. In contrast, in the factory setting, factory workers in a particular factory will use different linguistic codes from those who work in a different setting in a different region. It is evident that in the social-societal context, a speech event's physical setting will greatly determine the linguistic codes . In a sacred physical setting, such as the funeral service in a Javanese village, people tend to behave solemnly and speak in a hushed tone. It is culturally understood that, on such a sacred occasion, people speak softly and behave politely. In the church, the mosque, or other places of worship, people are expected to behave politely and speak softly in a hushed tone. The social setting determines the linguistic codes used in speech events. This cultural norm was a part of a social convention in the past. However, in the era when people believe in freedom of speech and democracy, the physical, social, and environmental settings seem to have changed dramatically . Social settings are no longer considered in social media speech events. In the digital era, people no longer care about the social setting, which required politeness from past participants. In this day and age, people are more concerned with the message's urgency . Once again, urgency is the sole determinant. In times of grief, people can send messages without considering the addressee's psychological setting. For instance, a student may text a lecturer through social media as if nothing special happened, even though he knows that the lecturer is in a grieving psychological state. A few decades ago, when a neighbor died, people in the whole neighborhood came to pay respect to the dead and to help around the house. Not a single kiosk or food stall opened, no one went to school or work, and children stopped playing. The whole neighborhood was mourning the bereaved family. In the digital era, the setting of the speech event has changed dramatically. It is no longer relevant to interpret the speaker's meaning in the current pragmatic study from the conventional external contexts as in the past. Someone is forced to compromise, delay, or even cancel her/his intention when he/she considers the conventional elements too seriously instead of the virtual external contexts . Thus, the evidence of a change of setting is very important to note to understand the speaker's meaning in communication in the era of disruption. This proves that pragmatic studies should focus on more complex dimensions, unlike systemic pragmatics, general pragmatics, and culture-specific pragmatics . Social-societal context changes in the element of settings . --- C. Evidence of Context Change in the Ends Element The "ends" element refers to the speaker's purpose in communicating with the addressee and other participants. In communicating and interacting with each other, people almost always have a goal or purpose in mind. A person who habitually greets his neighbor positively is considered a polite member of society. On the contrary, a person who does not socialize with their neighbors, let alone greet them, is considered arrogant, unfriendly, and antisocial. The Javanese people use the expression "ilang Jawane" to refer to someone who no longer observes the Javanese cultural norms. While waiting at a train station before a train ride, people normally strike up a polite conversation with fellow passengers. In such a setting, people tend to use phatic speech function to break the silence. When Indonesians meet fellow Indonesians in a foreign country, say in Europe or America, they will greet each other and strike up a conversation even though they have not known each other previously, simply because they are foreigners in the country. Thus, it is obvious that people greet each other and communicate for a purpose, whether the purpose is substantive, phatic, or others. People have a communicative purpose when speaking. The linguistic codes people use may serve the function of small talk, a phatic function, or a substantial referential function commonly happening in verbal communication . The notion of "ends" occurs in speech events and is considered to be true in sociolinguistics, systemic pragmatics, general pragmatics, and culture-specific pragmatics. The question remains whether in cyberpragmatics the notion of "ends" still applies. In other words, has the notion of "ends" changed in manifestations? It is safe to say that people have their "ends" in mind when they converse with others. However, the goals and purpose of communication in the present time are more multidimensional than the monodimensional goals and purposes of communication in the past . The multidimensional goal or "end" occurs because life is getting more complex nowadays. An utterance that was interpreted individually in the past may have multiple meanings now. The utterance may have multiple ends. Metaphorically, in the past, when people shot a target, the bullet only hit one target at a time. In the era of cyberpragmatics, when people shoot, the bullet may hit several targets simultaneously. Thus, interpreting the purpose of people's speech in the era of cyberpragmatics is not as simple as it was in the past. People may not show their true colors in social media posts . Behind their seemingly good image, they may have bad or even devilish intentions. This illustrates the changes from conventional external contexts to virtual external contexts. It is important to understand the new phenomena to warn people to be vigilant against strangers with hidden motives prowling in social media. Social-societal context changes in the element of ends . --- D. Evidence of Context Change in the Instrumentalities Element The element of social-societal contexts is related to the notion of a "channel" or medium to express the speaker's meaning to the addressee. In the past, to speak to a big audience, people used a loudspeaker or "amplifier" so that the meaning or purpose can be delivered well. As technology advances, the development of audio and video technology makes an impressive replacement . To measure the success of a stage performance, the quality of audiovisual equipment during the show becomes the main consideration. News broadcasts are conveyed through loudspeaker devices to ensure the good quality of channels or "instrumentalities" in delivering the message. It is common to consider this as the conventional external context that applies to various pragmatics outside cyberpragmatics . In today's internet era, the medium or "channel" to deliver meaning has changed dramatically. The medium to spread information is a small device within a hand's reach. Through smartphones, any information can go viral in a matter of seconds. In the past, tools to amplify the message were limited by time and space. However, from the perspective of cyberpragmatics, the channel is identified as smartphones. Through smartphones, the spread of the message or intention is done simultaneously by tapping the sharing button in the smartphone . The dissemination speed of the messages or information should go hand in hand with the quality of the information being delivered. Unconfirmed errors or inaccuracies in the information, no matter how small, will be irrevocable once they have spread to the world. The rampant spread of fake news or hoaxes in social media lately is disconcerting as they are available to be read and accessed by everyone all over the world regardless of their inaccuracies. In the present-day world, one person can have several smartphones. They are also constantly connected to social media through their smartphones. When a person gives inaccurate information, instantly, many people respond to the errors for everyone to see. The dramatic changes in the "instrumentalities" of the virtual external context should be seriously heeded by the speakers and other speech participants from the perspective of cyberpragmatics . Social-societal context changes in the element of instrumentalities . --- VI. CONCLUSION In a nutshell, it is important to restate the four pieces of evidence of changes in the social-societal context from the perspective of cyberpragmatics. The four pieces of evidence are in the elements of participants, setting, ends, and instrumentalities. The shift of elements of social-societal contexts from the conventional type to the virtual one determines the quality of communication in the disruptive era. Linguistics studies, especially about the branch of pragmatics, must refocus the core of the study, that is, the study of the speaker's meaning must be based on virtual external contexts. The impacts of informational, digital, and internet technology must be wisely considered in the study so that this branch of linguistics focusing on the study of speakers' meaning will run on the right track in the disruptive era. More extensive and in-depth observations of other elements of other social-societal contexts are called for to formulate evidence of changes more holistically. The detailed description and illustration of the evidence will significantly contribute to the growing research in the field of cyberpragmatics. Understanding the meaning of utterances in the digital and internet contexts will be easier to be done. Similarly, language will evolve with the community and culture where it is spoken, instead of being left behind due to limited research and observations of the community's language phenomena. Research rigors to examine the linguistic phenomena involving digital and digital technology must be encouraged in these modern times.

This article aims to describe evidence of change to the social-societal context seen from the cyber pragmatic perspective. The substantive data sources were texts on social media in which there were objects and data of this research. The data were collected by using the observation method equipped with note-taking techniques and recording techniques. Data were analyzed by applying the contextual analysis method. We applied the contextual analysis method or the extra lingual analysis method because of the cyberpragmatic perspective of this research with a virtual external context as the main determinant of its meaning. Before the analysis, data that were classified and typified properly were triangulated with the expert and consulted on relevant theories. The results show that the social-societal context element changes occur in the following context elements: (1) setting, (2) participants, (3) ends, and (4) instrumentalities. Setting element changes occurred in the aspects of venue, time, and atmosphere. The instrumental element changed in the aspects of the kinds of tools and the range of errand communication. The participant element changed in the aspects of perception of gender, age, and social status, and the last element, changed in the aspects of monodimensional goal manifestation and multidimensional goal manifestation.

Elder mistreatment is a public health problem that takes a devastating toll on the independence, health, and well-being of millions of older Americans. A national prevalence study indicated that six million people aged 60 and older are victims annually, or about 1 in 10 older adults . Estimates are even higher globally, reaching one in six older adults . If EM were thought of as an illness or syndrome, it would be recognized as a public health crisis . EM has been defined by the National Research Council as: "Intentional actions that cause harm or create a serious risk of harm, whether or not intended, to a vulnerable elder by a caregiver or other person who stands in a trust relationship to the elder," as well as "failure by a caregiver to satisfy the elder's basic needs or to protect the elder from harm" . Because this definition focuses on a caregiver or trusted other, it excludes self-neglect, but does pertain to physical and psychological mistreatment, financial exploitation, and caregiver neglect. In this article, we focus on mistreatment by a "caregiver," defined as "a person who bears or has assumed responsibility for providing care or living assistance to an adult in need of such care or assistance" . Although the magnitude of EM indicates a need for developing and testing prevention strategies, evidence-based interventions and primary prevention programs for EM are lacking . Primary prevention, which refers to interventions that intercede to stop abuse from occurring, has not been systematically studied in EM . However, research has been conducted on a small number of secondary and tertiary preventive interventions, which seek to stop ongoing abuse and/or mitigate its detrimental effects . EM research can, however, benefit from evidence developed in other areas of family violence. For example, primary prevention programs have been developed to address child maltreatment and intimate partner violence , offering possible applications for EM . Because family violence occurs throughout the lifespan and similar risk factors are found for child maltreatment, IPV, and EM, prevention strategies for EM should be developed by building on what has been learned from research into other areas of family violence. Indeed, Hamby & Grych describe abuse in families as "webs of violence" wherein exposure at any point in the lifecourse increases risk of eventually experiencing and/or perpetuating violence. They urge researchers to follow these complex webs across different parts of the lifecourse, comparing and contrasting risk factors, treatment approaches and relevant theories. --- EM and Caregiving Approximately one-quarter of adults aged 65 and older reported relying to some extent on assistance from a caregiver because of physical and/or cognitive impairments . Although studies indicate that most EM is committed by family members , the extent to which family caregivers are perpetrators of EM is unclear. Two studies found that 26% of those receiving care from a family member reported experiencing either physical mistreatment or exposure to potentially harmful behaviors at the hands of a family caregiver . In another study, which focused on those caring for family members with Alzheimer's Disease and related dementias , 47% acknowledged committing at least one type of mistreatment, often psychological mistreatment . Even as family members are increasingly called upon to provide care to older relatives . Training and support available to caregivers may be sparse or nonexistent, and some proportion of family members may have little ability to adequately provide care. Given the vulnerability of older adults who rely on family caregivers and the documented high rates of mistreatment, prevention programs geared to families would seem to offer a promising first step to reducing the high levels of EM. Given this and the limited state of knowledge on primary mistreatment prevention in this population, approaches to evidence-based prevention interventions in other fields of family violence seem to offer a useful starting point. --- Shared Theoretical Mechanisms in Family Violence Because EM shares several risk factors with other forms of family violence, strategies based on relevant theories and causal mechanisms from other areas can help researchers identify adaptable prevention strategies that lower risk when developing interventions . We discuss several theories divided into three categories: addressing relationship quality and skills, stress and reaction, and knowledge and experience. Although the precise mechanisms of each theory are distinct, intervention components and outcomes discussed subsequently broadly follow this trifurcation. --- Relationship Quality and Skills Social learning theory Previous exposure to family violence is associated with future perpetration of abuse. As the Adverse Childhood Experiences study demonstrates, exposure to negative experiences early in life increases the likelihood of participating in IPV; those with high ACE scores have five times greater odds of perpetrating physical IPV than those with low scores . As the web of violence concept suggests, patterns of violence developed over the lifecourse can yield devastating results in old age . Shedding light on such patterns though interviews, Pickering and colleagues observed that adult daughters who reported that they were abused as children often engaged in reciprocal aggression towards their elderly mothers, particularly verbal/psychological abuse . Experiences of childhood trauma have also been associated with increased risk of perpetration of neglect among caregivers . Social learning theory explains this patterned violence, wherein behavior is learned through observation and replicated . Extrapolating from this theory, one can posit that abusive behaviors are learned through observation, and are replicated after having been established as normative behaviors. --- Attachment theory Poor relationship quality between potential victims and perpetrators is also related to risk of child maltreatment, IPV, and EM. In cases of child maltreatment, this includes low levels of maternal warmth -associated with physical abuse and neglect-and low levels of attachment . Attachment theory indicates that positive caregiver interactions with a child promote emotional and behavioral regulation while increasing the parent's investment in the child's wellbeing . Poor attachment in grown children has been implicated as a factor contributing to neglect of their older parents . --- Stress and Reaction --- Caregiver stress High stress levels among perpetrators have been identified as a risk factor for family violence , including EM, although the evidence is mixed. Pillemer & Finkelhor's seminal study showed that higher care needs among recipients-believed to raise levels of caregiver stress-were not associated with EM. Rather, those who mistreated older adults were more likely to have experienced other life stresses within the year in which mistreatment occurred. A more recent study found that among caregivers for older adults, more hours spent providing care and higher levels of subjective burden were predictive of physical and psychological mistreatment . Similarly, parents who commit child maltreatment find parenting more stressful than other parents , suggesting that subjective assessments of caregiving are central to understanding the role of stress in risk of mistreatment. Specifically, parents who commit maltreatment are more reactive to stressful events than those who do not commit mistreatment , and may have lower capacity for emotional regulation. Although the EM field is still learning about mechanisms underlying the relationship between stress and mistreatment, recent findings in caregiving show similar results as research in child maltreatment. High subjective burden among caregivers is predictive of potentially harmful behaviors . Coping styles are another avenue that should be explored to explain the relationship between stress and mistreatment, as coping can alter the outcome of stress experiences . --- Reactive abuse Rather than care needs per se, certain behaviors are associated with risk. In child maltreatment , IPV , and EM , aggressive behaviors from the victim are associated with abuse, potentially eliciting negative reactions from perpetrators. Similarly, older adults who exhibit aggressive behaviors related to behavioral symptoms of dementia are more likely to experience mistreatment than those without dementia . --- Knowledge and Experience Low levels of knowledge and experience among caregivers are associated with increased risk of mistreatment. For example, parents with poor understanding of their children's behaviors are at risk of participating in harmful behavior towards children . Children displaying certain challenging behaviors are also at greater risk of maltreatment , possibly due to the higher degree of child-rearing skills required of parents. Poor understanding of recipient behaviors may similarly explain high rates of mistreatment among caregivers to people with dementia . Although high-demand caregiving can contribute to EM through stress, inexperience on how to provide care is another factor . Indeed, some cases of criminal neglect are nearly indistinguishable from those where caregiver is overwhelmed . --- Factors Unique to EM Although applying risk factors and theories from different forms of family violence, it is important to recognize dynamics that make EM unique. Aging is associated with illnesses and disabilities that may change an adult's status from independent to dependent, requiring different behaviors from others. For example, family norms of gift-giving may become exploitative if the older adult loses financial decision-making capacity . Further, although the dependency and need for care among some older adults may mirror that seen in children, an important distinction in older adults is the assumption of autonomy. Unless it is legally determined that the older adult lacks capacity, they retain their decision making autonomy, which includes the autonomy to make seemingly poor decisions or to place themselves in harmful or detrimental situations. --- Community-Based Intervention Models in Family Violence Like theories and risk factors, community-based prevention models for other types of family violence offer useful lessons for addressing EM. We largely focus on home visiting models seen primarily in the child maltreatment field. There are several reasons for doing so. As with parents raising children, family caregiving for older adults takes place primarily within the home. A home setting is valuable for an intervention because it gives providers insight into contextual risks that may not be apparent in other settings . Further, these approaches enable tailored interventions appropriate for the diverse needs of families, as opposed to more groupbased interventions . Home visiting models are particularly attractive in an EM prevention context, as a strong tradition of home-based intervention already exists for family caregivers . In reviewing approaches to community-based intervention, this section aims to identify components that contribute most to success. Table 1 summarizes the intervention models described. --- Home Visiting Programs to Prevent Child Maltreatment Home visiting programs in child maltreatment date back to the 1970s . Some programs provide generalized supportive services, such as the Healthy Start program, which connects new parents to community resources . Other interventions target specific mechanisms thought to underlie child maltreatment. This strategy is used in the Promoting First Relationships intervention, which aims to improve parents' sensitivity to child needs and promote attachment . Still other programs are multicomponent, offering a range of services and support to families . Despite variation, some designs and strategies are consistently more successful than others. For example, home visiting interventions that include a training component outperform those without training. Aiming to break cycles of violence and improve relationship quality using tenets of social learning and attachment theories, parents participating in Parent-Child Interaction Therapy are taught parenting skills over 12 to 14 training sessions . Other training approaches focus on improving parents' perceptions of children. In an adapted version of the Healthy Start program, at-risk families were trained to alter negative appraisals of children. At follow-up, the control group had a 24% rate of physical abuse at follow up, whereas those receiving the enhanced version of the Healthy Start intervention showed a rate of 4% . Although not specifically designed to prevent child maltreatment, the Family Check-Up model offers several promising approaches. A home visiting program addressing maladaptive behavior patterns in children, it is tailored to the unique needs of families over the course of three visits . Parents are interviewed about their situation during the first visit, assessed and observed interacting with the child during the second visit, and receive feedback during a final visit to discuss parenting strengths and weaknesses as well as available services and supports to meet the family's needs . Although its impact on preventing child maltreatment remains largely unstudied, recent findings by Dishion and colleagues show lower levels of neglect Originally aimed at parent with children aged two to seven with behavioral problems, this intervention is provided over the course of 12 to 14 weekly sessions and aims at improving relationship quality and disrupting coercive tendencies between parents and children. During the first seven sessions parents follow the child's lead during play in order to strengthen the relationship. In the final seven sessions, the parents are taught skills to direct the child's behavior, thereby better managing it. --- Toth et al., 2013 Family Check-Up Comprised of three sessions, the Family Check-Up begins with an interview with the parent to learn about the family situation and engage in motivational interviewing techniques. This is followed by an assessment, which includes a questionnaire and observation of the parent and child interacting. At the third session, feedback is provided on the family's strengths and weaknesses, as well as a discussion of accessing additional supportive services. Men respond to program recruitment invitations and are asked to complete a survey about their interpersonal relationships and engagement in aggressive or violent behaviors towards their partners. They are provided with a personalized survey summary which presents their responses alongside answers from the general population, which often serves to highlight participants' deviation from societal norms. Participants then engage in individual telephone counseling with study personnel trained in motivational interviewing techniques to promote change. --- Mbilinyi et al., 2011 Prevention Relationship Enhancement Program Couples are taught skills to better manage negative affect and improve communication. Skills are taught during five sessions, scheduled 1 week apart, with three to five other couples, lasting for 3 hr each. Sessions are devoted to one or two topics and led by trained psychology undergraduate and graduate students. Outside of sessions, couples are assigned reading and asked to complete exercises. Markman et al., 1993The Gerontologist, 2019, Vol. 59, No. 4 among families participating in the Family Check-Up program 2 years after the intervention. The Family Check-Up model is based on the transtheoretical model of change, a dynamic and person-centered approach that encourages behavior change in stages based on readiness . Moving forward through stages is believed to improve self-efficacy, which supports behavior change . In the case of the Family Check-Up model, motivational interviewing techniques prepare parents to change reactions contributing to difficult behaviors such as defiance and poor self-regulation . Feedback helps parents recognize areas of strength and areas where they can improve. --- Community-Based Programs to Prevent Intimate Partner Violence Although home visiting approaches are far more common in child maltreatment interventions, some IPV interventions offer valuable lessons especially given overlapping dynamics with EM that are not seen in child maltreatment . Drawing on child maltreatment, the Family Check-Up model has also been effectively revised to prevent IPV in the form of the Men's Domestic Abuse Check-Up, a secondary prevention approach. The program is like the Family Check-Up but provides the intervention via telephone to men who self-identify as engaging in violent behaviors toward their partners. As with the Family Check-Up, motivational interviewing and feedback are central . In addition, because perpetrators of IPV overestimate the normality of violent behaviors, the program provides participants with a brochure detailing the actual prevalence of interpersonal violence within the general male population. Revealing the discrepancy between participants' perceptions of IPV occurrence and what is normal promotes self-awareness and draws upon social learning theory, encouraging men to accept a more accurate perspective of IPV. In addition to or as part of home visiting, interventions to improve communication skills between partners reduce revictimization . In a seminal study, Markman and colleagues tested the Prevention Relationship Enhancement Program, where couples learned techniques such as active listening and expressive communication during therapy sessions . Researchers found that during 4-and 5-year follow-ups, participants in the intervention reported better communication and lower levels of violence in their relationships than controls. The success of these programs may be attributed to their accounting for varying stages of risk, an important factor to consider since potentially abusive behaviors often increase over time . --- Building on Family Violence Interventions to Identify Key Components for a Home Visiting Model to Prevent EM Drawing on lessons learned from child maltreatment and IPV, we argue for the adoption of a home visiting model delivered as a primary prevention intervention for EM. Based on successful models in child maltreatment and IPV the structure of an intervention should target caregivers facing new or changing conditions, take place over several months, and be tailored to individual and family needs, preferences, and culture. Components should include regular assessments, training, and strengthsbased feedback. How such an intervention might look is illustrated in Figure 1. Although not exhaustive, these approaches reflect the most promising practices from family violence. Readers seeking additional information about specific program structure and components should review MacLeod & Nelson and Whitaker and colleagues . Although we propose a standardized structure and several key components for an intervention, the exact intervention tools will depend on specific risk factors in each family. Previously we suggested risks fall into three general theoretical categories-those related to relationship quality, stress, and knowledge. Responses to risk factors identified during assessment should link to selected tools by way of underlying theory. Whereas training for someone who is experiencing high levels of stress might focus on behavior management and assistance accessing respite, caregivers who are struggling with a difficult relationship history will likely receive greater benefits from counseling. However, because there will likely be overlap in these categories, some cases responses will look very similar . Figure 2 illustrates an intervention based on a hypothetical case where stress-based risk factors are central; Supplementary Figures 1 and2 illustrate what an intervention might include for the other theoretical mechanisms discussed. --- Target Caregivers at the Start of Caregiving and Those Facing Changing Conditions As Teresi and colleagues point out, interventions in child maltreatment are most effective when delivered before maltreatment occurs. Similarly, primary EM prevention programs implemented near the beginning of a caregiving journey or at critical junctures after a change in care level occurs may be most effective. For example, an intervention could target caregivers following a hospital transition. However, unlike new parents who readily identify with this role, caregivers who are just beginning this role may require additional guidance during recruitment given a tendency to not yet perceive themselves as caregivers . --- Take Place over an Extended Period Further, in reviews of both child maltreatment and IPV interventions, the most successful home visiting programs were found to be relatively intensive and moderately long in duration. Based on studies from other forms of family violence, regular and frequent visits-perhaps 12 or more -over at least 6 months are most likely to improve the chances of success of a home visiting intervention. Even longer intervention periods may be appropriate, possibly lasting 2 years or more, depending on other design components . --- Tailored to Individual and Family Needs, Preferences, and Culture Interventionists should utilize the flexibility of a multicomponent approach to deliver services and supports that meet the unique needs of families wherever they are on the spectrum between a healthy and abusive relationship. For example, interventionists in the Healthy Start program individualize the intervention to families by guiding them to relevant services based on specifically identified needs, and decrease the number of visits as parent competency grows . Multicomponent options may also be preferable compared to "one-size-fitsall" approaches , as some families will be more amendable to and will engage with some interventions tools more than others. In addition, an EM intervention should meet the needs of culturally and ethnically diverse caregiving dyads and families. The Family Check-Up model's success in racially and ethnically diverse samples has been attributed to its ability to address multiple types of stressors linked to culturally specific norms and attitudes . --- Guided by Regular Assessment To effectively meet specific and evolving needs and identify whether progress has been made toward reducing risk, family assessment at baseline and regularly thereafter is necessary. Assessment, as described here, is both a means of evaluating a program, and a component of the intervention itself. In the Family and Men's Check Up models, assessment is the cornerstone of increased selfawareness underlying behavior change in these transtheoretical model-inspired programs . Although assessment should avoid being burdensome, it is important to include contextual factors such as family dynamics as well as characteristics of the caregiving situation that are likely to evolve . An assessment should inform a care plan that is specific to the needs of the family. . Success is attributed not only to skill-building in areas of weakness but also increase in self-efficacy supporting behavior change. Like training options offered in caregiver and child maltreatment interventions, content for an EM intervention might include practical demonstrations on how to provide care, money management, and managing difficult care recipient behaviors. Training could also address relationship quality, especially in situations involving recipients with dementia or those who may have difficulty appropriately expressing their needs. In line with Promoting First Relationships , relationship quality can be improved by assisting caregivers in understanding recipient needs. Dyad training on communication skills could be useful in some cases . However, because there are likely limitations of such training in families with long-standing relationship conflicts; such training, if used, should include realistic goals as to what can be accomplished . This approach is supported by the transtheoretical model, where progress is tailored to the individual . --- Inclusion of a Training Component --- Provision of Strengths-Based Feedback Strengths-based feedback builds self-efficacy, encourages participant retention, and motivates progress toward acquiring the skills necessary to provide quality care. Such feedback should be provided following assessment, reassessment, and while providing training modules. Components of social learning theory found in the transtheoretical model of change suggest that self-efficacy can encourage engagement in interventions , which has been illustrated in Family Check-Up programs . Although self-efficacy has long been considered an outcome of caregiver interventions , an EM intervention should also consider self-efficacy as a moderator. --- Conclusion Evidence-based primary prevention strategies from other fields of family violence offer a useful starting place for interventions to prevent EM. Of particular interest are home visiting models aimed at caregivers, which offer opportunities and approaches to target underlying risk factors shared among different forms of family violence. We describe starting points for such an intervention but acknowledge that any program will be shaped by the available resources and specific agency goals. Regardless, given what we know about prevention programs from child maltreatment and IPV, we believe there is adequate information to pilot a promising prevention intervention for EM. --- Supplementary Material Supplementary data are available at The Gerontologist online. --- Conflict of Interest None reported.

Elder mistreatment (EM) is a public health problem that harms millions of older Americans each year. Despite growing recognition of its occurrence, there are no evidence-based primary prevention programs. Although EM is distinct from other areas of family violence, including child maltreatment and intimate partner violence, common risk factors and theoretical underpinnings point to opportunities for prevention strategies. Drawing on evidence-based best practices found in other fields of family violence, we identify approaches that could be tested to prevent EM at the hands of family caregivers, who are among the most likely to commit mistreatment. Specifically, we examine home visiting approaches primarily used in the child maltreatment field and identify components that have potential to inform EM interventions, including prevention. We conclude that there is enough information to begin testing a prevention intervention for EM that targets caregivers.

Background Life-long treatment with antiretroviral compounds is currently the gold standard in the clinical management of human immunodeficiency virus infection. While antiretrovirtal therapy is effective in restoring health and minimizing secondary HIV transmission, it falls short of achieving virus eradication. Effective curative treatment could provide a sustainable solution to prevention and control [1]. Despite increasing research with the accumulation of scientific evidence for HIV cure, such therapy is not yet in sight [2]. Despite the uncertainties, a survey of over 400 people living with HIV in the United States showed that more than half were willing to participate in different HIV cure studies [3]. Elimination of the HIV virus in the "Berlin patient" offered new hope in the PLHIV communities about the ultimate goal of cure [4]. De-stigmatization was particularly valued by PLHIV who favored curative treatment when it becomes available [5,6]. In the interim, animal studies and human trials suggested that "functional cure" of HIV infection may become a reality in the coming years [7,8]. Unlike "sterilizing cure", functional cure strategy aims to achieve effective suppression of HIV viral load so that antiretroviral therapy becomes unnecessary. While "functional cure" could represent different strategies, the term is now consistently used in referring to the attainment of virus control without ART [7], while one's HIV infection status remains unchanged. Willingness of PLHIV in receiving non-eradication cure treatment or joining functional cure research would be an important consideration, especially that current generations of antiretrovirals are safe and extremely effective. PLHIV's decision may hinge on one's understanding of the concept of cure and how this is explained [9]. In the United States, many participants in a study did not consider functional cure as an improvement to conventional ART [5]. Experiences with ART may also affect PLHIV's decision about participation in functional cure research. In Hong Kong, ART coverage in PLHIV receiving care at the public service is high. With the increasing reports of the promising outcome of functional cure, it is timely that their attitude towards participation in functional cure research be explored. We hypothesized that the degree of knowledge and attitudes towards HIV cure and functional cure differed between ART-experienced and ART-naïve PLHIV. In addition, education needs in the specific area of functional cure among newly diagnosed patients could be identified, which may in turn ease subject recruitment for a clinical trial, and improve expectation management of participants. As such, we undertook to examine the attitude of MSM living with HIV towards functional cure in Hong Kong, and contrast the awareness and perceptions between newly diagnosed and veteran PLHIV who have been on long duration of ART. --- Methods Data for analyses came from a questionnaire survey administered on ART-naïve and ART-experienced men who have sex with men living with HIV participating in two separate cohort studies in Hong Kong. The former study comprised the collection of data from newly diagnosed MSM attending any one of the three HIV specialist clinics over a two-year period; while the latter study involved treatment-experienced MSM receiving ART at the largest HIV clinic during the same period of time as a follow-up round of a study on the sex networking behaviors following HIV diagnosis [10]. Informed consents were obtained prior to the conduct of the surveys. Approval of the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee and Ethics Committee of Department of Health were obtained. A self-administered questionnaire was designed with mutual topics in both surveys covering demographics and HIV diagnosis year, perceptions about HIV cure and their impacts, and awareness and anticipation of HIV functional cure, considerations of and concerns about participating in an HIV functional cure clinical trial . Functional cure was the specific theme for this survey, the definition of which was clearly described in order to avoid the misinterpretation of the results as there could be confusion between different forms of HIV curative treatment. Hypothesizing that sexually active PLHIV would have a higher awareness, anticipation and interest in HIV functional cure, sexual behaviors in the preceding year including sexually transmitted infection diagnosis, sexual activity, and use of psychotropic drugs for sex , were inquired. Clinical data were separately transcribed, which included the age of HIV diagnosis, acquired immunodeficiency syndrome status, and longitudinal CD4 cell count and viral load. The most recent readings prior to the trials, community education, updated information and appropriate medical advice would be needed. Safety is a major concern for potential enrollees in HIV functional cure trials. Keywords: HIV, Functional cure, Men who have sex with men, Awareness time of questionnaire for the latter two measures were recorded for analyses. On the perceived impacts of HIV cure, participants were asked to choose from seven options: no longer needing to take HIV medications, restoration and stabilization of effective immune function, not getting HIV for a second time, no longer needing to visit a doctor for HIV, no longer at risk of AIDS or HIV-related morbidity, no longer transmitting HIV to the others, and being considered as a person not living with HIV. Awareness of HIV functional cure was categorized into: knew of it and understood what it is, heard of the idea but not the details, and had no idea of it. After giving a brief introduction on HIV functional cure, participants were asked to rate their anticipation on a scale of 0 to 10. Willingness to join an HIV functional cure study was assessed by a 6-point Likert scale . Participants were inquired about the importance of each of the following factors which they would consider in joining an HIV functional cure trial: safety, duration of the clinical trial, incentives for participation, view and support from family and peers, advice from healthcare professionals, credibility of the research institution, and need for interruption of ART medications. Potential adverse consequences of participating in the trial included CD4 count going down, HIV viral load going up, becoming infectious to the others, presence of AIDS or other related complications, and side effects arising from the therapy. The levels of importance and concern were ranked in descending order as: very important/concerned, moderately important/concerned, somewhat important/concerned, and a minimally important/concerned. Differences in perceptions on HIV cure and HIV functional cure trial between ART-naïve and experienced groups were assessed by participants' sociodemographic and behavioral factors in logistic regression. Determinants of anticipation of HIV cure, awareness of HIV functional cure, and willingness to join an HIV functional cure trial were analyzed by univariable and multivariable logistic regression models. Participants were defined as having a high degree of anticipation if they gave a score of 10. Awareness of HIV functional cure was reflected by at least knowing the idea without details. The importance of factors associated with joining functional cure trials and the concerns were dichotomized by defining the lower two levels as zero and the rest as one. Only variables with p < 0.10 were included in the initial multivariable models for stepwise backward elimination based on the AIC value. Inter-correlations between cure-and trial-related consideration factors and concerns were explored using logistic regression models. All statistical analyses were conducted in R. --- Results --- Demographics Between March 2019 and January 2020, 153 treatmentnaïve MSM were recruited, which accounted for about one-third of all newly diagnosed male PLHIV in Hong Kong; while data from 203 treatment-experienced MSM were collected between April 2019 and November 2020. The median age of MSM included in the analyses was 37 years , with treatment-naïve ones at 30 years and treatment-experienced ones at 42 years of age . Among those receiving treatment, the median duration of HIV diagnosis and receiving ART was 9 years and 8 years , respectively. The median age of HIV diagnosis of all enrolled MSM was 31 years . As regards sexual behaviors, about one-third did not have any sexual activities in the preceding year, while 39% had engaged in chemsex, among whom 12% injected drugs. About 90% of treatment-experienced participants have achieved a viral load below 20 copies/mL, while almost all treatment-naïve counterparts gave a value of at least 1000 copies/mL. Regarding CD4 cell count, a smaller proportion of treatment-naïve participants had at least 500 cells/uL, whereas three-quarters in the treatment-experienced group had attained the same level. --- Perceived benefits of HIV cure Perceptions on HIV cure differed between the treatmentexperienced and naïve group . The most commonly perceived impact of HIV cure was "restoration and stabilization of effective immune function" , with lower odds among those on ART compared to ARTnaïve participants . About half considered untransmittability as an important impact of curing HIV infection. Lowered risk of AIDS or related morbidity was considered more important in the treatment-naïve group , whereas reduced need for clinic visit was accorded higher importance among treatment-experienced participants . Those who considered immune function stabilization as an important impact of HIV cure had a lower CD4 cell count and were less likely to have recently been engaging in sex . Perceived importance of reduction of clinic visits was associated with being sexually active and a higher CD4 level . Participants having attained post-secondary education level and younger than 40 years of age regarded being free from the risk of AIDS or HIV-related morbidity as an important impact for HIV cure. Older adults did not consider reduced transmittability important . De-labelling was an important impact for those with a higher education level . --- Awareness of HIV functional cure Overall, less than half of the participants were aware of HIV functional cure. Compared with those having no knowledge of it, treatment-experienced MSM were more likely to have heard of HIV functional cure but without detailed knowledge . Awareness was associated with STI history in the past year , and being sexually active . Those who were aware of HIV functional cure did not consider restoration and stabilization of effective immune function an important impact , nor were they concerned about potential adverse effects of therapies for functional cure . --- Anticipation of HIV functional cure Among all participants, the median anticipation score for HIV functional cure was 10 with more than half giving a score of 10, and it did not differ between ART treatment-naïve and experienced participants. Among treatment-experienced MSM, those having been diagnosed with HIV for more than 8 years had higher odds of scoring 10 for their level of anticipation for HIV cure . Those anticipating HIV functional cure were more likely to be of age 50 years or above , and to have a lower awareness of the idea . Higher level of anticipation was associated with the consideration of not needing to take long-term HIV medications following functional cure but negatively with untransmittability of HIV . --- Acceptance of an HIV functional cure trial Should there be an HIV functional cure trial, the vast majority would consider joining. Therapy safety , advice from healthcare professionals , and credibility of the research institution were the top three factors participants took into account when considering whether to join such a trial. Participants interested in joining the trial were more likely to have been diagnosed with an STI in the past year , and to take advice from healthcare professionals when making such a decision . When deciding whether to participate in an HIV functional cure trial, view and support from family and peers were important for those who ceased to have sex and who had attained lower education level . Credibility of the research institution was important to those who had received higher education . Participating in such a trial is not without concerns. Major concerns included the potential risk of developing AIDS and complications , HIV viral load going up and adverse effects of the therapy . They were concerned about adverse effects on their CD4 level , HIV viral load going up and becoming infectious following functional cure therapy. Considering trial duration important was associated with concerns about adverse effects . Participants who considered trial safety important were concerned about all five potential adverse consequences of participating in the trial. Those looking forward to the prospect of restoring effective immunity following HIV cure had concerns about CD4 count going down after joining the trial . --- Discussion HIV functional cure is a relatively new concept for PLHIV. Despite their generally low awareness, MSM living with HIV had, as shown in this study, a high level of anticipation for HIV functional cure, especially those who had been diagnosed and put on ART for some years. The effective achievement of functional cure carries the potential prospect of getting rid of repeated clinic visits and taking long-term medications [11]. The benefit of being untransmittable may, however, not be regarded as an important impact for HIV cure because in the current era of "undetectable equals untransmittable", adherence to ART is already effective in minimizing the risk of transmission. In our cohorts of MSM living with HIV, the attitudes differed by one's ART treatment experience. Treatment-naïve participants were more concerned about their immune function and AIDS progression than those ART-experienced, as deteriorated health was perceived by newly diagnosed patients an imminent consequence brought on by HIV [12]. Anticipation for HIV functional cure was apparently lower among those who were aware of it. Although some 42% had heard about functional cure, most had limited knowledge of it and cared little about the potential benefit of restoring their immune function and potential adverse effects arising from the new therapy. With the increasing likelihood that functional cure may soon become the next major advance in HIV treatment [7], there is the need for promoting community education on the rationale and strategy for HIV functional cure. As in many countries, MSM accounted for a high proportion of PLHIV in Hong Kong [13], and who continued to be significantly impacted by the epidemic [14]. The MSM community, especially those living with HIV would need to be targeted in anticipation of the introduction of functional cure therapy. In our study, MSM engaging in higher risk sexual behaviors were more likely to be aware of the recent advancement on HIV science, such as pre-exposure prophylaxis [15]. This echoed our results that sexually active PLHIV, particularly those engaged in condomless sex as indicated by their STI history, were more likely to be aware of HIV functional cure. There are diverse reasons for PLHIV to look forward to receiving functional cure therapy. It is intuitive that PLHIV with a weaker immune system were less likely to be sexually active and they wished to restore and stabilize their immune function and were at the same time more concerned about adverse HIV outcomes after joining functional cure trials. The fact that younger PLHIV cared more about disease development and transmittability was likely because they were newly diagnosed and would remain sexually active for a longer time in the future. PLHIV with higher education level regarded de-labelling of HIV an important impact of HIV cure. In Hong Kong, public stigma against PLHIV [16] and perceived discrimination [17] have been reported. It is therefore an ideal vision for them to remove the label of HIV after achieving some sort of cure. Discrimination in workplace and difficulty in career development could be experienced by PLHIV [18]. Although highly educated PLHIV were found to deal better with workplace discrimination and improve employment quality than those with lower education level [19], they were more concerned with their career development that removing stigma means removing a barrier in their career paths. From the perspective of the general public, increased awareness and knowledge in HIV science advancement might be able to reduce stigma. A previous study showed that heterosexually active adults who were aware of the untransmittable nature of PLHIV achieving undetectable viral load had a lower level of anticipated HIV stigma [20]. As this study focuses on PLHIV, a follow-up study in the general population and the MSM community would be needed to assess the impact of functional cure on general, and dating-and sex-related enacted stigmas. Overall, the idea of an HIV functional cure trial was well accepted by our cohorts of MSM living with HIV with over 90% indicating their willingness to join, similar to the high acceptance rate of 95% in a previous study [11]. Potential trial participants were more likely to have engaged in higher risk sexual activities as reflected by their report of STI in the preceding year. This might be precipitated by a prospect of continuation or re-engagement in adventurous condomless behaviors upon cure [21]. Safety is a major concern for clinical trials involving the use of novel agents like neutralizing antibodies or therapeutic vaccines. Careful monitoring of PLHIV in the trial would be warranted not just for protecting participants' health but also easing their concerns of side effects and adverse HIV outcomes. Hesitation would arise if the trial requires interruption of ART or if it takes a long time, in relation to the risk of rebound and potential adverse effects [11]. Trusting relationship between PLHIV, researchers and healthcare workers could be beneficial for ensuring the delivery of optimal health outcomes [22]. Based on potential participants' trust on healthcare professional advice, collaboration with HIV clinics on explaining the details and safety issues of the trial would be paramount. This study carries some limitations. Like other epidemiological studies, there were inherent recall and social desirability biases in the administration of behavioral questionnaires. We had minimized these biases by adopting a shorter recall period and omitting fine details about sexual activities. The adoption of a self-administering approach should have reduced embarrassments in response to interviewers' questions. Incorporation of clinical data by transcription benefited the study by providing objective measurements of participants' HIV outcomes. In the absence of any functional cure trials, the exploration of one's willingness to participate in a hypothetical trial without having one in place could be postulational. Nevertheless, the systematic inquiries into the awareness and anticipation of functional HIV cure among MSM living with HIV and their inclination in joining a hypothetical study have generated useful results which could be of reference for the planning of a functional cure trial. --- Conclusions To PLHIV, HIV cure means restoring and stabilizing their immune systems that they no longer need to take longterm medications, nor do they need to visit HIV clinics repeatedly. While virus eradication cannot be achieved, HIV functional cure is a promising strategy in which MSM living with HIV had a high level of anticipation despite relatively low awareness. While a high acceptance rate of a functional cure clinical trial was elicited, MSM living with HIV were concerned about adverse HIV outcomes. Their appreciation of the objectives of functional cure, understanding of the study procedures, and recognition of potential adverse events are crucial, and all of which need to be accessible well before a functional cure trial begins enrolment. --- --- Abbreviations HIV: Human immunodeficiency virus; PLHIV: People living with HIV; MSM: Men who have sex with men; STI: Sexually transmitted infection; AIDS: Acquired immunodeficiency syndrome; IQR: Inter-quartile range; OR: Odds ratio; aOR: Adjusted odds ratio; ART : Antiretroviral therapy. --- Supplementary information The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12879-022-07346-x. Additional file 1. Questionnaire. S1. Crude odds ratio for factors of importance to be considered and concerns related to participation in functional cure trial among MSM . Table S2. Crude odds ratio for important factors to be considered about and concerns related to the trial among MSM . --- Additional file 2: Table --- Authors' contribution The study was conceptualized by SSL. Data was collected by THK, CPC and NSW. Data analysis was performed by CPC. Results were interpreted by all authors. The first draft of manuscript was written by THK. All authors participated in the critical revision of the manuscript, and read and approved the final manuscript. • 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 --- • At BMC, research is always in progress. --- Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: had no role in the study design; in the collection, analysis and interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication. --- --- --- Competing interests The authors declare they have no competing interests to report. --- Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Background: Human immunodeficiency virus (HIV) functional cure is a novel biomedical strategy characterized by sustained viral suppression without the need for life-long medications. The attitude of people living with HIV (PLHIV) towards functional cure and clinical trials are understudied. We aimed to examine the awareness and levels of anticipation for HIV functional cure among men who have sex with men (MSM) living with HIV, and their willingness to join trials as differentiated by their antiretroviral treatment status. Methods: MSM living with HIV with and those without treatment history were recruited from Hong Kong's HIV specialist clinics. Self-administered questionnaires covering behavioral profile, perceived impact of HIV cure, attitude towards HIV functional cure and related clinical trials were collected. Clinical data were separately transcribed. Determinants of perceptions and attitudes were identified by logistic regression models. Results: Of 356 MSM living with HIV recruited, less than half (42%) were aware of HIV functional cure, but they had a high level of anticipation for it. Treatment-experienced participants were more likely to be aware of HIV functional cure. Awareness was associated with continued engagement in sexual activities after HIV diagnosis and sexually transmitted infection (STI) diagnosis. Higher anticipation was observed among older MSM living with HIV but it was negatively associated with one's awareness. Over 90% were willing to join functional cure trials, especially those who had previously been diagnosed with STI and had engaged in chemsex in the past year. Advice from healthcare professional was an important factor considered by those willing to join clinical trials. Younger, better educated MSM, and those with lower CD4 counts were more concerned about potential risk of AIDS and potential complications upon trial participation. Conclusions: MSM living with HIV, especially those sexually active, showed positive attitude towards functional cure and willingness to join related clinical trials despite low awareness. To enhance preparedness for HIV functional cure

INTRODUCTION European populations are ageing faster than populations in many other parts of the world. 1 Despite a continued increase in healthy life expectancy, a substantial proportion of 'people in early old age' in Europe suffer from some degree of impairment which often limits their work capability. 2 In fact, reduced health functioning and disability are major determinants of involuntary early exit from the labour market. 3 4 In view of the strong economic need for extending labour market participation of older people in European countries, 2 it is important to know more precisely the sociodemographic, socioeconomic and workrelated characteristics of older men and women whose work capability is limited due to impairment or disability. Such knowledge may provide useful information on entry points for preventive measures within and beyond occupational settings. Previous investigations documented a social gradient of disability prevalence in early old age, leaving men and women in lower socioeconomic positions at higher risk. [5][6][7][8][9][10] However, given the bidirectional pathways between health functioning and socioeconomic position, including employment status, in a life course perspective, 3 a clear-cut interpretation of this association is difficult. Moreover, measures of socioeconomic position in terms of occupational position or earnings are often confounded by distinct occupational exposures , which may trigger early onset of disability. 11 It is therefore important --- What the paper adds ▸ Previous investigations demonstrated a social gradient of disability in older populations, but were restricted to single countries or cohorts, were mostly cross-sectional, or did not use a comprehensive measure of disability. ▸ In this study, we followed the guidelines of the International Classification of Functioning and used multiple variables to construct and validate two different disability indices in the frame of a longitudinal cross-country study. ▸ Our findings based on employees aged 50-64 years from 11 European countries demonstrate a social gradient of both indices of disability which is in part mediated by adverse psychosocial working conditions. ▸ Investment into reduction of psychosocial work stress, in particular among employees in lower socioeconomic positions, is a relevant entry point of policies that aim at maintaining work ability in the ageing workforce. to disentangle these effects and to test their robustness in a large dataset which offers opportunities of adjusting for relevant confounders. In this study, we set out to test associations of sociodemographic, socioeconomic and work-related factors with disability in a prospective study design, using strictly comparable indicators, based on a large sample of male and female employees in early old age from 11 European countries. More specifically, we test the hypothesis that low socioeconomic position and exposure to a stressful psychosocial work environment contribute to later disability . Furthermore, we claim that the association of socioeconomic position with disability is partly mediated by exposure to a stressful psychosocial work environment, as measured by core components of two established theoretical models: the demand-control 12 and the effort-reward imbalance 13 models. The demand-control model defines work stress in terms of a distinct job task profile where jobs defined by high quantitative demands in combination with low decision latitude or low task control are stressful. 12 A complementary model, ERI, focuses on the work contract and the principle of social reciprocity lying at its core. 13 Rewards received in return for efforts spent at work include money, esteem and career opportunities . The model asserts that lack of reciprocity occurs frequently in modern economies and generates strong negative emotions and psychobiological stress responses with adverse long-term effects on health. A second aim of this study concerns the development of indices of core dimensions of disability, taking note of a more comprehensive approach to this notion, as documented in basic work resulting in the International Classification of Functioning, Disability and Health . 14 Disability is no longer interpreted as a fixed attribute of an individual, but rather as a dynamic continuum of experiences at different levels . As disability results from an interaction of vulnerable individuals with a broad range of environmental factors including working conditions, 15 16 it seems mandatory to assess this condition in a multidimensional, quantitative frame of analysis. At least two crucial components need to be distinguished, the level of bodily impairment and the level of restriction in activities and social participation, since 'people with the same impairment can experience very different types and degrees of restriction, depending on the context'. 16 A person with an amputation, may for instance, not be restricted in activities and participation when she is provided with appropriate prostheses and supports. Another reason for this distinction is related to their potentially different role in predicting morbidity and mortality. In a recent study of older British women, it was observed that restrictions in participation, and restrictions in complex activities, were associated with increased risk of mortality, but no such effect was found for impairment in the fully adjusted models. 17 Therefore, embedded in the framework of ICF, we aim at developing a comprehensive continuous measure of core dimensions of disability, based on available data. --- METHODS --- Data source Data were obtained from the first two waves of the 'Survey of Health, Ageing and Retirement in Europe' . SHARE is the first longitudinal research project comparing data on working conditions, retirement, health and well-being, and social position among people aged 50 years and older in a variety of European countries. 18 Data from the two first waves include 11 European countries . Data collection is based on probability household samples where all people above 50 years of age, plus their partners, were interviewed using Computer Assisted Personal Interviews. Due to different institutional settings in the participating countries, the sampling was either drawn as a stratified simple random selection from the national population register , or as a multistage sampling using regional and local registers or telephone directories . In SHARE, information is available for 28 517 respondents in wave 1, where the country average of household response rate is 60.6% for the total sample, ranging from 38.8% in Switzerland to 79.2% in France . In wave 2, 18 742 of these individuals remain in the sample, with an attrition rate of 27.9%. 35 Because we were interested in associations between working conditions in late midlife in wave 1 and disability in wave2, we excluded people not aged 50-64 years at wave 1 , and those reporting no employment at baseline from the longitudinal sample. Moreover, we conducted a complete case analyses, thus excluding respondents with missing data on any of the variables . This results in a final sample size of 4864 respondents. In order to compensate for unit non-response, and for attrition between the first and the second waves, calibrated longitudinal weights were applied. These weights are defined for the longitudinal sample only and are calculated for each country separately . --- Measurement Disability Based on ICF and its definition of disability, we divided between two indices of disability: one measuring impairment and another measuring restriction in activity and participation . 20 Accordingly, all SHARE modules were screened for questions related to disability, and respective items were linked to the two dimensions using established linking rules. 20 21 Since most original items were dichotomous, we dichotomised the remaining items, such that one category was indicative of having at least a moderate problem in its respective domain. As a next step, we applied a principal component analysis to evaluate the chosen items based on wave 1 data . In the results, two factors with eigenvalues above 2.0 were extracted confirming the assumed dimensions of disability. However, two candidate items for the A&P factor did not load as expected and were excluded from further analysis. Not considering the dropped items, the two factors explained around 34% of the variance of all items, and 45% of the A&P, and 23% of the impairment items were explained by the two-factor solution. Next, we performed a confirmatory factor analysis for item responses at wave 2 using the STATA module confa. 22 The first item of the respective index was used as scalar . Results show that our index solution meets the statistical criteria. However, one item on the impairment factor and two items on the A&P restrictions factor displayed marginal loadings only, probably due to their low prevalence, but were, nevertheless, included in the final index solution. In case of A&P restrictions, it should be noted that the included items turned out to be part of two other existing scales . Yet, not all items of these two latter scales were included, and instead of dichotomising both scales into 'No limitations' versus 'One or more limitations' , we created one single scale with additional information on the number of A&P limitations. We constructed sum indices for impairment and A&P restrictions 21 . Each of those is a count variable ranging from 0 to 12 . We plotted both indices against a Poisson distribution which was confirmed in both cases. Expectedly, the impairment scale showed a high correlation with subjective health , while the A&P scale showed a lower correlation of r=0.17. Both scales were correlated with r=0.2. This demonstrates convergent as well as divergent validity, that is, the scales seem to measure different constructs. --- Working conditions Quality of work was assessed by a short battery of items derived from the Job Content Questionnaire measuring the demandcontrol model 23 and from the ERI model questionnaire 24 . The psychometric properties of both questionnaires were previously tested. 24 Given the constraints of a multidisciplinary approach in SHARE project, the inclusion of the full questionnaires was not possible. Thus, items with best psychometric properties in their respective scales were selected. With regard to the first model, the measurement was restricted to the control dimension. This decision was based on evidence that the predictive power of 'control' by far exceeded the power of 'demand', and that tests of the interaction term 'demand'×'control' had produced inconsistent results. 25 Low control at work was measured by the sum score of two Likert-scale items ranging from 2 to 8, with higher scores indicating lower control at work. Scores in the upper tertile of each country were defined as representing poor quality of work in terms of low control. To measure ERI, two out of six items measuring 'effort', and five out of 11 items assessing 'reward' at work were included. For the selected items, all-item total correlations were far beyond the established threshold of 0.30, 26 ranging from 0.93 to 0.81 , and from 0.67 to 0.42 . 'Effort-reward imbalance' was defined by a ratio of the sum score of the 'effort' items and of the sum score of the 'reward' items . As previous analyses showed that quality of work in terms of this model varies across countries under study in SHARE, tertiles of the ratio were calculated for each country separately. 24 Participants scoring in the upper tertiles of this ratio of imbalance were considered experiencing poor quality of work. --- Additional measures Age, gender, income and education were included as additional measures. Income information is based on the total annual household income composed by the sum of different income components that were assessed in the questionnaire. In case income components were missing, information was obtained through imputation. 27 To adjust for household size, we divided the value of income in accordance with the OECD equivalent scale, and categorised it into country-specific tertiles based on all available cases in each country . Education was measured according to the International Standard Classification of Educational Degrees that was categorised into 'low education' , 'medium education' , and 'high education' . These two indicators are introduced as proxy measures of socioeconomic position, but are analysed separately as they point to different dimensions of social inequality. 28 --- Analyses As a first step, table 2 gives an overview of the study sample and the core variables. Additionally, associations between disability and all covariates are analysed using analysis of variance. Thereafter, we calculated Poisson regression models to predict disability 20 in wave 2. Given the multilevel structure of the data, we estimated a random intercept multilevel model with individuals nested within countries . 29 This model contains two components: a so-called fixed component that explains systematic variability in the data, and a random component which accounts for unobserved variability between countries . This procedure allows for accurate adjustment for country affiliation. Maximum likelihood estimation is used for parameter estimation. In the Results section, for each one of the two disability outcomes, we display findings based on three consecutive models, all adjusted for age, gender and level of impairment at wave 1. This procedure allows for exploring to what extent the predicting variables are associated with change in disability between wave 1 and wave 2. 30 Model 1 investigates the joint effect of income and education on disability. In model 2, the joint effects of the two work-stress measures are presented. Model 3 includes all variables of the first two models, with the aim of examining potential mediation effects. In respective tables, we present incidence rates ratios together with the level of statistical significance and CIs. Calculations were done using the 'xtmepoisson' procedure in STATA 11. --- RESULTS --- Descriptive findings In table 2, sample characteristics are given, and mean scores of the two disability factors, impairment and A&P restrictions, are displayed according to covariates. A total of 2655 men and Concerning the scores of impairment and A&P restrictions, we first observe that overall levels of A&P restrictions are much lower than those of impairment scores. Impairment scores varied according to gender , age and socioeconomic position . As shown in figure 1 , this social gradient is consistent across all countries, using income as an indicator of socioeconomic position. Less clear findings are obvious in case of A&P restrictions. As expected, we observe higher scores of the two disability factors among men and women experiencing chronic stress at work in terms of the two work-stress models. Again, relatively consistent trends are observed across the countries under study, as exemplified by measuring work stress in terms of ERI . --- Multivariate findings In tables 3 and 4, results of multivariate analyses are presented. In both tables, we present findings of the three regression models described in the Methods section, first for impairment, and second for A&P restrictions. Significant effects of gender , age , socioeconomic position , and work stress on impairment scores were confirmed in multivariate analysis . If the indicators of socioeconomic position and the two work-stress models are included simultaneously , we observe an attenuated effect of the two indicators of socioeconomic position on impairment, pointing to a partial mediation of the association between socioeconomic position and impairment by a high level of work stress. Results are not as consistent in case of A&P restrictions . Effects of gender and age are not significant in the multivariate model, whereas the remaining effects correspond to those of impairment. --- DISCUSSION Based on two summary indices of impairment and restrictions of activities and participation, we tested sociodemographic, socioeconomic and work-related predictors of disability among working men and women of early old age across 11 European countries, using data from two waves of SHARE study. After adjusting for baseline disability and relevant confounding variables, low socioeconomic position and chronic stress at work exerted significant effects on disability scores 2 years later. Effects were more consistent for impairment than for A&P restrictions. In addition, we found some support of the hypothesis that the association of socioeconomic position with disability is partly mediated by work stress . To our knowledge, no former study analysed these associations in a prospective design comparing data across a variety of European countries. A further innovative aspect of this investigation concerns the construction of linear indices of two core components of a comprehensive concept of human functioning that was developed in the frame of ICF. 14 Previous studies found similar associations of socioeconomic position with measures of disability, but were restricted to single cohorts or single countries. [5][6][7][8][9][10] Huisman, 5 for instance, found that inequalities were lowest in the oldest age group, particularly among women. However, due to the cross-sectional nature of this study, it cannot be determined if this is an ageing or cohort effect. Moreover, it is unclear if the reported socioeconomic differences in long-term disability are in part due to differential working conditions. Similarly, a few investigations tested prospective associations of job strain 31 32 or ERI [31][32][33] with reduced functioning, often measured by the SF-36 or short versions. 34 Stansfeld 31 found, for example, that high demands and ERI and negative aspects of close relationships were independent predictors of poor SF-36 functioning. However, other indicators of socioeconomic position, such as income and education, were not analysed. In this study, we add information on the combined effect of socioeconomic position and work strain on two core indicators of disability, and further on a partial mediating role of work stress in this former association. Several limitations need to be mentioned. First, most items we used for construction of the disability indices were based on self-reported data rather than on functional testing. Therefore, the validity of our indices may be limited, even more so due to potential reporting bias . However, graded associations with age and, in case of impairment, consistent variations with gender and socioeconomic position point to a valid assessment. As reporting bias is assumed to vary across countries it is unlikely that it inflates the findings of the total sample. Nevertheless, future studies should aim at integrating more 'objective' functional measures and determining differences between self-reported and observational data. A second limitation concerns the assessment of stressful work with abbreviated scales. A test of the full models using original measures of all respective scales may reduce the risk of underestimating their effects on prospective disability. Third, we restricted our analyses to the assessment of work stress at wave 1. Additional information on the duration of exposure available from the assessment of work stress at wave 2 could enrich the current analyses. Also, we did not consider data on non-work-related or extracurricular exposures or information on the cause of impairment, for example, injury versus disease. This will be possible with data from SHARE wave 3 that was focused on retrospective assessment of the participants life courses. Since previous analyses indicate that healthier people were more likely to participate in wave 2, 35 we also cannot rule out some selection bias, where people with low levels of disability are more likely to participate in survey research. However, this may rather lead to an underestimation of disability levels and its association with work stress due to health selection. Furthermore, while our analyses focussed on individual predictors, future research may also explore the role of distinct national disability policies and their interactions with working conditions. 2 36 Similarly, albeit the conducted multivariate analyses account for country affiliation, future research may invest in more detailed differences between countries and measurement equivalence of our two indices. In this regard, additional explorative principal component analysis by country showed that A&P items mostly loaded on the same factor which did not hold true for several impairment items. These limitations are balanced by several strengths. First, the SHARE study applied strictly comparable procedures in data collection based on identical measures and study designs across all countries. 19 Second, by applying multilevel statistical modelling in the analysis of data from two measurement waves, by including relevant confounding factors in multivariate analysis, and by providing quantitative indices of two relevant dimensions of disability, our results can be considered rather robust, even more so as they are based on a large sample representative of working men and women aged 50-64 years within the respective countries. Third, our analysis is based on two established, theoretically grounded notions of stressful work, lack of control and failed reciprocity between effort and reward. 37 Both models were previously tested in a range of prospective cohort studies and were shown to contribute to the prediction of stress-related disorders, such as coronary heart disease 38 and depressive episodes. 39 Available evidence on health-adverse effects of an adverse psychosocial work environment strengthens the significance of current findings with regard to reduced functioning in terms of impairment and restricted activities and social participation. Furthermore, additional analyses based on SHARE data demonstrate that the experience of stressful work in terms of these models increases the likelihood of intended early retirement. 37 In conclusion, low socioeconomic position and stressful work are prospectively associated with two core indices of disability in a large cohort of working men and women of early old age from 11 European countries. Investing in good quality of work and reducing social inequalities in health and functioning are relevant entry points of policies that aim at maintaining work ability in early old age. --- Competing interests None. Patient consent Obtained. --- Ethics approval Ethical commissions of the study countries. Provenance and peer review Not commissioned; externally peer reviewed.

Objectives Prevention of disability in the ageing workforce is essential for sustaining economic growth in Europe. In order to provide information on entry points for preventive measures, it is important to better understand sociodemographic, socioeconomic and workrelated determinants of disability in older employees. We aimed to test the hypothesis that low socioeconomic position and exposure to a stressful psychosocial work environment at baseline contribute to later disability. We further assumed that the association of socioeconomic position with disability is partly mediated by exposure to adverse working conditions. Methods We studied longitudinal data from the first two waves of the Survey on Health, Ageing and Retirement in Europe comprising 11 European countries. Sociodemographic, socioeconomic and work-related factors (low control, effort-reward imbalance) and baseline disability of 2665 male and 2209 female employees aged between 50 and 64 years were used to predict disability 2 years later. Following the International Classification of Functioning (ICF), disability was subdivided into the components 'impairment' and 'restriction in activities and participation'. Two multilevel Poisson regressions were fitted to the data. Results After adjusting for baseline disability and relevant confounding variables, low socioeconomic position and chronic stress at work exerted significant effects on disability scores 2 years later. We found some support for the hypothesis that the association of socioeconomic position with disability is partly mediated by work stress. Conclusions Investing in reduction of work stress and reducing social inequalities in health functioning are relevant entry points of policies that aim at maintaining work ability in early old age.

should always be made in education about EVD with a special focus on districts neighboring countries where the disease is consistently being reported. Keywords Ebola virus disease, Qualitative research, Focus group discussion, Cross sectional study, Rwanda --- Background Ebola Virus Disease , formerly known as Ebola hemorrhagic fever, is a severe and often fatal illness that affects humans. It was first discovered in 1976 near the Ebola River in what is now the Democratic Republic of the Congo [1]. The average EVD case fatality rate is around 50% and the case fatality rates have varied from 40 to 90% in past outbreaks [2]. Originally transmitted to humans from wild animals, the spread of EVD is driven by human-to-human transmissions [3]. Consequently, community engagement is pivotal in the fight against EVD and successfully containing and controlling outbreaks. For effective outbreak control, the provision of comprehensive interventions including case management; infection prevention and control practices; surveillance and contact tracing; reliable laboratory services; safe and dignified burials and social mobilization are required [4,5]. According to the World Health Organization , EVD vaccines have been used to help control the spread of Ebola outbreaks in Guinea and in the DRC [6]. There are currently two treatments approved by the U.S. Food and Drug Administration to treat EVD caused by the Ebola virus, species Zaire ebolavirus, in adults and children. The first drug approved in October 2020, Inmazeb ™ , is a 'cocktail' combination of three monoclonal antibodies. The second drug, Ebanga ™ , is a single monoclonal antibody and was approved in December 2020 [7,8]. The 2014-2016 West African EVD outbreak was the longest, deadliest, and most complex outbreak since the discovery of the virus in 1976 [5,9]. It was not until March 29, 2016, that the WHO lifted the Public Health Emergency of International Concern status of this outbreak. With an estimated total cost of 4.3 billion USD, this epidemic resulted in 11,310 deaths and 28,616 laboratory confirmed cases. This was reflected by the substantial decrease in economic output in the three affected countries [2,10,11]. Guinea subsequently declared a new Ebola outbreak in February 2021, necessitating the establishment of preventative measures and mandates among its neighboring countries [12]. Since August 2018, the DRC has been facing a largescale EVD outbreak in the eastern provinces of North Kivu and Ituri spanning as far as Goma and South Kivu provinces [13]. The DRC announced its tenth outbreak on August 1, 2018. Between April 30, 2018, and November 17, 2019, 3296 cases and 2196 mortalities were reported among the five most affected areas, those being Benin with 697 cases, Katwa with 674, Mabalako with 416, Butembo with 288, and Mandima with 344. This outbreak has become the second largest EVD outbreak documented followed only by the 2014-2016 West African outbreak [14]. The unprecedented growth in cases prompted the WHO to declare the Ebola outbreak in DRC a Public Health Emergency of International Concern [15,16]. From 1 June 2020 to 18 November 2020, a total of 130 EVD cases including 119 confirmed and 11 probable cases were reported in the country [17]. In February 2021, a new EVD case was detected in Butembo, a city in DRC's North Kivu Province [18]. Unstable conditions due to armed conflict [19], outbreaks of violence, and social/ economic problems in affected areas complicated the public health response and increased the risk of disease spread both locally within the DRC and to other countries in the region including Uganda, Rwanda, Burundi, Zambia, South Sudan, and Central African Republic [20]. On 20 September 2022, Uganda declared an EVD outbreak caused by the Sudan Ebolavirus species, after the confirmation of a case in Mubende district in the central part of the country. During this outbreak, a total of 164 cases , 77 deaths and 87 recovered patients were reported from September 20 to January 10, 2023 [21]. To date, Uganda has reported four SUDV outbreaks in 2000, 2011, two in 2012, and one in 2022 [22,23]. The increased frequency of EVD outbreaks in DRC, Guinea and Uganda might be attributed to increased human-wildlife contact caused by extensive deforestation, hunting, and mining practices , which demonstrates a need for a robust monitoring system to inform future preventative policy measures [14,24]. Rwanda's dense population and frequented public transportation system render it highly susceptible to the rapid spread of EVD. Sharing borders with the DRC and Uganda increases Rwanda's risk of cross border EVD transmission due to the unmonitored cross-border movement of people and goods and weak border control systems. The increase in magnitude of the 2019 DRC EVD outbreak in Goma and the Mwenge Health Zone in South Kivu further increased the risk of cross border EVD transmissions in Rwanda. Borders between Rwanda, DRC, and Uganda are particularly vulnerable as they are only separated by border check points. These borders, especially those shared with DRC, are characterised by intense trade activities and a high volume of population movement/ migration. The threat of cross-border EVD transmissions is an emerging and increasing cause for concern not only for Rwanda but Burundi as well. Rwanda's constant vulnerability to cross border EVD transmission necessitates strengthened EVD preventative and preparedness interventions, greater cross-border collaboration, and the flexibility to adapt to emerging needs [25]. The Government of Rwanda, through the Rwanda Biomedical Centre, launched the Multi-sectorial Risk Communication and Community Engagement Strategy for Ebola Virus Disease Preparedness and Response in August 2018 [25]. The strategy seeks to contribute to the national preparedness and response plan to halt EVD transmission in Rwanda, through effective evidence based social mobilization, community engagement, and public education that advocates for socio-behavioural changes. Although Rwanda remains free of Ebola, its cost-effective control strategy requires an up-to-date understanding of peoples' knowledge, beliefs, attitudes and behavioural patterns surrounding the virus. Apart from unpublished descriptive technical reports, this survey is the first to assess EVD-related knowledge, attitudes, and practices at individual, institutional, and societal levels in Rwanda. --- Methodology --- Study design This mixed-method cross-sectional survey was conducted from May to July 2021 in five selected districts from Rwanda . These districts were purposely selected from the districts identified as high-risk zones given their geographical proximity or with direct air links to DRC and Uganda. --- Quantitative approach This research was carried out in the initial phase of the COVID-19 pandemic, and due to the Government of Rwanda's COVID-19 travel restrictions, all interviews were conducted over the telephone by a trained member of the research team. The research team consisted of two groups. The first group was composed of data collectors who were stationed in all five study districts. Their role was to sample the study participants, obtain informed consent, and collect participants' contact information. The second group consisted of enumerators trained to administer the questionnaire. Later, the enumerators called the participants to conduct the interviews. Participants were randomly selected from urban and rural sectors in each village. The study participants were parents/ caregivers or household members of various backgrounds and occupations. One participant was randomly selected from each household selected for the study. Vulnerable groups such as child headed families/households , people living with disabilities, and others were deliberately selected and added to the study cohort. Participants who could not communicate, below 18 years old, were excluded from the study. The sampling frame in each village was developed with the help of community leaders. Assuming 50% of the target population have an attribute of interest, a minimum sample of 1000 was found to be adequate at 95% confidence level, 4.95% level of precision and considering a non-response rate of 15% [26]. In all five districts, the research team contacted a total of 1010 participants using a structured questionnaire: Gasabo , Karongi , Rubavu , Rusizi , and Burera . The questionnaire was originally designed in Kinyarwanda and was later translated to English by two bilingual experts, before the study team members validated the questionnaire. In order to assess the convenience and interpretation of the questionnaire, a pilot study was carried out on 35 participants from the general population and the questionnaire was modified accordingly. The questionnaire was predominantly comprised of: --- Qualitative approach A purposive sampling strategy was used to select KIIs and FGDs participants from the list of contacts provided by the district-based data collectors. A total of 98 people participated through KIIs and FGDs. The reasons for selecting these participants were mainly to ensure diversity in our cohort , availability and accessibility of participants, and willingness to openly express their opinions and experiences. In each district, KIIs were completed with teachers, Early Childhood Development care givers, Community Health Workers and Health workers in health centres. At the national level, KII participants were representatives from Government institutions and their key collaborators. All interviews were conducted over the telephone, or via the Zoom platform. In each district FGDs were conducted with community leaders, parents/ caregivers, and adolescents. Each FGD consisted of four people and were conducted through conference calls. For the quantitative study, the KIIs and FGDs guides were composed of questions related to knowledge and awareness surrounding the cause of EVD, signs and symptoms; Risk perceptions and beliefs; Behaviours' and practices; Information communication channels and sources'; Recommendations and vaccination Issues. --- Data analysis Quantitative data were collected using Kobo Collect Software and then converted to an excel sheet. Analysis was carried out using Statistical Package for Social Sciences version 20 and Python version 3.9.7. Chi-squared tests were used to determine the association between survey responses and district, marital status, and education level of the respondents. A logistic regression model was used to assess the relationship between the question "Do you think Rwanda is at risk of EBV an outbreak?" and the cofounder factors age, gender at birth, district, level of education, and marital status. The minimum p value for significance was considered 0.05. Qualitative data was managed using NVIVO 8 software. Interview transcripts were imported into NVIVO 8 for coding and subsequent analysis [27]. To ensure intercoder reliability and agreement, coding was performed by two researchers. Inter-coder agreement remained above 90% throughout the analysis. --- Results --- Quantitative findings --- Characteristics of the study population Out of 1010 respondents surveyed across the five districts, 20.5% were in Rubavu, 20.3% in Gasabo, 20.01% in Karongi, 20.0% in Burera and 19% in Rusizi. In all surveyed districts, 56% of respondents were male, ranging from 49.2% , 52.7% , 57.9% , 59.1% to 60.4% . For the total sample, approximately 70.3% of participants were married, 25.3% were never married, and the most respondents reported that they had obtained a primary level of education . --- Assessment of knowledge and awareness on Ebola The overall Ebola knowledge levels disaggregated by district are shown in Table 2. The analysis of these data suggests relatively high levels of awareness and knowledge pertaining to Ebola across the five study districts albeit with variations in geographical proximity to affected regions of DRC. The study participants responded that they receive Ebola-related awareness and knowledge from multiple sources, some of which are prevalent across the population sub-groups. The results are confirmed by the low p values obtained in a chi-squared test comparing answers from different districts, which indicates heterogenous answers across different places of the country. For instance, when people were asked if EVD is caused by a virus the highest proportion of correct answers was found in Rubavu whilst the lowest was found in Burera . This difference is computed by a low p value . --- Assessment of attitudes towards Ebola infected and affected persons Overall, the respondents in all districts showed a positive attitude towards EVD survivors. For instance, 88% of surveyed respondents reported that they would welcome back a neighbor who survived Ebola, with the highest proportion being in Karongi and the lowest in Burera , the heterogeneity in the answer is captured by a low p value in the chi-squared test. However, the proportion of respondents who would hug or touch a person who was previously infected with EVD was lower with the highest proportion to answer yes being found in Karongi and the lowest in Rusizi ; this discrepancy in answers is reflected by a low p value . --- Assessing the risk perceptions and beliefs about Ebola Table 4 shows the respondent risk perceptions and beliefs related to EVD. First, we report the DRC to be the country with the highest proportion of risk --- Assessment of Ebola related behaviours and practices The respondents were also asked to indicate the action they would take for themselves and family members to avoid Ebola. Table 5 presents the proportion of respondents with respect to various personal hygiene practices to avoid Ebola. The most common preventive measures, which were mentioned in each of the locations, were hand washing and limiting body contacts. We also found that practicing handwashing and good personal hygiene was a common behavior to avoid Ebola across districts, with little variation , while heterogeneously , people reported changing their behavior or practices to avoid being infected with Ebola, with the lowest proportion of "yes" answers found in Gasabo and the highest in Rubavu . When asked if people believe it is important to receive vaccination to prevent Ebola, 97.2% answered "yes" with non-significant variations across the districts . However, people from different districts showed variability in their answers regarding whether they believe vaccines against a Wrong response EVD are safe. Overall, 90.1% of participants answered "yes", while the lowest proportion of "yes" answers were found in Rusizi and the highest in Gasabo , the heterogeneity in the answers reflects a low p value for this question . --- General perception To assess the impact of the variables age, gender, district of residence, education level, and marital status in people answering the question "Do you think Rwanda is at risk of an EVD outbreak?", we performed a logistic regression analysis . We report significant effects on an answer for the question for the gender of the respondent and whether the respondent dwells in Rubavu or Rusizi. --- Qualitative findings --- Assessment of knowledge and awareness on Ebola Analysis of qualitative data suggests that there are variations on overall Ebola-related knowledge among the sub-population groups: Health Workers, Community Leaders, Adolescents and Teachers for both ECD and Secondary schools. We have learned that Ebola is caused by the virus which has its origin in animals: FGD, Community Leader Rubavu . I do not know the causes of Ebola, but I know its symptoms like vomiting, bleeding and strong fever, things like that. KII, Teacher Karongi . --- I wish I had knowledge on Ebola. I could explain to students, I do not have enough knowledge. I am not ignorant completely. I know little about it such as what we hear from radios on the nature of the epidemic. Teacher Karongi . Although Rwanda has not yet recorded an EVD outbreak, the study respondents were knowledgeable regarding the prevention of future outbreaks, and the knowledge was almost uniform across the districts, study population sub-groups, and geographical areas. Overall, the respondents reported receiving knowledge on Ebola mainly from their community leaders and community health workers. During community work called "umuganda" health workers are invited to share information on Ebola. However, since March 2020, there has not been any In Rusizi, there is a programme of vaccination against Ebola, so community leaders have information on Ebola. Also doctors from Kigali help people in the district on how to take measures to prevent Ebola. FGD, Community Leader, Rusizi . Several districts have been organizing EVD awareness campaigns to educate the public about how to better prevent the spread of the virus. For instance, community leaders in Rubavu indicated that they had received formal Ebola training from health workers at the district level as a part of the district awareness raising strategies. --- All Community leaders and opinion leaders have been trained and involved in Ebola education. The training was an alert in Rubavu district. We even did campaigns and the cars were driving around the town with loudspeakers with instructions on knowledge on Ebola and preventive measures. FGD, Community Leader, Rubavu . The analysis of the feedback from participants identifying as teachers across the study districts noted little to no formal Ebola-related dissemination of knowledge and/or training for teachers. For instance, teachers in Karongi reported limited exposure to educative activities on Ebola-related knowledge. One of the teachers had this to say: Honestly, I do not know much about Ebola. We simply just get some information from the radio, but this is not comprehensive for the role we play as teachers who should impart such knowledge to the children, who would in turn also share the information with their families at home. KII, Teacher, Karongi . --- Assessment of attitudes towards Ebola infected and affected persons Responses from the FGDs with parents and local leaders indicated that, in general, the community members have a positive attitude towards survivors. The following is an excerpt from one of the participants. Sometimes the person can be quarantined. However, when he or she is sent to the hospital and the hospital discharges him/her because he/she is cured, then there is no reason to fear or quarantine him/her. We can welcome him/her back. FGD, Parent/Caregiver, Karongi . Conversely, the results also show that some respondents have negative attitudes towards Ebola survivors, should they affect the country. These results are also supported by information gathered through FGDs. For instance, some participants will need more time to personally assess the risk of eating, hugging, or shaking hands with a person who has recovered from Ebola. For the recovered member, it takes time to go back into the community, the population would like to be sure that he has fully recovered, but there is no discrimination. FGD, Local Leaders, Rubavu, . Parents and local leaders also reported a willingness to adhere to Ebola burial guidelines and protocols when burying someone they suspect that he has died of Ebola . The following are excerpts from the FGDs conducted with the parents. For someone who died of Ebola, there is a burial team and because of fear of Ebola no one can approach the dead body, the community accept use of burial team, even if families are psychologically affected by not paying the last respects to the dead. FGD, Parents/Caregivers, Gasabo . --- In our customs we are used to make the last respect to the dead person, so it is too hard to accept the burial team, but it is a must, sometimes one member of the family can wear the protective equipment and be part of the burial team. FGD, Local leaders, Rubavu . One National level KII noted the need to conduct a refresher on EVD to raise awareness. My immediate suggestion would be to conduct a refresher related to Ebola especially when it comes to community sensitization approaches, latest information, latest science about Ebola, strengthen the knowledge of community health workers about the vaccine. Also because of turnover, this refresher will be new information for new health workers. We should also have a branded corner in health facilities where people can access Ebola information. We also need to engage teachers, ECD caretakers, young children to institutionalize knowledge about Ebola, plus we should constantly update to highlight the difference between COVID-19 and Ebola. We also need to have a refresher for religious leaders. KII, UNICEF, Gasabo . --- Assessing the risk perceptions and beliefs about Ebola Most of the respondents think that Rwanda was at risk of an Ebola outbreak. This is reflected in the FDG/KII findings with respondents citing both DRC and Uganda as the source of regional Ebola cases. "Yes, the epidemic can come in Rwanda because our neighbour DRC is regularly suffering from the epidemics and we are at a risk since they can happen to a person from DRC who can for instance enter and contaminate or spread the Ebola", KII, Teacher, Gasabo, . The fact that neighbouring countries such as DRC and Uganda had reported Ebola cases put the Rwandans at high risk. The following excerpts suggest the potential risk of Ebola as a result of illegal movement between boarders. Where we live, there are people who come from DRC to pick sand. We need to avoid contact with these people. They are picking sand at a place called Kiraro. FGD, Adolescent, Karongi . With new DRC, one method of transport is boat. Many people come from Goma and Bukavu and there it is possible for Rwandans to get Ebola. Also, here in Karongi, fishing is an important business. The lake is between two countries, but borders are not clear and there is constant movement between these people doing this business and other businesses which make the spread of Ebola possible. FGD, Community Leader, Karongi . --- "It has been said Ebola is coming from bats and in Rwanda we have those bats. So, it can come also. In addition, movement from Goma to Rwanda can facilitate the spread of the virus". FGD Adolescents, Gasabo . Most of the respondents from the FGDs and KIIs perceive that everyone in Rwanda takes Ebola as a very serious disease that spreads. However, isolated responses from FGDs with adolescents in Burera, and parents in Gasabo, Karongi, Rubavu and Burera mentioned that Ebola was a political disease, one that is 'made-up' by politicians. --- "Many people in the community know the seriousness of Ebola. Before COVID-19 pandemic, Health workers from several hospitals were sent into different churches to teach them how to take measures against Ebola virus". FGD, Community Leaders, Karongi . "Once I heard traders talking about a political disease", FGD, Adolescent, Burera . "People have different way of understanding. Some people have the misconceptions on Ebola. Some said that Ebola is a political disease, it does not exist". Parent FGD, Gasabo . The majority of myths and misconceptions were cited in KIIs and FGDs conducted in Gasabo, and the majority were captured in the KIIs with health workers. The persistence of myths and misconceptions also indicates a need for EVD awareness campaigns. "The myths and misconception about EVD, it is said that the epidemic is caused by a virus made from laboratories and brought in Africa". KII, Health Worker, Gasabo . "It is somehow said that it is an antichristian epidemic and aiming at imparting satanic stamp of 666". KII, Health Worker, Gasabo . --- Assessment of Ebola related behaviours and practices Respondents were asked to indicate whether they had changed their behaviour or practices to avoid being infected with Ebola. This change in behaviour was attributed to the community meetings/umuganda that were held before the COVID-19 pandemic. At these meetings, the members of the community were trained and equipped with various methods to mitigate contracting the disease. These training sessions were performed by health workers who gave speeches and handed out materials at these events. During the community work also, we were given the opportunity to talk to the people. When you suspect a person, you mention it on his transfer then you wait for the answer from the health centre to see if you were right. KII, Health worker, Karongi . In the event of onset of symptoms or confirmation of an Ebola case within a family, respondents across all the groups cited isolation as the recommended practice. Interestingly, parents in Gasabo went on to mention that isolation should be followed with testing of all family members to determine whether the disease had spread within the family. "In my opinion, in case families suspect Ebola, we have to inform our community health workers in my village, because I know, they have more information on health issues. Usually, we always seek help from community health workers because they are close to us and know how to behave in case of illness, we trust them, but in the meantime, we have to isolate the suspect". FGD, Parent/caregiver, Gasabo . "When we suspect one of member to be infected better to call the health workers while avoiding any contact and approach to that person". FGD, Parent/caregiver Rusizi . Some respondents indicated that they were previously vaccinated or willing to be vaccinated, highlighting the importance of EVD vaccines in preventing further infections and death. The vaccinated participants attested to experiencing no side-effects from taking the EVD vaccine. I saw many being injected especially those bordering Democratic Republic of Congo and I was wondering when we will get the same vaccines too, unfortunately we did not qualify for the first round, but many are ready for the EVD vaccines. FGD, Community Leader, Karongi . I have already been vaccinated and my family. If there are preventive measures, they are not 100 percent tight. The only way to be definite is to get vaccinated. FGD, Community Leader, Rubavu . My family has been vaccinated for Ebola. We did not experience any problem. We think the vaccine is safe because it is tested on those that received it. FGD, Community Leader, Rubavu . Most of the respondents were with the opinion that they saw no reason for communities to refuse vaccination. A few participants; however, cited myths and rumours as the cause for hesitancy but also believed this could be overcome through awareness campaigns and community mobilisation. --- Nothing in my community can prevent people to get vaccinated, sometimes rumours but with sensitization it was over. FGD, Community leader, Rusizi . The primary channel of communication that is most preferred by the respondents is radio, and in some cases, Television , but radio stands out across key informants and focus groups. "The Radio is the principal means since almost everyone has it. You can invite people in the community for a meeting and few will come but when it comes to the announcement made on the radio, everyone will attend. " KII, Teacher, Karongi . --- "Primarily they listen to the radio, they watch television and also listen to cars going around with big speakers. " KII, Health Worker Rubavu . Pertaining to trusted sources of information, the respondents collectively viewed the Ministry of Health and its personnel, Government Institutions, such as Rwanda Biomedical Centre, and local leaders as being more trustworthy in times of disease outbreaks. "Those who do meetings like Country leaders at the country level. We cannot accept the information from the district unless we hear ourselves from high rank officials. Unless it comes from the cabinet, otherwise we can't trust the rest. " KII, Teacher, Karongi . Regarding accurate dissemination of information, some participants indicated that there is need to strengthen the flow of educational information regarding Ebola. One of the key informants said the following. What is important is to reinforce the flow of information so that there is no cut in the communication to ensure that the right information is in the right place at the right moment. we monitor this very closely also to make sure that there are no rumours. KII, WHO, Gasabo . --- Discussion This study sought to assess the knowledge, attitudes and preventive practices concerning EVD in five selected districts at high risk of cross-border spread given their geographical proximity or with direct air links to DRC and Uganda. We observed relatively high Ebola-related knowledge and awareness levels across the five study districts albeit with variations by geographical proximity to affected regions of DRC and by population subgroups. Such a high level of awareness is a good indication of the awareness raising efforts that have been previously undertaken by the Ministry of Health. Consultations with health workers across the five districts indicated that they had received formal training on Ebola prior to the COVID-19 pandemic, which had covered topics on general knowledge surrounding Ebola, such as what the causative agent is, how it is spread, how it can be prevented, as well as the hallmark signs and symptoms. The training is likely to have contributed to a relatively high level of knowledge on Ebola. Apart from this training for health workers, most respondents across the five districts reported that they mainly received the information on Ebola through media sources, including radio and TV. Another proportion of respondents reported receiving information on Ebola from posters/notice boards at health facilities. People who came from the DRC were also a source of knowledge for the adolescents. The channel of communication that is most preferred among respondents from all districts is radio, which may be attributed to its increased accessibility . This finding, while not surprising, underscores the importance of radio in the context of health-related communication, especially in rural areas where other media sources are less accessible. To the best of our knowledge, this is the first comprehensive KAP survey on EVD in Rwanda. Our findings indicated a heightened level of Ebola risk perception in Rwanda as a country because it shares a border with DRC, which is the EVD epicentre in the East African region. Such a study is essential for developing preventive interventions for community members considering the increased risk of cross border EVD spread between both countries. Previous reports and studies have also showed that the risk of cross border EVD spread is exacerbated by the high volume of people traveling across the different borders between Rwanda, DRC and Uganda [25,28]. Despite the generally high level of Ebola-related knowledge, gaps remain in respondents' knowledge and attitudes towards EVD that were likely to impact preventive practices and behaviours. This was observed in survey results where a large proportion of people were apprehensive about whether they should physically and/ or socially interact with EVD survivors. For example, a small percentage indicated that they would not have a personal or close interaction with EVD survivors. Our findings align with another previous study that has shown how EVDrelated discrimination is largely based on community fear that EVD survivors are still contagious [29]. Such discrimination has led to EVD survivors being mocked by their communities, being evicted from their homes by their property owners, losing their former jobs, and being divorced by their spouses [29][30][31][32][33]. Most of our study participants indicated that one's re-acceptance into the community following an EVD infection depends largely on being able to provide a certification of treatment and recovery from an Ebola treatment centre. Similarly, previous findings demonstrated that EVD survivors were restricted from visiting public places, such as public toilets, and have experienced difficulty in trading commodities at their local market due to a community reluctance to physically handle their items or money [29,32]. These results demonstrate that comprehensive advocacy and awareness programs are still needed, especially pertaining to preventative measures and treatment options. Ebola virus disease was historically perceived as a near certain killer in the absence of treatment; however, that is no longer the case. Supportive care, rehydration with oral or intravenous fluids, and treatment of specific symptoms are now used to improve the likelihood of survival. In addition, the provision of supportive medical care for EVD infected patients, combined with approved monoclonal antibody therapy leads to recovery for most infected individuals [7,34]. Variability in behavioural practices was observed by district, where, for instance, residing in closer geographical proximity to affected areas resulted in residents being more alert in ensuring they protected their family members from contracting the virus. Our findings complement one prior study indicating that EVD-related fear can influence inhabitants to act with increased caution and to resort to preventive measures against the disease. These changes in human behaviour resulting from fear of the virus may lead to a decreased human to human, and human to pathogen contact rates [35]. The majority of respondents consulted during the study felt that increased personal protection could be achieved via vaccination and that the EVD vaccine was safe. The Government of Rwanda, in collaboration with the WHO and other partners, have been implementing public health measures to protect its population against EVD. Vaccination was one of the measures taken to protect health care and frontline workers in the districts at highest risk of EVD in Rwanda [6,36]. Raising awareness of the Ebola virus, educating people about appropriate prevention measures, and vaccinating the frontline is of utmost importance for successful EVD outbreak prevention and control efforts. --- Limitations and mitigation strategy The COVID-19 pandemic and the Government of Rwanda's COVID-19 restrictions necessitated altering data collection methods from face-to-face interviews to virtual interviews. Although contextually expedient, conducting interviews virtually created unforeseen challenges, where roughly 50% of selected respondents refused to provide informed consent. This is probably because participants may be apprehensive about sharing sensitive information or being recorded during the virtual interviews. Consequently, the research team was required to recruit more participants to replace those who had chosen not to participate in the study. Finally, conducting interviews online might also introduce specific methodological limitations related to information bias. For instance, the absence of face-toface interaction may lead to misinterpretations or misunderstandings of questions, potentially influencing the responses given by participants. To address this, the research team was requested to clearly communicate with study participants. --- Conclusion Overall, the study noted a relatively high knowledge level among the study respondents with 99.6% reporting having heard of Ebola. Such high level of awareness is a strong indication of the effectiveness of the awareness raising efforts that have been previously implemented by the Government of Rwanda. The results of this study also suggest that future communication and media efforts should be focused on knowledge dissemination in the public domain, such as available treatment options, case reporting, and the importance of positively interacting with EVD survivors. These findings, along with existing data and previous experience in communicating EVDrelated information throughout the country should help guide other high-risk countries in creating an effective, evidence-based framework for controlling EVD. --- --- Abbreviations --- --- --- --- --- Competing interests We declare no competing interests. ---

Background The overall goal of this survey was to understand the knowledge, attitudes, and practices related to the Ebola Virus Disease (EVD) in Rwanda. Methods This mixed-method cross-sectional survey was conducted in five selected districts of Rwanda. Quantitative data were collected from 1,010 participants using Kobo Collect Software and the analysis was performed using SPSS and Python software. Qualitative data were specifically collected from 98 participants through Key Informant Interviews (KIIs) and Focus Group Discussion (FGDs). Interview transcripts were imported into NVIVO 8 for coding and subsequent analysis.As per our quantitative findings, we report that from the 1,010 respondents, 99.6% reported having previously heard of Ebola, 97.2% believed that vaccination is important in combatting the disease and 93.3% of individuals reported a willingness to receive vaccination should one become available. Around 54% of the respondents were correct in identifying that the disease is of a viral origin which originates from wild animals (42.1%). When asked if they believed that Rwanda is at risk of an EVD outbreak, 90% of the respondents believe that the country is at risk of an EVD outbreak, and the cofactors *gender* and *whether people dwell in Rubavu/Rusizi* were found to significantly impact their perception of threat. As per our qualitative findings, the respondents mentioned that both geographical proximity and relations with the Democratic Republic of Congo place Rwanda at risk of developing an internal outbreak. Although the respondents seemed to be aware of the Ebola prevention behaviours, it was noted that some of them will require significant time before reintegrating into the community an EVD survivor, as they will first need assurance that the patient has fully recovered. Therefore, the qualitative findings reinforce what we originally reported in the quantitative approach to this study.Our results show that there was high EVD-related knowledge and awareness among the general population in Rwanda. However, for strong public health awareness, preparedness, and protection, a massive investment †

Introduction Stillbirth has a long-lasting impact on parents and families. This study examined socio-economic predictors associated with stillbirth in Nepal for the year 2001, 2006 and 2011. --- Introduction Stillbirth refers to the birth of a baby with no signs of life at or after 28 weeks' gestation [1]. Globally, stillbirth is a major public health problem, with more than 2.7 million stillbirths occurring annually; of these, 98% are from developing countries [2]. Sub-Sahara Africa and South Asia account for the highest numbers of stillbirth [3]. The long-lasting impact of stillbirth remains a large burden for parents, families, policy makers and public health practitioners [4]. Evidence has shown that stillbirth is associated with physical and psychological morbidity, and remains a significant source of cost for the affected family and community [2,5,6]. Despite the huge burden of stillbirth on families and global health, progress made in lowmiddle-income countries to reduce stillbirth is considerably slower than the decline in child mortality [3]. Stillbirth rates vary within and between countries; with economically disadvantaged communities having higher rates compared to their economically well-off counterparts [3,7]. In developing countries, the major risk factors for stillbirth include advanced maternal age, maternal educational status, infections, fetal development, environmental hazards, diabetes, malaria and umbilical cord complications [8][9][10][11]. Recent studies from developed countries have also reported that psychological issues are associated with higher stillbirth rates [12,13]. The major causes and predictors of stillbirth in South Asia are not well understood because of the huge variation in data availability and quality that underestimates the true number of stillbirths [5]. Recent case-control studies [14,15] conducted in Nepal found that stillbirth is associated with older maternal age, lower level of maternal education, coming from the poorest households, inadequate antenatal care and antepartum haemorrhage. Similarly, a verbal autopsy study conducted in Nepal revealed that obstetric complications which included prolonged labour, antepartum haemorrhage and pregnancy induced hypertension were associated with stillbirth [16]. A community-based study from a rural area of Nepal found that a history of prior child loss, maternal age above 30 years and low socio-economic status were associated with higher stillbirth rates [17]. The 2011 Lancet series on stillbirths suggested that for better estimation and intervention, the epidemiology of stillbirths should be at a country level instead of at the regional level because of the regional variations [4]. A major limitation of these Nepalese studies is that the findings cannot be used to inform initiatives and policy responses at the national level because the samples do not represent geographically diverse population across the country. Hence, the aim of this study was to provide nationally representative evidence on the socio-economic predictors associated with stillbirths in Nepal, using pooled data from the 2001, 2006 and 2011 Nepal Demographic and Health Surveys . Findings from this study would enable public health professionals to inform different policies and programmes to reduce stillbirth, with subsequent improvements in maternal and newborn outcomes in Nepal. --- Methods --- Data sources The NDHS is nationally representative, collected by the Nepalese Ministry of Health and Population, in collaboration with New ERA and ICF International, USA using a multi-stage cluster sampling design. Data on fertility, mortality, family planning, and important aspects of nutrition, health, and health services were collected for the years 2001, 2006 and 2011 using standard model questionnaires designed for, and widely used in developing countries [18][19][20]. For the 2001 NDHS, 8726 women aged 15-49 were interviewed, of these 7089 pregnancies were 7+ months' gestation. Similarly in the 2006 NDHS, 10,793 women aged 15-49 years were interviewed. Of these, 5921 pregnancies were 7+ months' gestation. In the 2011 NDHS, 12,674 women aged 15-49 years were interviewed; of which 5376 reported pregnancies 7+ months' gestation. A total sample of 18,386 pregnancies 7+ months' gestation five year prior each survey was included in the final analysis. For the year 2006 and 2011, pregnancies were identified using calendar information such as pregnancy outcomes and duration of pregnancy; whereas for the year 2001, pregnancies were identified using information such as pregnancy history index, and outcome and duration of pregnancy. Further detail of the survey methodology, sampling procedure, and questionnaires are reported elsewhere [18][19][20]. In all surveys, the response was more than 98%. --- Study outcome The outcome variable was stillbirth, defined as the birth of a baby with no signs of life at or after 28 weeks' gestation [1]. The outcome was recorded as a binary variable in the data set coded as 1 for 'stillbirth' and 0 for 'Alive at birth'. Information on stillbirth was obtained using reproductive calendar ; and pregnancy history and outcome of pregnancies . --- Exploratory variables The exploratory variables selected for this study were based on previous studies from developing countries [8,14,16,[21][22][23] and the information available in the pooled data sets. Mid-Western and Far-Western) and place of residence . The socio-economic level factors considered were maternal education, literacy level, occupation , paternal education, mother's current work status and household wealth index. The household wealth index measures the economic status of the household. As a measure of household wealth index, we pooled the wealth index factor scores in each of the three original Individual Recode data files as calculated by original DHS. The pooled original household wealth index factor scores were then categorised into three: the bottom, 40% of households was referred to as poor households, the next 40% as the middle households and the top 20% as rich households [25]. Maternal factors encompass maternal age at first birth, previous death of a baby, mother's current age and maternal marital status. We also considered environmental factors consisting of drinking water source and types of sanitation facility for each household classified based on the WHO and UNICEF Joint Monitoring Program guidelines [26]. Based on JMP guidelines, we categorized sources of drinking water as: piped water on premises , other improved drinking water sources , unimproved water sources and surface drinking water sources . Similarly, we categorized types of sanitation facility as improved and unimproved. Improved sanitation facilities included . Unimproved sanitation facilities were traditional pit toilet, pit latrine without slab or open pit and bucket toilet and open defecation . --- Statistical analysis Frequency tabulations were first conducted to describe the frequency and relative frequency of all potential confounding factors. This was followed by calculating the stillbirth rate and 95% confidence interval, using 'the number of stillbirths divided by the number of live births multiplied by 1,000'. Generalized linear latent and mixed models with the log link and binomial family [27] that adjusted for cluster and survey weights were used to identify those socio-economic predictors associated with stillbirth. A staged modelling technique [28] was adopted. Community-level factors were first entered into the baseline multivariable model with manual backward elimination process to keep statistically significant variables with p-value <0.05 . Second, socio-economic factors were added into community-level factors associated with outcome variable and those factors with p-values < 0.05 were retained after backward elimination process was conducted . Third, maternal factors were added into model 2. After applying similar approach as above, variables with p-values < 0.05 were retained in the next model . In the final stage, environmental factors were assessed with a list of significant variables from model 3. Variables with p-values < 0.05 were retained in the final model . Only those factors significantly associated with stillbirth at a 5% significance level in the final model were reported in the study. In the final model, collinearity was tested and reported. The analysis was restricted to five years preceding each of the survey. A total of 57 missing values were excluded from the multivariate analysis, and GLLAM estimates were translated to relative risk and 95% confidence interval. All analyses were performed using STATA statistical software, version 14.1 with 'Svy' commands to allow for adjustments for sampling weights and cluster sampling design. --- Results --- Basic characteristics of the study participants The majority of mothers who reported higher rates of stillbirth were from rural and mountainous areas, poor households, parents with low levels of education, households with unimproved sources of drinking water and unimproved toilet facilities . We also noted that mothers whose major occupation was agriculture had more stillbirths compared to those mothers who worked in non-agricultural sectors. The prevalence of stillbirth across three ecological zones indicates that the rate was 28 per 1000 amongst mothers who resided in the mountains whereas this rate was 17 per 1000 in the terai, and 19 per 1000 in the hills . --- Predictors of stillbirth The univariate analyses revealed that ecological zone ; religion ; mother's literacy ; parental level of education , currently not working mothers, mother's whose major occupation was agriculture, mother's age , types of drinking water source and types of sanitation facility were all significantly associated with higher stillbirth . Multivariable analysis revealed that factors associated with stillbirth were mothers in the age bracket , mothers who lived in mountains or hills, mothers whose religion was Hindu, Muslim, Christian and others, mothers who had no schooling or only primary level of education. Further we found that mothers whose major occupation was agriculture and those who used open defecation reported higher stillbirth. In the final model, we removed maternal education level and replaced it with father's education level; the result indicated that stillbirth increased significantly among fathers with no schooling . --- Discussion This study reports the predictors associated with stillbirths in Nepal by using pooling the three most recent Nepal demographic and Health survey and found that maternal age , low levels of education, sanitation and ecological zones were predictors for stillbirth. Additionally, when mother education was replaced by father education in the final model, father with no education reported significantly higher stillbirth. This current study provides an evidence-base that could be used to inform the design of effective interventions, policies and programmes aimed at health professionals and individuals recognising stillbirths. Primiparity is an established risk factor for stillbirth in both high and low income countries [29], and our results also found this association. This study demonstrated that mothers aged 25 years and above at the time of their first birth were more likely to experience stillbirth. This finding was supported by case-control studies conducted in Nepal, Bangladesh and Canada, which indicated that older mothers significantly reported higher stillbirth than younger mothers [14,22,30]. Similarly, a hospital-based study conducted in Nigeria also revealed that mothers aged 35 years or older were significantly more likely to report higher rate of stillbirths [31]. Studies conducted in high-income countries showed a significant relationship between stillbirth and maternal age [7,[32][33][34][35]. Higher stillbirth rate in older women has been attributed to the increase likelihood of congenital anomalies, chronic hypertension, placenta praevia, uterine rupture, and breech deliveries in older mothers which may contribute to an increased fetal death [36][37][38][39]. Studies have also shown that advanced maternal age has been associated with an increased risk of abnormal chromosomes, and or decreasing uterine and hormonal function [40,41]. Parental education is considered as one of the important determinants of health. Previous studies from Pakistan and Bangladesh reported that education could increase the uptake of health service utilization [42,43] with subsequent reduction in stillbirth. This study found that women with only primary level of education or no schooling had higher risk of stillbirth compared to those who had secondary or higher levels of education. There are very few studies from developing countries that have examined the relationship between maternal education and stillbirth. A study conducted in a province of Thailand revealed women who had low levels of education were at a higher risk of having stillbirths [44]. This finding is also consistent with previous studies from Canada and Denmark, which found that lower level of maternal education was associated with higher risk of stillbirths [45][46][47]. Plausible reasons for this finding may be that educated mothers are more likely to practice healthy behaviours, including health seeking, which may contribute to reducing their risk of stillbirth compared to mothers with no education. Likewise, fathers with no schooling were also associated with higher risk of stillbirth. Our study demonstrated that the risk of stillbirth was significantly higher among women who worked in an agricultural sector, similar to a finding from a hospital-based case-control study conducted in the Nguyen province of Vietnam [44]. Our finding of higher risk of stillbirth among mothers residing in the high altitude mountains or hills was similar to a retrospective births record obtained from four regional centres in Peru, which indicated that after controlling for potential confounding factors, mothers who lived in high altitude were significantly more likely to report higher stillbirths than those mothers that lived in low altitude [48]. Whether our finding is related to altitude or access to antenatal and birth service is unknown due to the limitations of the DHS data. However, literature has shown that management of pregnancy complications through quality antenatal care [49] and provision of skilled birth attendance around labour time [4] help to prevent stillbirth. Based on these evidences, it can be argued that the focused antenatal care as well as targeted skilled birth attendance for women residing in the mountainous region would help to reduce the number of stillbirth. Our study also found an association between stillbirth and mothers religious affiliation. Mothers whose religion was Hindu and others including Muslim and Christian reported significantly higher stillbirth compared to those mothers whose religion was Buddhist. Analysis [50] conducted in India using the National Family Health Survey reported differences in child mortality based on religious affiliation. In Nepal access and utilization of birthing services differs by religious affiliation and this may contribute to the increased stillbirth in some religious groups [51]. Unimproved water and sanitation contributes 0.9% to the global Disability Adjusted Live Years [52]. It is not surprising that women who reported open defecation were at greater risk of stillbirth compared to mothers who reported improved sanitation facilities; similar with the finding from a population-based prospective cohort study conducted in India that revealed open defecation among pregnant women was associated with adverse pregnancy outcomes [53]. The study has a number of strengths. Firstly, the analysis was based on nationally representative data ; thus, estimates from this study are generalizable to the Nepalese population and can inform national policies and initiatives in Nepal. Secondly, the response to the surveys was high , reducing a likely chance of selection bias from the observed findings. Thirdly, measurement bias is unlikely to affect the observed results as the data were collected using a standardised questionnaire developed for developing countries including Nepal [18][19][20]. It is however retrospective data, and there may be some bias in reporting stillbirth. Despite these advantages, this study is limited in a number of ways. Firstly, the diagnosis of stillbirth was based on self-report from mother and is subject to recall and misclassification bias. Secondly, formal verbal autopsies were not conducted on stillbirths. Finally, no information on health services factors or other factors such as tobacco, gestational diabetes and genetic abnormality that may have been associated with stillbirth were included in the NDHS data. --- Policy implications To close the equity gaps, community-based interventions need to be formulated and implemented in order to improve maternal and child health in Nepal. At the individual level intervention, uptake and quality of antenatal care should be encouraged among mothers from low socio-economic group and those mothers from hilly and mountainous ecological zones. At the community level intervention, increase awareness and access to basic and emergency obstetric care to women from hilly and mountainous ecological zones. These interventions will improve prevention strategies that could have massive and far-reaching improvement on Nepalese mothers and children in order for the country to accelerate progress towards achievement of ending preventable stillbirths by 2035 [54]. --- Conclusions Our findings suggest that antenatal care service should be targeted to women from low socioeconomic status and those who lived in the mountainous ecological zone in order for Nepal to further reduce the rate of stillbirth to a target of 12 stillbirths per 1000 births by the year 2030. --- The data was obtained from http://www.dhsprogram.com/data/ available-datasets.cfm. --- Ethics The DHS project obtained ethical approval from the Nepal Health Research Council-Kathmandu. The first author communicated with MEASURE DHS/ ICF International and permission was granted to download and use the data for his doctoral dissertation with the School of Science and Health at Western Sydney University, Australia. and MKN is supported by Early Career Researcher Postgraduate Scholarship, Sydney Medical School. The first author gratefully acknowledges the comments made on first draft by Dr. Amit Arora from Western Sydney University. ---

The Nepalese Demographic and Health Survey (NDHS) data for the period (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011) were pooled to estimate socio-economic predictors associated with stillbirths in Nepal using binomial logistic regression while taking clustering and sampling weights into account.A total of 18,386 pregnancies of at least 28 weeks gestation were identified. Of these pregnancies, 335 stillbirths were reported. Stillbirth increased significantly among women that lived in the hills ecological zones (aRR 1.38, 95% CI 1.02, 1.87) or in the mountains ecological zones (aRR 1.71, 95% CI 1.10, 2.66). Women with no schooling (aRR 1.72, 95% CI 1.10, 2.69), women with primary education (aRR 1.81, 95% CI 1.11, 2.97); open defecation (aRR 1.48, 95% CI 1.00, 2.18), and those whose major occupation was agriculture (aRR 1.80, 95% CI 1.16, 2.78) are more likely to report higher stillbirth.Low levels of education, ecological zones and open defecation were found to be strong predictors of stillbirth. Access to antenatal care services and skilled birth attendants for women in the mountainous and hilly ecological zones of Nepal is needed to further reduce stillbirth and improved services should also focus on women with low levels of education.

INTRODUCTION PR's 100-year anniversary4 passed quietly, amid the health crisis caused by the Covid-19 pandemic and potentially overwhelmed by both the visibility and impact of their work . PR's first century has brought variety of paradigms , theories and themes present , as well as in the increase of publications dedicated to the profession -whether national, regional, at times bi-or tri-lingual, international and interdisciplinary. This confirms that PR not only "needs constant revisiting, revising and reconsideration" but does so, with consideration to the past, present, and future. The profession too, grew tremendously and is expected to continue doing so, the size of PR market being projected to reach an estimated 133 billion US Dollars by 2027 more than 6% compared to its current revenues . The last estimates of the number of people working in PR are from 2006 . Such expected growth in the current context, shaken by crises of many kinds that are increasingly complex and wicked, is due to pose new challenges to communication practitioners. An online survey carried out by the CIPR among UK based professionals in 2021 during the height of Coronavirus restrictions found an overhelming number of respondentes to suggest that the pandemic has "increased the value and reputation of public relations". German practioners seem to agree as they observe an increasing importance of PR/Comms within their organisations: Comparing a set of indicators used to measure the internal value of the communication function over 14 years Seidenglanz & Fechner conclude that the crisis caused by the global pandemic has strenghtend PR/Comms claim for leadership. Additionally, developments in technology, work and society will also play a role in shaping the demands and expectations of and towards individual practitioners and the profession itself. As a result, research and practice have focused more recently on the future of PR/Comms as well as on the potential differences in expectations, perceptions and projections formulated by different age groups. Considering the long way the profession has come, the present study -embedded in the German market -pursues similar goals. --- THEORETICAL FRAMEWORK --- The Future of PR The concern for the future and development of PR/Comms has emerged in the 1990s with renewed and more intense interest emerging in the mid-2000s. It might have been initially linked with efforts to establish PR as an independent profession or as a natural development to the descriptive research documenting the rise of PR degrees . This has led on the one hand to the review to educational offers , and on the other hand to reflections and assessments of emerging trends. This latter focus aimed to provide suggestions and solutions for how to address emerging trends, business demands and university degree offerings, all aiming to secure the longevity and success of PR/Comms. Following her literature review and analysis of 580 departments listing PR courses, Neff suggested that "future program development in public relations curriculum will be most likely led by the communication departments [...] the concept of a 'management' orientation as being the 'ideal' focus for the public relations professional" not being satisfactorily established in the literature of the time. Later studies reviewing the perceived relevance of PR education and its content consistenly pointed out to the need to focus more on research, ethics, reflecting more on their increasingly strategic role and the continuous globalziation of the profession . What research also pointed out at the time, were contradictory views, especially among practitioners regarding PR education and its role in shaping the profession: "For example on the one hand employers perceive public relations education as over theoretical and lacking practical application; on the other, it is blamed for failing to instill the analytical flexibility found in non-PR graduates" . This has thus resulted in industry led initiatives aiming to establish benchmarks for the practice. It is in this context that Global Alliance commissioned the Global Body of Knowledge Project, aiming to identify the knowledge, skills and abilities of PR practitioners depending on their role and career level. An international project in its nature, what the GBOK also implied is that professionalization of PR remained a concern. From here, two different areas have emerged -both future related. One aims to ensure the professional status of PR is attained and maintained . The first includes capturing the reflections of practitioners themselves about their own field and practice while the second investigates the context in which PR exists and the trends influencing its application . Here too, some common areas of concern emerge: VUCA , super-wicked problems, trust in the profession and in organizations, polarization of society, and technological developments, more particularly digitalization, social media, and artificial intelligence. Each of these topics is then reconsidered within the context of communications practice. In 2023, the Edelman Barometer's main finding is that distrust breeds polarization. USC Annenberg Center for Public Relations Report shares similar findings: 77% of the respondents agreed with the statement "Polarization is bad for the country which is bad for our company" with 61% of the respondents attributing polarization to partisan media outlets. Communicators predict that organizations will increase their public engagement in dialogue around one or more social issues with 83% of them agreeing or somewhat agreeing that business has a powerful platform it can use to speak from on important issues. Thus, the report suggests that "professional communicators are the pioneers in this unfamiliar territory: 93% are spending more time navigating a growing list of complex societal topics" . Based on examples like Procter & Gamble's #likeagirl campaign in 2014 , Patagonia's documentary DamNation , and PayPal's and Deutsche Bank's decisions to cancel plans involving business operations in Charlotte, North Carolina, after the state passed an anti-LGBTQ law known as House Bill 2 in 2017 , one can expect businesses to become more vocal and visible in social issues. Both reports agree on the future directions of concern and potential focus for communicators and maintain the same points made by research into PR education two-three decades earlier : trust, ethics, research, measurement and evaluation, all on an increasing background of volatility and complexity and a higher expectation of responsible involvement from businesses. Adi`s PR2025 study took the questions about the future of PR/Comms further, inviting participants to reflect on how they perceived their role as communicators and then to identify the trends that they considered influential for the profession. In doing so, participants were then invited to focus on the solutions, in this case competences, attributes, responsibilities and frameworks to ensure appropriate preparation and response. In identifying that technological and digital changes are of concern for the near future , PR/Comms practitioners agreed that it is digital literacies and associated ethics that are of more value as learning focus in the future rather than specific platforms. Equally, guaranteeing both time and budgets for learning is what it was suggested it was needed the most to ensure success. --- Demographics and generational divides The multiple crises of our time are likely to create shared experience that could be defining for the current adolescents. The same was true for the old generation adolescents as "baby boomers" were shaped by the post-WWII prosperity, Generation Y by the invention of the internet and Generation Z by the ubiquity of the digital world. As a result of the same historical, political, technological, cultural and social conditions under which they grew up age cohorts developed similar characteristics in which they differ from other age cohorts. That exactly is the main idea behind the social generation concept . Empirically speaking the generation is a non-disputed variable identified and coded based on self-reported age groups and/or year of birth . The distinctions between digital natives and digital immigrants, those who grew up with internet and digital technology and those who did not, respectively, have been widely recognized . Several studies have suggested that there are similarities between different generations within the digital natives group. Nonetheless, Generation Z differs considerably from its predecessor, Generation X, more than it does from Generation Y. Members of Generation X tend to be more individualistic, sophisticated, and hedonistic than previous groups, and they seek engagement in tasks they find meaningful, interesting, and varied . These characteristics have implications for their perceptions of their profession and its organizational and societal role, as well as for their work expectations and desire for recognition. Consequently, Generation X represents a high-maintenance workforce requiring a leadership style that Klein has characterized as "helicopter management." While some researchers view these generational categorizations with skepticism and others even reject the concept , in PR research they are rarely questioned or critically considered. In fact, even when focusing on generational research and differences, the focus stems either from generational bashing and interactional frustrations with one particular group: the millennials . The present study aims to address the frustrations in the field of PR by critically examining a specific age-defined group: the likely next generation of PR leaders. Specifically, this study focuses on German-speaking PR and communications practitioners aged up to 35, who will potentially assume leadership roles and shape the profession in the upcoming years, hence the focus on the near future but also on future as in desirable and ideal. To investigate their professional attitudes, behavior, and opinions, the research question reads as follows: How do young PR/Comms professionals in Germany view their profession today and in the future? --- Research framework The study draws upon occupational area research, such as the latest editions of Seidenglanz and Fechner and Zerfass et al. , as well as the tradition established by PR2025 . The study explores four aspects of how the next generation of PR/Comms professionals act, think about, and reflect on their profession. The study presented here aims to explore and build upon the concerns surrounding the PR profession by critically approaching a specific age-defined group within the industry: the likely next leadership generation. The population under investigation consists of all Germanspeaking PR/Comms practitioners aged up to 35. To explore their professional attitudes, behavior, and opinions, the study draws upon occupational area research and the tradition established by the PR2025 study. In doing so, it formulates 4 research theses guiding the analysis, aimed to facilitate the understanding of the views and expectations of young, German-speaking professionals: • Research thesis #1 : Practitioners aged 35 and younger will consider VUCA and its post-pandemic version, BANI , as shaping their work and the future profession. Both VUCA and its post-pandemic version BANI were identified in previous studies as shaping and influencing the profession . • Research thesis #2 : Building and maintaining trust will be considered a major challenge by young PR/Comms professionals, who recognize the influence of VUCA and BANI on their work. • Research thesis #3 : Young professionals perceive their current occupational practices to be lacking in social focus and they envision for the future a higher focus for such issues. This is thesis links and explores argument about PR/Comms still not meeting all the requirements a professional status. • Research thesis #4 : Young PR/Comms experts will consider flexibility and a balance between work and private life as the most important factors when choosing an employer. Further details about their respective family status will be inquired to check the appropriateness of generational generalizations. --- METHODOLOGY The current study was part of a much wider research project study. A standardized questionnaire was used to explore the research questions and looking more closely at the appropriateness of research theses . A list-based online survey was thus implemented. Data collection took place between the 17th of May and the 10th of July 2022 and produced 160 completed answers. The sampling for this research was purposive. The databases used to recruit the participants were provided by the German PR/Comms Association , Quadriga University of Applied Sciences and Quadriga Media . All invited participants indicated previously that they would be interested and willing to support any research carried out by either organization. Having cleared the ethics disclaimers and demographic questions all participants indicating that they were 35 years of age and younger proceeded with questions about their employment expectations, challenges facing PR/Comms, the purpose and role of the profession . 111 of 160 respondents identified as female , 37 as male, and one as diverse . 5 Most of them -exactly two thirds -work for companies. With an average age of 31 years at least two of three live in a relationship , but only a few already raise children . Overall, there is no indication that the characteristics of the sample vary considerably from the population -that is young PR/Comms professionals aged no more than 35 years. Due to the nature of the sample, which is outlined above, no claim of representativeness can be made. Consequently, the means of inferential statistics cannot be applied. Instead, the benefit of this study lies in its focus on a younger demographic, which is likely to take over leadership roles in the profession in the upcoming years, and critique the variables so far taken for granted in previous studies including their assessment of desirable and ideal practices. Like with the PR2025 study, the projection of upcoming trends and portrayal of the ideal future is seen as a proxy to desirable behavior. The use of descriptive statistics to evaluate the data is thus the most appropriate in this case. --- RESULTS --- RT1: VUCA/BANI All participants of the study have been confronted with eight descriptive statements operationalizing if and to which extent phenomena associated with VUCA and BANI are visible specifically within their own PR/Comms practice . The task given was to indicate their level of agreement, considering the current status quo on the one hand and the future situation on the other. The mean scores for agreement are higher for the VUCA dimensions. This means that despite the impact the pandemic and global health crisis BANI items have received less agreement. Stakeholder relationships are still considered stable and not brittle , causes and interdependencies in public communication are generally comprehended and the impact of communication measures can still be predicted albeit a few uncertainties . Only a considerable degree of anxiousness among stakeholders -operationalized by the BANIitem the complexity of public communication has made stakeholders increasingly uneasy -is identified as a more serious issue . Participants consider the accelerating change of public communication -that means volatility of societal developments -as most pertinent trend and thus potential threat. Other development typically associated with VUCA -that is more complexity , uncertainty and ambiguity -are also considered relevant. Moreover, nearly identical mean values for uncertainty, ambiguity, and anxious could be related to the increasing complexity of PR/Comms work. Source: own elaboration. Finally, mean differences between the status quo and the expected situation after five years are not major. All items record a higher agreement for the forward-looking statements indicating that participants expect all aspects of VUCA and BANI to be exacerbated in the future. Both the average VUCA and BANI scores confirm VUCA/BANI as current and future phenomena and thus relevant for the PR/Comms profession. However, the next generation of practitioners does not predict a change from VUCA to BANI as the average VUCA mean score remains well above its BANI counterpart . --- Specific challenges Participants were asked to rank the most crucial challenges in the PR/Comms field. The results showed that most practitioners considered establishing and maintaining trust as the top challenge, with an average rank of 2.8 . This was followed by communicating the environmental responsibility of organizations in social, economic, and environmental contexts and creating transparency with regard to organizational actions and explaining the respective contexts . The challenge related to fast and flexible adaptation of stakeholder communication to constantly changing circumstances ranked fourth . Source: own elaboration. --- RT3: Objective and purpose of PR/Comms Participants were asked to assess the current status quo and their -possibly idealisticexpectations for the objectives and purpose of PR/Comms . The results revealed that building public trust was seen as the second most sought-after task of the profession, with a mean score of 4.4 out of 5.0. The predominant objective identified was providing public information, which scored slightly higher . Notably, young practitioners believed that the PR/Comms industry is currently doing too little to bring society together, as reflected in their desire to promote social cohesion. The mean values for public information , building public trust , establishing transparency , and promoting social cohesion increased between 0.2 and 0.9 when considering the ideal state in the future. These findings indicate that young German practitioners view PR/Comms as having a crucial role in creating a more cohesive society. While the relevance of organizational objectives might decrease in the future, they still hold value: building organizational reputation and trust scores 4.5 on first question , but only 4.3 on second . Same goes for avoiding negative organizational publicity and persuasion in the organizational interest . Other items such as balancing the organisation's and external stakeholder's interest and creating dialogue opportunities for external stakeholders on the other hand gain significance for the next Generation. between the desire for intrinsic values and the extrinsic motivation of a well-paid job . The next generation of PR professionals demonstrated both a progressive and a hedonistic approach when selecting an employer . They valued employer-provided training for skill development , flexible working hours , career opportunities , and sufficient free time . The compatibility of family and career was more important for practitioners with children compared to those without children . Table 4. Relevance of criteria for choosing an employer. Question: What criteria were decisive for you personally when choosing your current employer? Assessments measured on a scale from for "not relevant at all" to "crucially relevant". In summary, young practitioners in PR/Comms prioritize intrinsic values and seek a pleasant work environment. They also value a balance between work and personal life. While organizational objectives still hold relevance, the next generation intends to bring a more societal focus to the profession. --- DISCUSSION The current study confirms that VUCA dimensions explored in studies like PR2025 remain of concern even for younger PR practitioners . More specifically, the comparison with a study conducted four years earlier by Bentele, Seidenglanz & Fechner shows some interesting developments although it has to be noted that the latter sample represents PR/Comms professionals of all ages. Taking this potential bias into account all VUCA items are deemed more relevant today as they have been back then. This confirms the pertinence of the associated phenomena. Moreover, volatility became by far the most relevant aspect of the theory indicating a steady acceleration of communication. Finally, this highlights the need for practitioners, their departments, and organizations to pay more attention to well-being to cope and address potentially higher stress levels. A change in mindset is also noted among the sample of this study. While trust, and transparency are top-of-mind topics that younger practitioners share with their more senior colleagues , it is their focus on and concern with sustainability that sets them apart, which corresponds with the findings of various non-PR/Comms related studies investigating generations Y and particulary Z such as Beckers & et al. or FIRST INSIGHT . Clearly, progressive attitudes affect how young practitioners see the purpose and mission of the PR profession as well. Nevertheless, the intent is not to reinvent, but to change the focus of the profession in the direction of more societal contribution is obvious. In fact, the reported status quo for PR/Comms objectives very much aligns with a previous study which evaluated the whole profession regardless of age . Young German practitioners do not dismiss PR's organized persuasive work, but strongly believe that social cohesion and transparency should have more weight within the professions work, which justifies making stakeholders a priority. This also echoes both Bleile's projections of PR/Comms becoming "a tool at the public's disposal, with the best communicators working for the public to provide value and legitimacy" and Adi and Stoeckle's call for a more accountable and responsible PR practice. The potential change young communicators might bring to the profession has to be viewed at in the context of another observation. Typically for Generations Y and especially Z the personal "pursuit of happiness" becomes a central issue when deciding for a job and choosing an employer. Although idealistic values certainly do matter the hedonistic attitude makes them look for a pleasant work occupation, where they can do their own thing, are getting paid well and still enjoy a beneficial balance between work and private life -a contradictory situation, which Plastino generally observes within that particular age group. Interestingly, life situations also seem to also play a role here, as priorities seem to change once PR practitioners become parents. Hence it is not impossible that other factors as well may result in a change of attitudes. Overall, although this study is able to identify a few clear-cut trends there is still a lot of riding on the future of PR/Comms. --- CONCLUSIONS This study offers insights into the future developments and trends of the PR/Comms profession in Germany, as perceived by 160 young professionals who represent the potential next generation of PR/Comms leaders. The findings show notable difeerences between the current state of the profession and its desired state, particularly in relation the purpose and organization of the profession. The participants have demonstrated an idealistic mindset, driven by the pursuit of personal fulfillment, work-life balance, and financial stability. However, family circumstances can potentially influence this approach. The study also confirms the relevance of VUCA characteristics in the profession. While participants acknowledge that the field has not fully entered the BANI territory, characterized by complex and unpredictable public communication, they anticipate increased challenges in the future. --- Limitations The nature of the sample and sampling method obviously limit the degree to which the presented results can be generalized. Statistical representativeness thus cannot be claimed, which is a limitation faced by much of the current PR/Comms-related surveys -research fatigue, increased workloads, and difficulty maintaining respondent databases that are GDPR compliant being contributing factors. Despite this, the study offers a unique opportunity for research on a specific subgroup within the PR/Comms profession. Moreover, the results do echo the findings from other studies indicating to a developing mood and a set of common concerns shared by young communicators. Looking in the crystal ball is always problematic. This is especially true for surveys with forward looking questions. It is just not within the capacity of the human brain to predict future behavior precisely and reliably. Complicating the situation some of the questions trigger a social desirability bias. So, what is measured are intentions based on limited sample, which only allow conclusions for likely tendencies. Moreover, the hedonistic attitude which became apparent while exploring the deciding factors when choosing an employer cannot be underestimated and may lead to a more pragmatic and less progressive approach to PR/Comms among some practitioners. Nevertheless, since advocacy for more reflective, transparent, accountable PR/Comms practice is shared widely, the assumption that idealistic values among the next leadership generation PR/Comms will lead to a change of the profession within the foreseeable future seems realistic. But the degree of change probably will be less obvious as today's data suggest. --- Further research Having reflected the limitations future research should continue to challenge and explore these generational aspects while also integrating other variables -first language, country of birth vs. country of residence, personal values preferences. Additionally, more regional, and national data, including more European data is still highly needed. Methodologically, it would make sense to explore the attitudes, behavior and opinions of the next PR/Comms generation more deeply using qualitative techniques such as single in-depth-interviews, focus groups or Delphi designs. --- Source: own elaboration. --- RT4: Decisive factors choosing an employer To understand the factors influencing the choice of an employer, participants were asked about the criteria that were decisive for their current employer. The results indicated that intrinsic values were prioritized, with a high importance placed on a pleasant work atmosphere and assuming responsibility and decision-making freedom . The compatibility of the organization's purpose with personal values was also deemed important . However, there was a conflict

This study investigates the perceptions of young PR/Comms professionals in Germany regarding the profession's current and future state, amidst the significant impact of the COVID-19 pandemic on the job market. The research builds on previous studies exploring the future of work and the PR industry, aiming to address gaps in literature regarding generational differences in attitudes and a critical reflective approach. The study used a standardized survey and purposeful and snowball sampling methods to select 160 PR practitioners under the age of 36. The results reveal that the next generation of PR/Communications professionals values intrinsic and idealistic factors when choosing an employer, but those without children pursue a more hedonistic agenda. Respondents perceive themselves as dealing with the ramifications of a VUCA (volatile, uncertain, complex, ambiguous) world and consider trust as the most significant challenge within the next five years. The study highlights a significant gap between the societal and internal dimensions of PR/Communications and its actual implementation. The implications of the study suggest a need to address contentious questions surrounding the role and mission of PR/Comms in society, focusing on social impact and social value, and advocating for more collaborative and interdisciplinary approaches in training and learning practices.

INTRODUCTION During the last 15 years, adolescent birth rates have decreased dramatically in the United States . Despite declines, since the mid-1990s Latina teens have maintained higher birth rates than Black and White teens . In 2009, although California's Latina teen birth rate was below the national average for Latina teens , it was nearly two times higher than the average teen birth rate in the state . That same year, 72 percent of all teen births in California were to Latina adolescents . Numerous factors have been identified as influences on adolescent childbearing , including contraceptive access and method choice ; sexual partnership characteristics ; social and cultural norms regarding teen childbearing; and parent-child communication and parental monitoring . Socioeconomic disparities and poverty at the individual, household, and neighborhood levels consistently have been found to be associated with teen pregnancy and sexual decision-making . The weathering hypothesis, which attributes adverse health outcomes to cumulative exposure to socioeconomic and political inequities, offers one theoretical explanation that may underlie these associations ; adolescents' future discount rates affected by perceived life chances has been hypothesized an alternative explanation of health behaviors . Likewise, acculturation, the process whereby immigrants and children of immigrants adapt to multiple cultural and social norms in their home and community environments, may be a particularly influential factor for Latina youth. Such findings suggest the potential importance of addressing structural factors that shape the environment in which youth make decisions about their futures, including decisions around childbearing. Socioeconomic opportunities, measured as educational and career aspirations and expectations as well as by poverty indicators, have been shown to affect the reproductive health of U.S. adolescents . Adolescent girls with low educational expectations are more likely to become pregnant and to give birth than girls with high expectations . Youth with aspirations for the future are more likely to abstain from or delay intercourse and to use contraception if sexually active . Parents' educational expectations for their daughters are associated with adolescent sexual behavior and pregnancy occurrence, with national data documenting that low parental expectations are tied to higher likelihood of teen pregnancy and childbearing . One qualitative study conducted with Latino parents concluded that although parents preferred that their children defer childbearing to permit greater educational opportunities, the value placed on the role of motherhood and the attention given to teen mothers sent conflicting messages to their daughters . Other research suggests that Latina teens may view childbearing as motivation to achieve educational and occupational success or as a way to develop in a social environment that offers few alternative opportunities for the transition to adulthood . However, few studies have focused on how socioeconomic opportunities influence Latino teen childbearing, a relationship that may be uniquely shaped by immigration and acculturation, family environment and cultural norms . In response to the persistently high rates of teen births among Latinas and a need for improved understanding of socioeconomic and racial/ethnic disparities in teen childbearing, we conducted a qualitative study examining contextual influences on pregnancy and childbearing among pregnant foreign-and U.S.-born Latina adolescents in two California counties. We included a comparison group of young Latina adult women in the same communities who had delayed childbearing until adulthood. Our primary objective was to explore how individual and family educational and employment aspirations and expectations affected teen childbearing among young Latina women. Further, we assessed how contextual factors -family, immigration and subsequent acculturation shaped these relationships. --- METHODS Sixty-five pregnant foreign-and U.S.-born Latina women ages 15 to 19 or 22 to 35 were recruited at eleven community-based clinics or health education programs in Los Angeles and Fresno, two California counties with high teen birth rates. All participants were pregnant with the first child they planned to carry to term and were in their second or third trimester of pregnancy. All research activities were approved by the California Health and Human Services Agency Committee for the Protection of Human Subjects and by the Committee for Human Research at the University of California, San Francisco. A detailed description of the study's methodology and analytic approach has been published elsewhere and entered into QSR NVivo software. The qualitative analytic work was organized first by structural codes corresponding to study questions; subsequently a list of emerging thematic codes was identified and refined through iterative reviews of study transcripts, structural coded sections, and memos produced by the research team. For this paper, the research team analyzed text corresponding to two thematic codes: 1) educational and employment aspirations and expectations, including sub-themes related to school commitment and participation, childbearing and socioeconomic aspirations and expectations, and barriers to education and 2) influences on aspirations and expectations. --- RESULTS --- Study population characteristics The 65 study participants were evenly distributed between Fresno and Los Angeles, and between U.S.-and foreign-born women . --- Expectations for education and employment: influences on teen childbearing Nearly half of teen respondents experienced a common trajectory during their early teenage years that involved a gradual disengagement from and disinterest in school. Prior to their pregnancy, a handful of teens had already left high school to complete educational requirements through home study or continuation school. For U.S.-born teens and youth who immigrated to the U.S. prior to adolescence, this trajectory was influenced by environments largely void of economic opportunities. A number of teens described pregnancy as a motivation to focus on education and career and as presenting them with an opportunity for maturation. One teen described her pregnancy as a positive influence on her development: "I feel that I'll actually have a life now. Cause before, it was just fun and games. Now I'll have a responsibility." Adult respondents, in contrast, were more likely to describe high school as an enjoyable place where their ability and intelligence were recognized and as a critical step in achieving future educational goals. Furthermore, they tended to describe teen pregnancy as an event that would have interfered with their educational and career development: "I had goals way before I was even thinking about kids….I would think if I had a kid right now I wouldn't be able to go to school. I won't be able to become a doctor that I wanted to be." Some adults recognized that, as teens, childbearing constituted an important goal; however, they noted that their own education was critical to their success as a mother. As one respondent stated: "My main goal was the family, but I thought if I didn't educate myself first, how was I going to educate my children?" Future aspirations of several recent immigrant teens were shaped by traditional gender norms that aligned their education and childbearing trajectories with those of teens in their communities of origin. These respondents articulated low employment expectations resulting in reduced emphasis on schooling. This was sometimes directed by their male partners who supported them in completing high school but had stated clearly that they could not work and develop professionally. Furthermore, for some recent immigrant youth, early childbearing was regarded as typical in their home country and the later teen years (ages [17][18][19] were considered an appropriate life stage for pregnancy; educational goals, as a consequence, were sidelined. --- Family expectations for education and employment: influences on teen childbearing Most respondents stated that education was important to their families; however, expectations for the level of educational attainment and the support provided by parents varied between teen and adult respondents. Whereas about half of teens noted that their parents' expected them to complete high school, adults commonly reported that high school graduation was considered by their parents to be a given rather than a goal and most families supported education beyond high school. Adult respondents generally reported they had more familial support available to them during their teenage years then did teen respondents and related specific examples of how their parents provided emotional and tangible support in school. Emotional support included encouraging school attendance, vocalizing the desire to see their child succeed, and warning teenage daughters about the challenges of finding employment without a good education. Tangible support included assistance with homework, meeting with teachers and school staff, and helping with school paperwork and applications. One U.S.-born adult commented how her father contributed quite substantively to her learning: "I remember my father always showing interest in my schoolwork. He always wanted to read my writing assignments….He looked so forward to when it was time to do projects because it got his creative mind going…he'd sit down with me and we'd come up with incredible ideas." Though some teens remarked that their parents advocated education: "[My mom and dad] always push me to go to school and tell me that if I want to have a good life that I will have to go to school," others, all U.S.-born, commented explicitly that their parents expressed little interest in their life or future. As stated by one U.S.-born teen: "My dad was abusive…he couldn't take care of us 'cause he had like three or four other jobs. He was never there for us. I was basically taking care of my brothers and sisters. He would be out in the streets until one, 2:00 in the morning." Foreign-born adult and teen respondents, particularly those who had immigrated to the U.S. prior to adolescence, articulated a strong influence from parents to attain a college education and to take advantage of opportunities presented to them by their having immigrated to the U.S. Among teens, for example, nearly two-thirds of foreign-born compared with one-third of U.S.-born respondents indicated expectations from their parents to obtain higher education. A foreign-born adult reflected on how her parents encouraged education as a way to achieve a more secure future: "when I was a teenager, my parents took me to the fields to see how you earn money…to work in the fields. I didn't like it, and that kind of encouraged me more to go to school and have a better future." However, having a parent place tremendous value on education was insufficient, alone, in ensuring teens achieved their educational goals and delayed childbearing. Immigration-and acculturation-related barriers presented obstacles for some to actualize these goals, which ultimately affected childbearing. --- Immigration influences on education: effects on childbearing Legal and linguistic barriers were cited by many pregnant foreign-born teens as inhibiting educational attainment and perceived professional opportunities. For teens who immigrated during adolescence, immigration-related factors constituted a primary influence on shifts in educational expectations and attitudes toward school. Motherhood, therefore, constituted a viable alternative. Immigration laws that limit access to financial aid and higher education also presented constraints. As one foreign-born teen who had lived in the U.S. since childhood stated, it is "really hard…for Latinos…especially people that don't have papers, you know. You do want to finish high school, but going to university or college…it's difficult cause you don' got papers, you [don't] got money to pay the college." Language barriers constituted a second challenge for participants who immigrated to the U.S. during their teenage years. An 18year-old participant who moved to the U.S. at age 15 highlighted the challenge presented by linguistic barriers: "I didn't want to [go to school] anymore. I was embarrassed because I'm an adult already and I didn't speak any English." For some youth who immigrated as teens, immigration itself disrupted their schooling. These youth described difficulties in acculturating to a new social environment, coupled with lack of English comprehension and language skills in the classroom, ultimately limiting their ability to sustain previously high educational performance. Finally, reasons for immigration were noted by several teen participants as contributing to low educational expectations upon arrival in the U.S. and likely affected early childbearing. --- Family influences on education: effects on childbearing Economic instability within the family, household disruption due to parental separation or divorce and family care taking responsibilities presented great stressors on teens; foreignborn teens in particular. These barriers were prompted frequently by immigration itself or by movement between relative's households after arriving in the U.S. As one foreign-born teen commented after her move, "it was just my mom and myself. And when we got there she couldn't find a job, and we didn't have anyone else there who could help us out, and I tried to find a way to help her too in that way. So I looked for temporary jobs as a way to earn money." Family reunification of adolescent children and their parents, highlighted by one teen who had been raised by her grandmother in Mexico and then moved to the U.S. to live with her mother at age 13, highlighted the potential complexity of reconciling a parent-child relationship alongside adapting to a new country of residence. Other immigrant teens had to assume substantial childcare responsibilities for extended family members, which interfered with their ability to stay in school. As an immigrant teen whose cousin had two children reported, "[My cousin] would watch one of the kids and I would watch the other one." This directly inhibited her ability to attend school regularly. In contrast, family instability due to illness, neglect, substance use and incarceration presented substantial challenges for U.S.-born teens in attaining their educational goals. These issues rarely surfaced as major educational barriers for adults. One U.S.-born teen commented on why her mother's illness prompted her to drop out of high school: "I guess because my mom needed somebody to be with her at the house and she wouldn't have nobody. And I would be like the only one that will help her around like to cook and stuff." Another statement underscores the level of familial disruption faced by some U.S.-born teen respondents: "Well, my dad…he used drugs a lot. He's in prison. He's in there for life. My mom…we never got fed and she always had different guys over. And just, she wasn't really a good mother." --- DISCUSSION Pregnant Latina teens in this study described few tangible socioeconomic opportunities. Thus, for many teens, pregnancy, even when unintended, offered an opportunity for a meaningful transition from adolescence to adulthood. Despite general family support for education, teens faced substantial socioeconomic and family barriers to achieving their educational goals. For foreign-born teens, legal and linguistic barriers predominated, with immigration itself prompting disruption in household and socioeconomic stability. For U.S.born teens, lack of emotional and/or tangible support from family as well as instability within the household presented substantial challenges to educational attainment. Many of these factors reflect underlying household and neighborhood poverty that shaped the more proximal determinants of early childbearing, including pregnancy intentions and ambivalence and patterns of contraceptive use. Many teens felt that while their family members generally supported their educational and career goals, they lacked the knowledge and skills to help them realize their goals. In contrast, relationships with engaged and communicative parents during the teen years helped many adult Latinas defer childbearing until adulthood. Previous literature underscores the importance of engaging parents in their child's education throughout adolescence , yet this study highlighted distinct barriers to achieving this goal. For U.S.-born teens in particular, such an approach would require consideration of family structure and neighborhood instability due to economic insecurity, substance use and community violence and the effects these environmental factors have on childbearing perceptions and desires within the life course. For foreign-born teens such efforts would need to address linguistic and legal barriers to education. We found variability in the factors affecting educational attainment and childbearing outcomes within the foreign-born population represented in this study. Different acculturation trajectories appeared to account for some of this variation.. Participants who had immigrated as young children tended to report receiving strong family support for their education, often because their families' migration to the U.S. was motivated by their parents' desire to improve their children's socioeconomic opportunities. In contrast, gender norms expressed by some more recent immigrants favored childbearing over education and career, particularly among teens and young women who were already married. Other recent immigrants articulated strong values for education, but noted financial and linguistic barriers that inhibited their ability to attend high school and/or continue career training they might have begun in Mexico. These findings highlight the importance of tailoring pregnancy prevention programs, including those that address socioeconomic opportunities, to the diverse range of norms and experiences reflected in the population of foreign-born adolescents. U.S. immigration policy also has an important effect on the educational experiences of Latinos residing in the U.S. The federal DREAM Act, which would permit states to provide in-state college tuition for students without regard to immigration status, grant conditional permanent residency and permit later naturalization, addresses a structural barrier to educational opportunities faced by immigrants without permanent resident status . Several study limitations should be noted. First, adult respondents were asked to report retrospectively on their educational experiences and parental attitudes/resources during their teenage years. Thus, we focused comparisons on issues that would be less subject to recall bias . Second, to protect participant confidentiality, we collected little information on the circumstances of migration and the reasons for immigrating to the United States, which likely influenced individual acculturation as well as family structure and its effects on setting educational and career expectations. Third, study participants were volunteers from a clinic population predominantly, and we were not able to assess the representativeness of study participants compared to all pregnant women seeking prenatal services at the study sites. Therefore, results of this qualitative study should be considered descriptive and exploratory. This study offers an in-depth examination of contextual factors that affected perceived socioeconomic opportunities and childbearing among Latina teens in two California counties. The study's results support a need to provide tangible future opportunities for Latina youth by promoting youth development strategies, strengthening educational resources for youth and their families, and encouraging youths' educational and career development as a strategy for reducing teen childbearing. Though immigration presented distinct structural and cultural barriers to education that appeared to influence teen childbearing, the fundamental lack of hope for educational achievement and professional development was profound. Comprehensive pregnancy prevention efforts that address youth's socioeconomic and family environments , including factors affecting immigrant adaptation and acculturation, constitute a critical area for intervention. b Parental educational attainment when the participant was 13-14 years old.Does not sum to 100% owing to participants who reported not knowing their parent' educational attainment.

Background-The decrease in adolescent birth rates in the United States has been slower among Latinas than among other ethnic/racial groups. Limited research has explored how socioeconomic opportunities influence childbearing among Latina adolescents. Methods-We conducted in-depth interviews with 65 pregnant foreign-and U.S.-born Latina women (31 adolescents; 34 adults) in two California counties. We assessed perceived socioeconomic opportunities and examined how family, immigration and acculturation affected the relationships between socioeconomic opportunities and adolescent childbearing. Results-Compared with women who delayed childbearing into adulthood, pregnant adolescents described having few resources for educational and career development and experiencing numerous socioeconomic and social barriers to achieving their goals. Socioeconomic instability and policies limiting access to education influenced childbearing for immigrant adolescents. In contrast, family disintegration tied to poverty figured prominently in U.S.-born adolescents' childbearing. Conclusion-Limited socioeconomic opportunities may play a large role in persistently high pregnancy rates among Latina adolescents.

Introduction More than 10 million prisoners are available worldwide, and approximately 30 million prisoners are released annually . More than 7 million prisoners are living in low-income and middle-income countries , comprising about 70% of the world's total prison population . Many of whom are imprisoned due to drug-related crimes. About 225,000 Iranian prisoners are above 18 years old, with 3.91% of them being female , while under-18 years-old prisoners are kept in the juvenile correctional and rehabilitation center . The number of prisoners differs from one country to another, and it is difficult to compare the number of criminals in different societies because police and judicial institutions in every country have different rules for imprisoning criminals , and their rules may change with the passage of time . The number of male prisoners is much higher than that of females all over the world . Moreover, the pathways to prison differ between male and female , and socio-psychological problems, HIV prevalence, and related risk behaviors, such as drug injection among male prisoners is higher than that in female prisoners. Therefore, it is important to study male prisoners . Prisoners are vulnerable to multiple problems, with many restrictions and deprivations in prison adding to these problems . They suffer from poor relationships and social rejection because of the norms and subcultures in prison and separation from the family. Such problems decrease their personal and social adjustment, as well as their mental health and safety, but some prisoners are efficient and young and must return to social environment after their release . Prison is a potentially important stressor in either the formation or exacerbation of prisoners' psychological problems, such as low level of psychological wellbeing . Psychological wellbeing is among the most important determinant of global wellbeing and health of prisoners. The PWB is defined as an attempt to realize one's potential capabilities and includes six characteristics of "positive relationships with others, " "self-acceptance, " "personal growth, " "purpose in life, " "environmental mastery, " and "autonomy, " which indicate one's mental balance and health . Wellbeing generally predicts a variety of vital outcomes, including a sense of life satisfaction, and greater success in job and relationships . Studies have shown prisoners' low level of PWB, so its assessment in prisoners is important because they are vulnerable . Poor family cohesion contributes to one's diversion and criminal behavior, delinquency and consequently imprisonment. Upon entering prison, this problem has destructive effects on family relationships . Family cohesion , a determining factor for PWB , is defined as the emotional bond between family members. FC is associated with adolescents' internal and external problems and predicts their future behaviors. External psychological symptoms, such as academic failure, isolation, violence, and hyperactivity, and internal symptoms, such as anxiety and depression, may also transmit to adulthood . Imprisonment seriously disrupts planning and purpose in life . PL is a positive structure that can better predict and promote well-being . PL is an important religious, philosophical and psychological topic and has various definitions in different fields; however, it is usually defined as one's sense of function and meaning in life. "Purpose in life" and "meaning in life" are often used interchangeably . A review of the literature showed that PL was an important and positive personality trait and predicted health outcomes as well as a higher level of PWB in present and future . Goodman et al. showed a significant positive relationship between family attachment and PL in men and women. Alhussain et al. reported a significant relationship between PWB and FC in young people . Another study on students showed that the higher the FC, the higher the PWB . Problems of prisoners are usually ignored and neglected by community health nurses and social workers, who can render professional services to prisoners and their families by assessing the PWB, FC, PL, and their inter-relationships. They also can design social work interventions and educational programs to increase the PWB, FC, and PL. Given the cultural and contextual nature of these factors, studies are required in different communities to design comprehensive educational interventions. No study has simultaneously examined the three variables of PL, FC, and PWB in prisoners. Most of the studies mentioned above were conducted on non-prisoners and those who had clear differences with prisoners, so the current study aimed to evaluate PWB, FC, PL, and their correlations in male prisoners in Iran. --- Materials and methods --- Design and settings This cross-sectional study was conducted on male prisoners in a large prison in southeastern Iran. --- Sampling The target population of this study included all prisoners aged above 18 years old, who were available at the time of the study . According to Cochran's formula, the sample size was 252 people . With about a 5% probability of dropout, 265 male prisoners were selected using simple random sampling. Inclusion criteria included prisoners, who could read and write to complete and understand the concepts of the questionnaires, were imprisoned for at least 6 months, had at least one member family, and declared no acute mental and physical illnesses at the time of data collection. The first researcher selected samples using simple random, so he prepared the list of prisoners and recruited the samples using random number table. --- Data collection and instruments Data were collected using anonymous, self-reported, and paper questionnaires from March to May 2021. To collect data, the first researcher referred to the study setting, distributed the questionnaires among the eligible participants, and explained the study objectives. He also trained prisoners on how to fill out the questionnaires. Finally, 259 prisoners completed the questionnaires and six prisoners did not complete questionnaires . four tools were used in the current study. --- Demographic information questionnaire It includes age, education, nationality, occupation, number of marriages, length of stay in prison, history of imprisonment, history of drug abuse and its type, history of mental illness, psychiatric drugs, and smoking and type of crime. --- Ryff psychological wellbeing scale Ryff developed this self-report scale in 1989. The short 18-item PWBS with six subscales was used in this study: selfacceptance , autonomy , positive relationships with others , purpose in life , personal growth and environmental mastery . Ryff confirmed the reliability of the original version of PWBS using Cronbach's alpha coefficient, as well as its validity through content validity . Researchers confirmed the reliability of the Persian version of this questionnaire using Cronbach's alpha coefficient and experts confirmed its validity . The PWBS is scored based on a six-point Likert scale, including strongly agree , somewhat agree , a little agree , a little disagree , somewhat disagree , and strongly disagree . The total score in this questionnaire is between 18 and 108, with scores of 18-48, 49-78, and above 79 considering as poor, moderate, and favorable PWB, respectively. --- Family standardized cohesion questionnaire Fischer et al. developed this questionnaire as part of the California Family health Project. The questionnaire has 13 items and four dimensions, including cohesion , cooperation , clarity of rules , and leadership in the family . the questionnaire is scored based on a six-point likert scale ranging from strongly agree , somewhat agree , a little agree , a little disagree , somewhat disagree , and strongly disagree . Fischer et al. reported the internal consistency of the original version to be 0.78 using Cronbach's alpha . Experts confirmed the validity of the Persian version of this questionnaire, and its reliability was above 0.70 using Cronbach's alpha method . In this study, scores of 0-2, 2-4, and 4-6 were considered as poor, moderate, and favorable FC, respectively. --- Purpose in life questionnaire Krombugh and Maholick developed this scale in 1969 to measure individual's sense of purpose or meaning in life. The questionnaire consisted of 20 items and participants chose one answer between one and seven. Scoring, type, and content of the items in this questionnaire are different with the subscale of purpose in life of PWB. In this questionnaire, the minimum and maximum scores are 20 and 140, with scores of 20-60, 60-100, and 100-140 considering as poor, moderate, and high PL, respectively. The reliability coefficient of the original version of PIL was 0.88 using Cronbach's alpha coefficient. The questionnaire has content validity based on Frankl's logotherapy theory . A study on 250 Iranians reported that the reliability coefficient of the Persian version of the PIL questionnaire was 0.92 using Cronbach alpha. The validity of the Persian version was confirmed by studying the correlation of its scores with life satisfaction, vitality, and positive-negative affection . --- Data analysis The data were analyzed by SPSS 21. First, data normality was determined by the Kolmogorov-Smirnov test. Regarding the normality of quantitative data, the Pearson test, and multivariate linear regression were used. The significance level was considered 0.05. --- Results --- Demographic information Majority of the participants were 20-40 years old , married , Iranian , and had diploma . Table 1 shows all demographic information of the male prisoners. --- Description of PWB, FC, and PL The results showed that the total mean PWB score of the male prisoners was moderate , with purpose in life subscale and environmental mastery subscale receiving the highest and lowest mean scores, respectively. The total mean FC score of male prisoners was moderate , with the clarity of rules and cohesion subscales receiving the highest and lowest mean scores, respectively. The total mean PL score was moderate in male prisoners; the item "I wasted the rest of my life after retirement. I did the most exciting things I ever wanted to do" received the highest mean score, while the item "I am very responsible. . .. I am very irresponsible" received the lowest mean score . --- Correlation between variables The results of this study showed a direct correlation between male prisoners' PWB and FC scores , but PWB had no significant correlation with PL score. In addition, FC score was not significantly correlated with the PL score. All dimensions of PWB had a direct correlation with FC except for positive relationships with others. In addition, all dimensions of FC had direct correlation with PWB. None of the PWB and FC dimensions were related to PL . --- Regression analyses The multivariate linear regression showed that FC , marital status , and length of imprisonment were the significant predictors of PWB. Therefore, married prisoners with a higher level of FC and imprisoned between 6 months and 2 years had higher scores in PWB . --- Discussion The current study showed moderate PWB, FC, and PL, as well as a direct correlation between PWB and FC in male prisoners. One study on young prisoners in Iran supported our results and reported moderate PWB of the male prisoners . Modarres et al. indicated that counseling and training were effective in promoting mental health indicators among prisoners . They believed that prisons were not closed and passive environments and do not merely seek to punish individuals; rather, they considered other aspects of their lives. Such results may be due to the organizational goals of prisons, including therapeutic, psychiatric, social work, rehabilitation, and correctional activities in prisons. Therefore, relevant experts perform a wide range of activities, such as cultural, religious, artistic, scientific, and vocational programs, which can be effective in improving the PWB of prisoners. In addition, the factors and conditions that prisoners deal with before imprisonment may affect their PWB. As PWB develops over time, it can be concluded that prisoners have suffered psychologically before their imprisonment. Moreover, depending on the psychological and personality characteristics of prisoners, they might have had various exciting experiences in their past lives. Purpose in life, the PWB subscale, received the highest mean score, which may be due to the type and content of the items in this subscale. For example, "Sometimes I feel I have done everything in my life" indicates that the prisoners are constantly worried about employment, family, financial status, and other important issues of post-release life, so they have the opportunity to think about their lives as well as learn the psychological training provided in prison. Environmental mastery received the lowest score, which was expectable because of prisoners' conditions and limitations. Prisoners are in a limited environment, have no choice in sleeping, waking up, exercising, training, or other programs, and must perform all their daily activities collectively and systematically. The results revealed moderate FC of the male prisoners; social work and cultural activities in prison may have contributed to improving outcomes. The social work sector in prison attempts to improve conditions by strengthening prisoners' family relationships and interests through phone calls, in-person visits, and temporary release from prison. Datchi et al. confirmed the role of family support and family -centered programs in bringing prisoners back to society with favorable wellbeing and family conditions . Hall et al. concluded that identifying prisoners' skills and strengths strengthened their cooperation and involvement with their family members , but a study on family members of individuals with substance use disorder reported a poor level of FC . Our results showed that the male prisoners' PL was moderate. Various factors may affect prisoners' PL, such as socially adverse conditions, poverty, mental health problems, and lack of awareness of the PL skills before imprisonment and disruptions in the process of living and planning due to imprisonment. When people are imprisoned, their planning and PL are disrupted. The highest mean score was related to item "I wasted the rest of my life after retirement. I did the most exciting things I ever wanted to do". This result can be explained by prisoners' environmental and psychological conditions. Due to the physical and social constraints during their imprisonment, prisoners feel sad about exciting recreation that cannot be done in prison. Therefore, doing something exciting has become a wish for someone who has been in prison for several years. The lowest score was related to item "I am a very responsible person. I am an irresponsible person." Prison limitations made prisoners not feel good about responsibility, so they did not get good scores. The results of this study showed that PWB scores of male prisoners were significantly and directly correlated with their FC scores and FC was a predictor of PWB. These results are consistent with those of other studies . One reason for this similarity may be related to the important role of the family in different aspects of life, including mental health. Boyraz and Sayger emphasized that increasing parental FC and wellbeing might contribute to the health and wellbeing of children . Farajzadegan et al. found that family functioning had a direct and indirect impact on quality of life and wellbeing and that family was important in improving wellbeing . Our results showed no significant correlation between PWB and PL in male prisoners. These results were not consistent with previous studies , indicating that PWB was significantly correlated with PL score. One reason why these studies are inconsistent with the present study may be that the studies were conducted on normal groups, while prisoners are vulnerable with special circumstances. The results of this study showed that the FC score in male prisoners had no significant correlation with their PL score. This result was not consistent with two studies , which reported an important role of the family function in identifying and finding meaning in life. Another study emphasized the impact of family on PL with the mediating role of sense of belonging . The reasons for this discrepancy may include the type of training and social work provided in prison, as well as the conditions of prisoners. In addition, several factors may affect the correlation between these variables, which have not been examined in this study. For example, PL is based on the time and environment of the individual, and prisoners' environment and decision making are limited. --- Limitations The study population was male prisoners in a prison in southeastern Iran. Therefore, generalization of the results to other communities should be performed with caution. It is suggested that longitudinal studies be conducted to provide more detailed information about the changes in the studied variables from prisoners' imprisonment to their release in different cultures and contexts. --- Conclusion The results showed that male prisoners' PWB, FC, and PL were moderate, and the prisoners' PWB had a direct correlation with their FC. The results highlighted some important practical implications for the healthcare providers, in particular, community health nurses and social workers. They are recommended to design specific interventions, such as rehabilitative, supportive, cultural, and educational programs for prisoners. Researchers should explore the effectiveness of these interventions to improve prisoners' PWB, FC, and PL and determine factors affecting these variables, as well as the investigative needs of vulnerable groups, such as prisoners. --- Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. --- Variables Family cohesion Purposeful life --- Cohesion Cooperation Clarity of rules --- Leadership --- Total family cohesion --- Ethics statement The studies involving human participants were reviewed and approved by this manuscript, derived from a master's thesis in nursing , was approved by the Ethics Committee of Kerman University of Medical Sciences . At the request of the Ethics Committee, the present study was conducted in accordance with the Declaration of Helsinki and Ethics Publication on Committee . The researcher presented the letter of introduction to collect data and received permission from the administrative division of prison. He explained the research and its goals to the participants and obtained their consent. The confidentiality of information was respected and the results of the research were reported to the relevant setting. 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.

Background: Psychological wellbeing, family cohesion, and purposeful life are important determinants of the prisoners' overall wellbeing and health; therefore, their evaluation is extremely important in prisoners as a vulnerable group. Objective: This study evaluated psychological wellbeing, family cohesion, purposeful life, and their correlations in male prisoners. Methods: This cross-sectional study used simple random sampling to select 259 male prisoners. Data were collected using questionnaires of Ryff psychological wellbeing, Fischer family cohesion, and Crumbaugh and Maholick purpose in life. Results: Majority of (78%) the participants were 20-40 years old and married (59%). The mean scores of psychological wellbeing, family cohesion, and purposeful life of the male prisoners were moderate. Psychological wellbeing was directly correlated to family cohesion in male prisoners, but it had no significant correlation with a purposeful life. Family cohesion was not significantly correlated to a purposeful life.Regarding the moderate level of psychological wellbeing in prisoners, it is suggested to pay more attention to educational and supportive programs in prisons for promoting such indicators in prisoners.

In Government's review of waste policy in England , it pledges to work in partnership with local authorities and businesses to facilitate the uptake of best practice in waste prevention and resource management, amongst other ways through: -Reducing barriers to innovation and wherever possible the burden of regulation on compliant businesses, -Supporting capacity building in local communities and giving them freedom to take initiatives in service design and provision. The Review also includes reference to using the Localism Act 2011 and Big Society concepts to "empower local communities". It promotes a partnership between local communities and allows community stakeholders to participate in the decision making processes with regard to how the community deals with environmental issues, including waste management . This approach builds on the recommendations of the Sustainable Development Commission that concluded that Government policy should prepare the ground for communities to deliver sustainable actions, coordinate support and provide access to funding . Within the context outlined above the role of voluntary and community sector organisations have been viewed as the agents of change at the local level, to deliver and facilitate more sustainable strategies . However, while there has been much research on community level initiatives to promote sustainability, there is limited practical evidence to guide policy, or indeed to inform successful public engagement with such initiatives . According to Weerawardena et al. , the issue of building long-term sustainability within the third sector is fragmented and relatively under developed. Using three towns in the East Midlands of England as case studies, this study aimed to examine and understand the levels of public engagement with VCS organisations and the factors that influenced this engagement. --- VCS organisations and the sustainability agenda VCS organisations are "groups within which individual members associate of their own volition with others in the pursuit of common objectives" . The term 'Voluntary and Community sector' is generally applied to entities that are value driven and which principally reinvest their surpluses to further social, environmental or cultural objectives . The sector ranges in scale from small, local charities to nationally important bodies. Due to its range and diversity, segmenting it by type of organisation can be helpful, particularly for sharing best practice and maximising the impact of sector activities. Voluntary organisations working internationally are sometimes referred to as nongovernmental organisations . Community groups may be differentiated from voluntary organisations . These entities are collectively referred to as voluntary and community organisations or the voluntary and community sector , . Various VCS organisations form a cornerstone of civic society and an institutional basis of society's 'third sector'. In many countries they have grown considerably in scope and scale within recent decades, and play a key role in engaging with and empowering society in a manner in which top down approaches are unable to . For example it is at the community level that notions of climate change adaptation and mitigation, and wider sustainability should best be developed and implemented at the local levels . In this way, opportunities to embed these concepts into existing mandates, budgets and governance structures at the local and regional levels can best be realised . There are also emerging narratives around the significance of community structures as means of maintaining the fabric of the community by embedding and facilitating notions of urban resilience . Thus community governance structures such as third sector bodies and engagement with these structures play a vital role not only in promoting sustainability, but also in enabling the capacity for change within communities and the society at large . --- 2.Methods Two main approaches were employed in the project, namely a: meta-analysis and questionnaire surveys. The meta-analysis utilised both academic and grey literature to examine the nature and key operations of VCS organisations, in the UK and beyond. The selected papers were obtained mainly from the Science Direct database, as well as key UK Government agencies such as the Waste and Resources Action Programme , Department of the Environment, Food and Rural Affairs , the Audit Commission, the Office of the Third Sector, and publically available environmental consultancy reports. The questionnaire surveys were undertaken in three adjacent towns, in the East Midlands region of England, namely: Luton, Milton Keynes and Northampton. The aim wasn't to be undertake a statistically significant sample from each town, but rather to capture a snap shot in terms of the socio-economic profile of the population, as well as the main activities of the organisations operating in the three towns. The surveys were conducted during the first three weeks of April 2013, with Milton Keynes being first, followed by Northampton and lastly Luton. Each of the three surveys employed 50 questionnaires, with respondents randomly selected from people walking in the streets, usually the high street. Respondents were handed the questionnaire and it was collected back on completion. A total of 149 questionnaires were completed and returned. The main purpose of the questionnaire was to determine public perceptions, attitudes to and use of the organisations in the case study areas. The questionnaire, a four page document with an introduction and contact information for reference and feedback purposes, was divided into four sections, namely: Generic questions on recycling and shopping from charity shops; Evaluation of shopping practices at furniture reuse shops; Examination of general environmental attitudes and beliefs; and Socio-demographic information. The majority of the responses utilised tick boxes and Likert scales ranging from 1 -5, with 5 indicating agreement and 1, non-agreement. There were also spaces for open questions provided. To facilitate the analysis, all questions were coded. Likert scales ranging from 1-5 were used to code the questionnaires. The coded information was initially entered into MS Exel and these data were then transferred into SPSS for analysis. The normality of the data was first ascertained using the Shapiro-Wilk Test. As p>0.05, the data were assumed to be parametric. Descriptive analyses were first performed to determine frequencies. Bivariate analyses, using Pearson's Correlation were then undertaken to examine and evaluate the nature of the key antecedents to the perceptions and attitudes of the participants towards the organisations, and their use of the shops. --- Results --- Meta-analysis Reported numbers of the organisations in the UK vary. For example, according to Curran and Williams , there have been approximately 400 such agencies set up in the UK since 1970. However, Defra note that there are about 1000 such entities in England alone. While LR Solutions and London CRN state that there were 693 third sector reuse organisations and initiatives in London, divided into six main categories, namely: • Domestic furniture and appliance reuse organisations • Computer and other IT reuse organisations • Charity shops • Other reuse organisations • Internet-based exchange forums • Locally organised swap forums -give or take days Despite the discrepancy in numbers, it is generally agreed that most are small and localised, with income of less than £250,000/annum . However, there are around 7% which have an income of over £1m and 20% operate regionally and nationally. For example, half of the principal reuse organisations in London serve one to three boroughs, with only 11% serving seven to ten boroughs . They have diverse scales of operation, activities and objectives. For example, they deliver public services , including waste management and in particular re-use , bulky waste services , community composting and waste minimisation activities . In 2009, nationally, the sector was diverting around on average 500,000 tonnes of waste from landfill per annum . For example, Featherstone states that reuse operations in London collect around 12,000 tonnes. The types of materials managed by the sector varies, with only 4% of items donated to charity shops ending up in landfill. According to Defra , textiles recycled through charity shops added up to 20% of the tonnages of material handled by other community organisations. By 2013, some 96% of materials were recycled or sold on for reuse . Approximately 20% of furniture collected at the kerbside is reusable and around 40% of the materials at household waste recycling centres are reusable . They also undertake a significant social role. For example, they redistribute the furniture and electrical appliances to vulnerable households, thus alleviating poverty, and combating social exclusion . Indeed, as illustrated in Fig. 1, poverty alleviation is key factor in their objectives. Featherstone argues that reuse organisations play a vital role in social housing, by facilitating the sustainable management of housing stock . In the UK, there is an increasing demand for reuse furniture, with child poverty a pressing problem, with two in five households having no working adult. Thus their benefits to society include : -The ability to deliver services in response to local demand, particularly with respect to niche materials and/or particular social needs -Making goods and materials available that would otherwise be unaffordable for those on a low income -An emphasis on, and ability to meet, the specific needs of service users and the local community --- Questionnaire survey Over half of the sample was female . There were more females, 64% interviewed in Luton than any other town. However, according to the ONS , the 2011 population census shows that the number of males and females in Luton is almost similar . The high percentage of female respondents could be attributed to reluctance to participate by men. The majority of the respondents for the three towns were in the 16-30 years age group followed by the 31-45years age group. Some 50.3% of the population was in some form of employment, with Northampton having the highest part time employed and Milton Keynes the lowest . However, the figures show that there are far less people in full time employment than part time across the three towns. Luton had the highest number of students . Some 57.7% of the sample had an annual income of less than £20,000. All three towns showed similar levels of income, except in the £30-50,000 range where Northampton had a significantly higher figure than Luton. Over 50% of the individuals in all three towns resided in rented houses, with Luton having the highest rentals at 65%. The most common houses were semi-detached and flats. House ownership was highest in Northampton at 58%. --- Recycling and use of charity shops Eighty-one percent of all the respondents said that they recycled. Table 2 indicates that residents of Northampton were the most likely to use charity shops , followed by Milton Keynes and Luton . However, on buying or taking items to furniture reuse shops, Luton had the least respondents , Milton Keynes had 62% and Northampton was highest at 80% . As shown in Fig. 2, clothes were the most donated item at a combined percentage of 31% for all the three towns, while furniture donation was 6.7%. Other items included books and toys. Over 34% of the respondents said they made monthly donations and 28% donated on an annual basis. Fig. 2: Items donated and bought across the three towns --- Awareness of the existence of furniture reuse shops An overwhelming majority of respondents , were aware of the existence of furniture reuse shops in their area and had shopped in them. Residents in Luton had the highest lack of awareness about them . Overall, 34% only came to know of the existence of the shops either by walking or driving by. As demonstrated in 2, about 50% of the individuals stated that they would take their unwanted sofa to a charity, while 40% said they would take it to the tip or throw it away. --- Environmental attitudes and beliefs The main reasons why individuals used the shops were because of the value offered and quality, as demonstrated by these correlation factors being highest . There was also some correlation with gender, with women expressing a greater engagement, compared to men. However, this may relate to the larger percentage of women surveyed. Perceptions about helping the environment and also living in close proximity to furniture reuse shops were also key deciding factors. Alternatively, of those that said they did not buy from the shops, 30% preferred to buy new items. However, they could not unequivocally say whether items from the shops were of good quality or not. Most reported strong beliefs on the positive role of recycling on the environment. An overwhelming 78% thought that their consumption patterns had an impact on the environment and 88% agreed that it was their personal responsibility to look after the environment. Asked what would encourage them to use furniture reuse shops, 81% suggested quality of items, while 63% thought awareness of the existence of a shop within their local area might have an effect on their shopping habits . --- 4.Discussion While accurate numbers of VCS organisations working in waste management across the UK, vary, they evidently play a significant role in society. This is evidenced both in terms of their environmental role of collecting, reusing, minimising waste, but significantly also, their social role helping vulnerable households . Across each of the three towns there was high awareness of furniture reuse shops, and most used them to either buy or sell items. Clothes were found to be the most donated item for all three towns, with just over a third of residents stating that they donated items at least once a month. This confirms the importance of textiles as found in other studies and suggests the key role played by textiles in achieving reuse targets . The socio-economic composition of the population appeared to have had some impact on engagement with the shops. Indeed, most residents across the three towns were in P/T employment, earning less than £20,000/annum, and within the age range of 16 -30 years old. These factors would have had a bearing on disposable income thus affecting lifestyles, as well as behaviour and attitudes towards resource consumption . It's important to note though that use of charity shops was highest in Northampton, where incomes were highest. This perhaps could be related to the higher quality of items in the shops. It would be interesting therefore to extend the study into areas where disposable incomes are higher. An overwhelming majority of the respondents were aware of charity shops in their towns, primarily via either walking or driving by. This would suggest the importance of location to such shops to ensure good public engagement. The perceived value offered in shopping at these stores would also suggest that price is a significant contributing factor in public engagement. Interesting, perceived benefits to the environment were also reported as a key reason why residents used the shops . Presumably this relates to perceived minimisation of waste through reuse and recycling of items, which if they had not been donated/sought/bought from the shops, would have been disposed to landfill. At the same time, it may simply be respondents telling the researchers what they believed they wanted to hear. Nevertheless, the findings suggest that notions of value, quality and environmentalism should form keycomponents in the marketing strategies of these entities. At the same time, the results showed that costs and good quality were key influencing factors, with 30% of the respondents preferring to buy new items. This raises two key issues with regard to continued and future public engagement. First, there exists a dichotomy between price and quality. If the quality of items donated to these entities continues to fall, then regardless of the 'low' price, residents may decide not to engage with them and not use the shops. Second, these organisations need cash flow to sustain operating costs and if there is no one to buy and no items to sell, they will struggle to fund their operating costs and remain viable . Thus while the study suggests that there was generally good awareness of the organisations in the three towns and they were fulfilling a need for individuals seeking value, there are a number of issues that need to be addressed to further improve public engagement. --- Conclusions VCS organisations working in the waste management sector have a key role to play in maintaining the environmental and social fabric of communities and the society. While, there was strong public awareness and engagement with the furniture reuse shops across the three towns, there were also a number of challenges identified that impacted upon public engagement. Key amongst these included overcoming perceptions of the quality of items, value and the location/visibility of shops. It is vital that they not only offer value, but also good quality items. A good location to ensure visibility is evidently also crucial. Greater awareness building and perhaps even a 'rebranding' exercise to move away from public perceptions of the organisations offering primarily 'poor' quality items, may very well be required. Faced with competition from entities such as eBay and low cost retailers, these are issues that need to be urgently addressed if the public is to be better engaged and ultimately the long-term viability of these organisations assured. Crucially, outside of their role in waste management, these organisations evidently also play a key role in addressing the social needs of society. Indeed, the meta-analysis illustrated that poverty alleviation was a more important issue than waste management. Indeed, glimpses of this are also seen from the survey in that value was a key contributing factor for why individuals engaged with the organisations. In addition, income levels in the three towns were low, thus the shops were a manifestation of the lifestyle choices of residents. It is possible that the timing of the study, during the economic down turn in the UK may have impacted upon perceptions and public engagement. An understanding of whether this is indeed the case would require a follow up study. Given their environmental and social roles, it is vitally important that the sector remains relevant, in tune with the needs of the public and ultimately therefore, resilient.

Within the context of shifts towards the concepts of resource efficiency and circular economy, voluntary and community sector organisations are increasingly being viewed as agents of change in this process. Using questionnaire surveys across three towns in England, namely Northampton, Milton Keynes and Luton, this study aimed to understand public engagement with these organisations. The findings suggest that there were generally high levels of awareness of the organisations and strong engagement with them. Clothes were the items most donated. Key reasons for engagement included the financial value offered and the perception that it helped the environment. However, potential limitations in future public engagement were also determined and recommendations for addressing these suggested.Recent European Union (EU) and UK Government policy and legislation have sought to encourage a more sustainable approach to the management of resources (EC, 2008; Defra, 2011;Williams et al., 2012). The revised EU Waste Framework Directive (rWFD), (EC, 2008), transposed via the Waste (England and Wales) Regulations 2011, identifies waste as a resource giving greater priority to waste prevention and ''preparation for re-use''. Indeed, for some time, Government has sought to implement policies and to encourage enhanced efficiency and recovery measures (Defra, 2005;2007;2011).

Introduction In the context of our country's long history of social injustice, the issue of systemic racism in higher education has recently gained a lot of attention. Constantly, African American students have suffered injustice that has reach a tipping point. The call to action for combating systemic racism has sparked a cultural phenomenon that has made inroads in all walks of life. Consequently, this study seek ways to explore, compare, and contrast the factors or reasons linked to the systemic racism that influences African American students' decision-making when deciding to join an institution of higher education. The African American students' experiences in the U.S. are unique. Most importantly, there are many factors that are beyond students' control that have shaped it, such as historical and economic issues. However, this is a top priority in respect to student's college choice. Historically, the academic journey from early school for African American students has not been straightforward. In schools, African American students are more likely to be labeled mentally challenged or learning disabled and overrepresented in special education programs. In addition, they are less likely to appear in talented or gifted programs such as Honors classes or Advanced Placement courses . Furthermore, they are subjected to expulsions and suspensions and the retaining rate is alarming. One out of five African American students received an out-of-school suspension, and they are 3.5 times more likely than their Caucasian peers to be suspended or expelled . Consequently, many African American students lack adequate opportunities to develop positive classroom experiences. They are often berated for minor classroom behavioral issues that become magnified by teachers and administrators . Over the years, researchers have identified three major problems that have emanated for African American men before they enter into higher education such as: lack of quality teachers in K-12 education, overrepresentation of African American male students in special education programs, and a lack of African American male students as well as other students of color in advanced placement courses at the secondary education level . According to Bottiani, Bradshaw, and Mendelson , students' perceptions of differential treatment, exclusion, and discrimination by teachers and other adults in school appear to play a significant role in the poor outcomes among youth of color. However, there are lot of barriers when it comes to admitting more Black students to the nation's selective universities. The climate change surrounding racial issues and the U.S. economic downturn have had a significant impact on the commitment of students of color in respect to higher education . Historically, Black Colleges and Universities have played a significant role in expanding educational opportunities in the United States . This study attempts to answer the overarching research question: How does systemic racism play a role in African American students' college choice decisions and the factors affecting the decision? --- Theoretical Framework to Lead this Study This study is grounded in Critical Race Theory and College Choice/decision theory. The conceptual framework for this study was provided by the Critical Race Theory . This is because CRT explores and analyzes the experiences of students of color, which can inform crosscultural inclusivity practices in a number of ways. However, CRT was appropriate for this study because it presents practical techniques that can be used by academic advisors and other higher education administrators when working with students of color. CRT can also serve as a theoretical framework through which a person can understand various ways on advising interactions and practices with students of color in order to offer help or exert harm . This study applied CRT because it challenges the status quo and examines racial difrences in myriad contexts, such as the many ideologies and norms embedded in educational institutions . Furthermore, CRT has been used to normalize and analyze racialized experiences in research and practice by challenging traditional paradigms, methods, texts, and discourses on race, gender, and class. Based on the different models within each of the college choice and decision theories, there is an understanding on why African American students might choose to attend an HBCU or Predominantly White Institutions . Models by Iloh , Glasser , Hasen and Litten , and Freeman provide an explanation of college selection and decision-making aspects within each example to support the study objectives. --- Methodology The present study is a non-experimental qualitative comparative case study analysis. This study will use qualitative research, which entails collecting and analyzing non-numerical secondary data, in order to understand concepts, opinions, experiences, and phenomena. The aim is to gain in-depth insights into the problem, answer the research question, or generate new ideas for future research. Qualitative researchers engage in naturalistic inquiry and explore real-world settings by inductively developing rich narrative descriptions of the cases they are investigating. Inductive analysis across case studies yield patterns and themes which is the fruit of qualitative research . The purpose of this study is to analyze data for insights within, across, and between cases . Therefore, this comparative case study design involves comparing, contrasting, analyzing, and synthesizing the similarities, differences, and patterns across three selected case studies that share a common focus, goal, or phenomenon. This study also used secondary data sources from all three cases, which includes documentation, archival records, interviews, physical artifacts, direct observations, and participant observation. The researcher has collected and integrated data within the case study research, which is unique to other qualitative approaches. This has help in facilitating a holistic understanding of the phenomenon being studied. --- Data Collection This qualitative research is designed to reveal the meaning that informs the action, practices, or outcomes using rigorous and systematic transcribing, coding, comparing, analyzing, and interpreting the trends and themes with a focus on retaining the true meaning. However, such research is used to understand how people experience the world. Qualitative researchers often consider themselves "instruments" in research because all observations, interpretations, and analysis are filtered through their own personal lens . This case study analysis involves intensive analysis of the following three selected case studies to understand and gain an in-depth holistic view of the research problem. i Consequently, these three research studies, conducted within ten years, were selected through extensive database searches using specific words. The titles and research questions from the selected case studies were extracted and compared using a color-coding method to show internal consistency. --- Data Analysis Procedure The researcher extracted the relevant data from the selected case studies and utilized this data for secondary data analysis. The selected case studies were semantically coded to assess their inter-relatedness, understand the text data, label the emerging sections with codes, examine the codes for similarities, and then combine the codes into broad themes. To initiate the data analysis, the following preliminary themes were developed: Student, College, University, African American, Black, Racism, and Identity. The selected studies were uploaded to ATLAS ti9 to arrange and code the major themes. A word cloud was assembled by ordering the importance of the phrases from all three case studies. Next, a hierarchical network was constructed where the coding revealed an emerging themes. Furthermore, the codes were arranged into themes or categories to create a database of topics that were widely discussed in the case studies . This process give room for the researcher to observe the density, comprehend indepth understanding to answer the research questions, and ensure the credibility of the data analysis. Subsequently, themes were developed by utilizing the six-step processes suggested by Caulfield . --- Results --- Findings on Theme One: Systemic Racism in Higher Education In analyzing the three case studies, the first theme that emerged was Systemic Racism in Higher Education. In Case Studies 1 and 2, the findings of these research studies displayed an acknowledgment of racism as a factor in African American students choosing to attend an HBCU or a PWI. --- Findings on Theme Two: African American Students Decision Making In Case Studies 1 and 3, the process of choosing a college to attend was another theme that emerged. The decision-making factors of African American students are prominent themes in each case study. In this theme, three small sub-themes have been discovered: college choice and decision factors, reasons for choosing HBCUs, and reasons for choosing PWIs. --- Findings on Theme Three: Experiences of African American Students in HBCUs and PWIs In study 1, recent high school graduates from HBCU and PWI-bound have a high expectations of college life that are typical of experiences that will duplicate high school in a larger community. In study 2, the students at the historically black colleges and universities reported a better psychological and functional faculty support levels than their PWI counterparts. --- Findings on Theme Four: Minority Students' Stress and Anxiety in HBCUs and PWIs In case study 2, the report from African American students at PWIs were more stress related and exposure to racial discrimination are being underrepresented on these campuses. As codes emerged, the researcher could see linkages between the factors influencing college choice and the selected theoretical frameworks of college choice decision and critical race theory. In this study, CRT was used in analyzing how race and privilege play a major role in society and also how it plays a role in higher education specifically. In this study, college choice and decision theories were used. In addition, several college choice and decision theories were used to analyze how college students think and arrive at their decisions to attend a college or university. However, the findings in case study 1 helped the researcher to connect the case findings to other cases utilized in this study. In case study 1, the theme of systemic racism in higher education was also discovered. This is connected to the findings and other themes from other case studies. In case 1, "African American students use racial climate and the significant presence of a diverse or same-race student body to validate their final selection" . In case study 2, there were discussions about how African American students have more positive perceptions at an HBCU and also how the experiences of black students at PWIs were negative. These findings have helped in discovering minority students' stress and anxiety at HBCUs and PWIs, as a key theme emerging from the analysis of case study 2. African American students on an HBCU campus would report lower minority status stress levels than their counterparts on a PWI campus. Collectively, the current analysis suggests that institutional racism affects African American students' college choices. --- Conclusion While answering the research question, the findings of this study have revealed that the role that systemic racism plays in college choice decisions of African American students is displayed in critical race theory. This is because it is a crucial factor in the analysis of choice theory. In the context of systemic racism, the critical race theoretical lens that affects the college choice decisions of African American students shows that race is an embedded factor in college choice decisions because of a deeply rooted and racist normalized traditions. The cross-analysis of the studies confirms that African American students use racial climate and the significant presence of a diverse or samerace student body to validate their final selection . This research confirms that racism and many other factors play a crucial role in the college choice decisions of African American students in attending predominately white institutions of higher education. It is apparent from this study's findings to note that racism is an issue for African Americans in their college choice decisions. Consequently, the students' decision-making for higher studies or deciding on their careers is a complex process. However, there is still much to learn about how African American students choose a college education for themselves. It is instructive to engage in a racial analysis of the access and choice processes to better understand the oppressive structures encountered by college-bound African American students. Contrarily, the college enrollment decision-making process for students can, on the surface, appear fair and inclusive across racial lines, although this is not the case . --- Conflict of Interest: The author reported no conflict of interest. Data Availability: All of the data are included in the content of the paper. Funding Statement: The author did not obtain any funding for this research.

A lot of attention has been given to the problem of institutionalized racism in academia lately. From a young age, the road to success in school has not always been an easy route for African American students. According to Bottiani, Bradshaw, and Mendelson (2016), students' perceptions of differential treatment, exclusion, and discrimination by teachers and adults in school appear to play a significant role in the poor outcomes among the youth of color. The climate change surrounding racial issues and U.S. economic downturn have had a significant impact on students of color as regards to their commitment to higher education (Karkouti, 2016). This paper attempts to answer the overarching research question such as: How does systemic racism play a role in African American students' college choice decisions? The results from this paper are displayed in critical race theory, as it is a crucial factor in the analysis of choice theory.

. Introduction In the past decade the struggle for water and sanitation justice has been linked with alliances deploying informal strategies and campaigns for formal recognition of the human right to water and sanitation . The focus of this paper is on the struggle, discourses, and politics in Spain, departing from the European Citizens' Initiative "Right2Water" and examining links and similarities between the European Right2Water movement and Spanish water movement. The European Citizens' Initiative "Right2Water" ran and collected 1.9 million signatures across Europe in 2012-2013, uniting cities, villages and civil society organizations against water privatization . With that result it became the first successful ECI, simultaneously building a Europe-wide movement and putting the water issue high on the European political agenda. "Right2Water" proposed to implement the human right to water and sanitation in European legislation, as a strategic-political tool to fight privatization of public water utilities . Implementing such water justice notion is part of an ongoing socio-political struggle. This struggle had been going on in Spain before the start of the ECI and it has continued afterwards. Deploying a political ecology lens, this paper examines how "Right2Water" influenced and impacted ongoing struggles for access to and control over water in Spain; how it fuelled the debate on privatization of water services and what heritage it has left in Spain. Moreover, it looks at how social movements developed, reinforced or complemented each other in their struggles for social justice and against austerity measures that have marked the second decade of this century. Rise of the "Indignados" movement coincided with the rise of a water movement like the Red Agua Publica . In the past decade a growing number of cities and regions have decided to re-municipalize water supply. The human right to water and sanitation framework was instrumental in this process, as we will show in a few cases. The Right2Water movement changed water policy discourse in Europe and proved to be a support to the Spanish water movement, that continues to fight for a more participatory and democratic water governance and policy. In Section 2 we describe our research framework and methodology. In Section 3 we will look at the history of water policy and privatization in Spain. In Section 4 we will look at the rise of social movements and in Section 5 we will highlight a few cases of water struggles in Spain. In Section 6 we examine the stakes and stakeholders that, for the time being, have won and lost in these struggles, and the overall outcomes of the water policy and politics battles over the past decade. We draw our conclusions in Section 7. --- . Conceptual framework and research methodology The conceptual base of this research is political ecology, which is a political approach toward the study of natural resources management. Ecological problems are seen as deeply interwoven with and produced by the socio-political and economic context . This paper will analyse historical and political processes of water management in Spain, looking for underlying power relations . Directly connected to this, the notions of environmental justice and contestations/non-contestations are central. In academic and activist debates, environmental justice is defined in multiple ways-for instance, by integrating and balancing or contrasting the arenas of distributive justice, cultural justice, representational justice and socio-ecological justice, for human and non-human communities and agents . In this research, we concentrate on three of its conceptual dimensions: recognition, redistribution, and representation . We focus therefore not just on the distribution of access to water, but also about who participates in political decision-making processes and whose values and interests are recognized by, for example, social norms, languages, and institutions . With regards to public water services this means that citizens must be recognized as political actors and therefore co-decide and cogovern in water services management. All dimensions are distinct but interlinked and overlapping . Hydro-technological modernization processes and legalinstitutional policy reforms coalescing in diverse forms of neoliberal water governance provide the background to these last decades' powerful trends toward Europe's water governance model . Understanding neoliberal water governance conceptualization is key. In brief, neoliberal thinkers and policy-makers advocate treating water as an economic good. Water resources' service and/or property, therefore, need to be private, transferable, and priced-commodified. According to neoliberal logic, policy measures such as privatizing water service provision, granting concessions to operate distribution networks, and implementing full-cost recovery in water service pricing would lead to improved water service, and more efficient operation and maintenance. In water governance, the assumption is that such neoliberal incentives lead to efficient water use . In practice, however, many studies have shown that privatizing public utilities has often failed to benefit water users. Rather, tariffs hiked, investments in infrastructure lagged behind, quality of service provision did not improve, and the environment was jeopardized . In recent years, protests have been organized in various parts of Europe by citizens and civil society organizations to stop privatization of drinking water utilities or demand cancelation of these contracts, challenging the neoliberal governance structure and calling for a vision on water as a common . Finally, water justice movements address injustices both at the individual, community and supralocal levels . This is the case for Right2Water as well. Water justices involve both quantities and qualities of water, as well as access to and distribution of water privileges and forms of control over water . This entails also that water conflicts include questions about decision making, authority and legitimacy. These are intimately linked to the struggle over discourses, favoring particular water governance notions and policies while obliterating others-in terms of thinking about and acting upon "water" . As discourses sprout from, substantiate, and actively constitute policy arenas, they become a necessary part of the struggle of social movements: discourse change is not only a trigger for social mobilization and protest but also a precondition for policy change . Media attention, which is one of this paper's fields of scrutiny, is an indicator for the domination of particular discourses in a specific period of time, and it also reflects how governance processes and social movements are framed, while it indicates resonance or influence in society . The main research question that we seek to answer is how the European Right2Water movement has related to the Spanish water movement and how the human right to water and sanitation has been instrumental in their struggles against privatization of water services. Literature and archival research and activist debates in the Spanish institutional and movements' landscape, have been the main sources of information on human rights to water, privatization and re-municipalization of water services in Spain, and how social movements have campaigned for public water provision. Five cases are analyzed as examples of local struggles with diverse settings and histories of privatization and movements. The cases show different types of water governance and ways of activism and provide lessons for future debates over water services provision, governance and remunicipalization. Besides the cases, data collection was done for this study on drinking water service privatization, re-municipalization of water services in Spain and on debates around these issues and the human right to water in Spanish newspapers. We explored these issues and social movements more in depth by applying bibliometric methods and data mining. With text mining software we searched and analyzed the number of times that these issues and movements were mentioned in Spanish national newspapers and how this developed during the past decade between 2010 and 2022. We searched for unique articles on key words "human right to water" , "15-M, "ECI Right2Water" and "Marea Azul" in combination with "water services." Besides this we searched for "New Water Culture, " "privatization, " and "remunicipalization" also in combination with "water services." Thirdly we checked mentions of two social organizations that played an important role in water debates: "AEOPAS" and "FNCA" also in combination with water services provision. We visualized trends in media attention in two graphs. Given that we consider newspapers as relevant media that reflect public debates and public issues of interests in Spain, we looked at the number of unique articles that dealt with these issues. With this media analysis we expected to show if, and how, social movements have gained influence and attained to change discourse in the Spanish newspapers with regards to water services management. . History of water services provision in Spain . . Water policies, privatization, and rise of the New Water Culture Water management in the twentieth century in Spain was characterized by a hydraulic paradigm stemming from the Franco regime that aimed to "develop" Spain through water infrastructure imposed upon the population and based on use of water resources to maximize State economic benefit . In 2001 the Spanish government introduced the National Hydrologic Plan and proposed construction of over 100 dams with a similar view on water as an economic good . The NHP puts emphasis on technology to regulate rivers and use of water where it is the most profitable, fitting tightly to the ideology of Franco's regime . The plan was adopted after massive protests throughout Spain, and modification by the Spanish parliament in 2005 as Act 10/2001 on the National Hydrological Plan . Since the 1990's, there has been a trend in Spanish municipalities to privatize water services, mainly through public-private partnerships . This term is frequently used by multinational corporations to disguise a privatization through contractual arrangement, often a long-term concession. The Spanish privatization model of water services over the last three decades is often described as "mercantilizacion" . While being a natural monopoly, this marketization of water services tends to favor the interests of powerful groups at the expense of the public interest and the commons, as it is characterized by the absence of competition . In the last decade of the twentieth century and beginning of the twenty-first century the EU saw a wave of utility privatization . Spain was no exception to this phenomenon. Private companies in the water industry increased their share in provision of water services to the population from 37% in 1996to 48% in 2006to 55% in 2015 by focusing on water supply in urban areas because that can generate most profits in a "market" with many "customers" in a small area. Local governments often saw privatization as an easy way to increase their income at short term because of the concession bonus and by keeping water supply off their balance sheets . This gave rise to an intense debate about the desirability of this privatization of urban water services with a clear oligopolistic dominant position for multinational companies . Political decisions were substituted by market instruments that favored corporations because of the structural imbalance between these players and municipalities . Simultaneously, since the 1990's, a coalition of academics, social activists, and water managers in Spain and Portugal has been promoting a shift from the hydraulic paradigm and seeing water as an economic good . This New Water Culture placed the emphasis on ecosystem protection and ecological health as a means to guarantee the availability of sufficient good-quality water to sustainably meet needs. In 2000 the Water Framework Directive came into force in the EU . This European law needed to be transposed into Spanish law and water activists saw this as an opportunity to shift the priorities of the traditional economicminded water management system to a more holistic approach that recognizes ethical, socio-political, and environmental concerns . The New Water Culture movement was founded in opposition to dam building and the inter-basin transfer of water from the Ebro River. It emphasized participatory and transparent decision-making in water management and provided an alternative policy framework as it positioned water as a common and a human right. The movement coalesced in the New Water Culture Foundation or FNCA . One of the founders was Pedro Arrojo, former professor at Zaragoza University and now the UN special rapporteur on the human right to water and sanitation, who explained it as follows: "Water services as services of general interest of society go beyond the logic of the market" . "For this reason, faced with the privatizing policies of neoliberalism that transform these public services into simple businesses, we must face the challenge of building new models of public and participatory management, based on principles of sustainability, justice, transparency and citizen participation" . The economic crisis of 2008 generated an increase in situations of social vulnerability and problems related with affordability and accessibility to water in urban contexts . Privatization processes of water services coincided with austerity measures that European institutions promoted in the countries most affected by the economic crisis . European austerity policies aimed to reduce government budget deficits through cuts in public services, which usually include water services, and decrease government spending. These savings on government costs came at the expense of providing vital services to citizens . The European Commission imposed privatization of water services as one of the conditions of bailouts to crisishit countries Greece and Portugal [Corporate Europe Observatory , 2012; Zacune, 2013;Kishimoto and Hoedeman, 2015;Bieler and Jordan, 2017] and also in Spain the pressure of European austerity measures was high . In 2011 the European Commission made a new attempt to further liberalize the services sectors in Europe by means of a proposal for a "Concession Directive." This directive aimed to align concessions contracts with economic water operators and single market rules . The directive promoted to open municipal water services for EU-wide bidding and aimed to create a European water services market which would open the door to privatization . This was conceived by water activists as a "privatization through the back door" and in Spain the directive was therefore called the "Privatization Directive" . The Right2Water movements discontent with Europe's water policy and governance was exactly with this proposed directive and earlier attempts to apply internal market rules to water supply and the management of water resources, because it would mean the commodification of water . --- . . Privatization of water services in Spain In Spain, the process of privatization was instigated by the search for funding by municipalities in crisis. It happened mostly through concessioning as the perverse mechanism of the "concession fee" allowed for a rapid injection of money into the municipal treasury, but the backside of this coin was a decades-long privatization and loss of control over the service . This process was supported by European Commission liberalization policies. Transformation of a public-service ethos to a profit-serving ethos, combined with technocratic and "neutral" solutions to complex societal and political problems and commodification of natural resources , are part of the neoliberal project of completion of the European Single Market . Privatization occurred by two major corporations: Aguas de Barcelona and Aqualia , both multinational enterprises . These two companies account for almost 90% of the total private water supply . This privatization and the oligopolistic character led to an over emphasis on profit margins and resulted in environmental degradation by failing to comply with the environmental terms and conditions in contracts as market logics do not value or protect human rights, nor sustainability of ecosystems . The natural monopoly that is inherent to water supply; the power imbalance between municipalities and multinational private providers and the drive for profit maximization form a hazard to the fulfillment of the human right to water and sanitation in case of privatization . In the Spanish model of privatizing urban water management all these risks became visible and reality. Costs were born by the citizen while private companies took benefit from water resource exploitation by setting, on average, higher prices than public companies. The privatization process in Spain was often accompanied by increasing rates, sometimes up to over 60% . Local authorities lost control over the service and access to water provision was jeopardized for vulnerable people by greater pressure on users with payment problems and 300,000 shut-offs in 2013 . The relations between the causes and consequences of the crisis and privatization, as well as the emergence of situations of water poverty , have led to the rise of a social movement committed to defend the human right to water as well as the model of public and democratic management . . /frwa. . . Social movements for public water in Spain . . A movement for public water A number of Citizen Water Networks have emerged from the New Water Culture movement with an aim to defend the social and ecosystematical values of water . The FNCA is seen as one of the most influential movements in water policy in Spain . In the beginning of this century these citizen networks were mostly active in opposition to dam construction and river management but also raised a voice for, in their view, marginalized groups in water policies. They demanded influence in decision-making, that was dominated by economic interests of the state and powerful groups . The citizen networks created large-scale awareness raising campaigns that incorporated many different types of calls to action, but also local social mobilizations against dam projects . One of these networks was the "Red Agua Publica" formed in 2012 which focused exclusively on urban water privatization, building on the Plataforma Contra la Privatización del Canal Isabel-II that saw light in 2010 in Madrid. This case will be discussed in the next section. The RAP rose from a local opposition to privatization of the water company in Madrid to an extensive network of organizations in Spain. Until today it promotes a vision of water as a common and a public service with the following objectives: defending the integral water cycle as a public good; supporting struggles against privatization of water services and for remunicipalization of already privatized services; transparent public water management with citizen participation and effective achievement of the human right to water. These objectives were completely in line with the objectives of the European Right2Water movement that emerged in the same year. The recognition of the human right to water with resolution A/65/254 by the United Nations General Assembly in 2010 was a victory for global water activists as well as an inspiration for local activists . The Italian Water Movement and the referendum they organized, inspired the European Public Services Unions to organize a European Citizens' Initiative on water framing it as a struggle for water as a human right and as a public good. By asking civil society organizations and local water and social movements for support to this EU-wide campaign the Right2Water movement was born. It reciprocally gave support and inspiration to other networks calling for citizen participation in water management and fighting privatization of water supply . From 2012 onwards, the European Citizens' Initiative "Right2Water" united people and social organizations against water privatization and liberalization policies across Europe . The Italian Water movement and their successful referendum against water https://redaguapublica.wordpress.com/about/ privatization also motivated Spanish activists as they maintained a distance from traditional political parties and called for water to be considered a common . Spanish water movements also used the phrasing "write water, read democracy, " introduced by the Italian Water Movement to emphasize this link . Widespread disappointment with the service that was provided by private companies resulted in reversals of the decisions to privatize water services. Such a reversion has happened in municipalities in developed countries , as well as in developing countries . The remunicipalization of water services in Paris in 2010 was an example for other cities and water activists . Municipalities wanted to take back control over their resources so they could better provide for the needs of their citizens and remunicipalization turned to a global trend in the second decade of this century . Whereas, Spain counted only one case of "remunicipalization" in 2006, the number of re-municipalizations of water services had risen to 14 in 2014 and 40 in 2022. Compared to the total number of 135 cases of water remunicipalization worldwide it shows that the Spanish water movement has taken a leading global role in water remunicipalization . The main reason for remunicipalization in Spain was the same in all cases: need for investments, guaranteeing universal provision, improving quality of services and taking back control over the service . Moreover, water prices in Spain are on average lower when services are managed by governments and tend to be higher under private management . Average tariff in Spain was about 0.89 e/m 3 in 2012, but the highest tariff at that time was in Murcia , followed by Barcelona and Alicante . In all three places water was provided by subsidiaries of Suez . Another reason for re-municipalization was the demand for more transparency and democratic control in order to fight corruption , which was seen as a democratic revolution at the local level, that went beyond water and sanitation management . --- . . Confluence of austerity protests and water protests Protests against the austerity agenda spread, adopting an innovative format in the so-called "indignados" and Occupy movements that occupied public spaces throughout member states of the EU . In Spain austerity related policies triggered many protests early 2011 where "Democracia Real Ya" developed and grew within 3 months into a protest platform with over 200 organizations affiliated to it . DRY called for demonstrations to take place in cities across Spain a week before the country's regional and municipal elections of May 2011, demanding radical changes in Spanish politics and an end to austerity policies. Mass demonstrations were held across the country on 15 May 2011. The largest of these protests was held in Madrid where demonstrators chanted "we're not merchandise in the hands of politicians and bankers" . After attempts by the police to remove the protestors, thousands of supporters occupied squares across Spain to express support for the activists in Madrid. From these solidarity camps, 15-M was born. The activists demanded a reform of the political system and an end to corruption . The involvement of many different actors in the 15-M movement made democratic participatory processes critical to maintain engagement. Since actors can have different backgrounds and motives for joining a movement it is essential to check and debate whether all participants are in the struggle together and to move forward . Protests continued over 2012 and a confluence of tides occurred after the 15-M movement opened up to a new social ecology of critical spaces to protect vulnerable groups . Different social movements saw their interest threatened by the same neoliberal ideology that increased pressure on workers and citizens, hitting hardest to the poor and caused a wide spread sentiment of people losing control over their lives . Trade unions in Spain joined their forces with 15-M social activists, ecological and water activists. Protests against austerity confluenced with the "marea azul" protests against privatization of water and in defense of water as a common good . In 2013 the RAP called on all European public water operators to leave EUREAU after it had denied the success of the ECI Right2Water, showing its concern of the interests of private operators over the public operators. An impressive list of around 100 groups and organizations from Spain signed the appeal . The Right2Water European Citizens' Initiative was transferred to Spain by the trade unions affiliated to EPSU-FSC.CC.OO and UGT.SP-and the New Water Culture Foundation. Pedro Arrojo was member of the "citizens committee" of the ECI with an extensive network in Spain. According to him "the ECI was at that time the motor of the struggle for the human right to water in Europe" . At first the campaign did not strike a chord but that changed when the unions and the water movement bonded with the anti-austerity 15-M movement and then reached a successful number of supporting signatures. The cooperation between the movements was complicated as both the austerity measures as well as the ECI were seen as European Union policy. How to combine action against EU policy and at the same time in favor of another EU policy? This psychological barrier had to be overcome. In the end over 65,000 people signed the Initiative , whereas 40,000 were necessary to surpass the quorum, despite of practical barriers . The encouragement of the Association of Public Water Supply and Sanitation Operators , the state-wide Public Water Network , Ecologists in Action, the Federation of Neighborhood Associations and the Platform against the Privatization of Canal de Isabel II played a role in the success for Right2Water. The Spanish case is a national example of a political achievement reached by uniting diversity as the Right2Water movement did in Europe . The collaboration of trade unions with social movements was very important. This had not occurred before. Also the combination of social and ecological values and objectives showed to be fundamental for a broad support for implementing the human right to water beyond anthropocentrism . Spanish water movements have been effective in linking the issue of water with other campaigns e.g., around housing or democratic platforms . These organizations continued their collaboration in "#Initiativa2015" by which they campaigned to implement the human right to water and achieve a "100% public" service provision in Spain . This initiative tried to prohibit supply cuts; to ensure a minimum of between 60 and 100 l per person per day in case of justified non-payment, and to eliminate the participation of private companies in water services. It was about developing a new model of public, democratic, and participatory management . This initiative was materialized a year later in the Social Pact for Public Water signed by 75 organizations at the start . It led in several municipalities to measures to guarantee a vital minimum of water and prohibition of supply cuts . The Pact addressed water justice issues and demands that went beyond the topics included in the human right to water and sanitation and provided a level of specification for its effective implementation . Several propositions to implement the HRTWS in legislation had been made but these efforts did not result in its implementation . It shows that the process of incorporating the HRTWS into the national water law, is complex and troubled. Until 2020 and despite the strong social movement and sometimes formal political support, this incorporation has not occurred . Until today no mention is made of the human right to water in Spanish water law, but the PSAP forms a basis for water policy in municipalities and public utilities as shown e.g., in the Cadiz Declaration in which several majors, water utilities and social organizations in Andalucía state their commitment to implement the human right to water and to ensure that water services remain in public hands . The PSAP is also at the basis of the declaration for public water services management by majors from 10 Spanish cities at the meeting for public water in Madrid on 3-4 November 2016 . --- . . Media attention Attention in the media shows how social movements have been able to take the stage in water services debates widely . The human right to water as recognized by the United Nations General Assembly was mentioned in the media since 2010 but in 2011 attention to water services provision expanded thanks to the 15-M movement. More attention for the human right to water followed in the wake of the European Right2Water citizens' initiative and the "marea azul" manifestations. The Spanish water movement managed to maintain an important voice in water services debates but media attention decreased. --- . Examples of local water struggles in Spain When the Right2Water campaign was launched in Spain, it could build on an already present network of social movements . These networks appeared indispensable to get people to join the European campaign. This was not easy as anti-European sentiments were also strong because of austerity measures imposed upon Spain by the European Commission that led to increasing unemployment and worsened the economic crisis situation . Only when European policies were connected to local struggles, the Right2Water campaign started to gain support. In several cities and regions in Spain this local struggle was present in 2012. In this section we explore five cases in which a local social movement was already struggling against privatization or for a more democratic public water supply. Four cases are about cities and one case is about a region . Madrid shows a fight against privatization; the other three cities show successful cases of remunicipalization and a failed struggle for remunicipalization . The case of Andalusia highlights how a social movement for public water was built and changed discourse and policy on water management. --- . . The case of Madrid The Platform against the privatization of the local water company "Canal de Isabel II" was set up in 2010 to oppose the planned privatization of the company in Madrid. It was an urban movement in which social organizations, unions, environmental groups, political parties, and citizens belonging to 15-M participated. The incorporation of the 15-M to the PCPCYII made it possible for the water movement to bond with other social groups-the platforms for the defense of public health and education-that also defended "a right to the city, " imagining the city as a place to live, contrary to, for example, real estate agencies and banks that see the city as a place to exploit . It succeeded in 2012 to organize a popular consultation in which 165,000 people, an overwhelming majority, in Madrid voted against the privatization of the water company and in favor of a 100% public water supply . This was the first time that a blue wave "Marea Azul" had gone over the city . These blue waves came back in manifestations in 2012 and 2013 for the recognition of the HRTWS in EU law, when the Spanish movements joined the European Right2Water movement. As the struggle unfolded, the movement switched from the earlier emphasis on opposition toward a more prepositive stance, seeking not just to stop privatization but also to make the public utility "100 percent public, " with participatory management involving citizens, making its goals more consistent with the principles of the human right to water. Although formally the government froze the privatization plans in June 2015, because of the heavy defeat experienced in the municipal elections in May that year, the Platform argues that the mechanisms that would allow the privatization to go forward are still in place . From the PCPCYII the Red Agua Publica grew that made the Right2Water campaign collaborate with the creation of the Spanish #Iniciativa2015 and the Social Pact for Public Water , in close cooperation with AEOPAS and the trade unions. --- . . The case of Valladolid In 2015, municipal elections brought in a new government and the opportunity to terminate the private water management contract with Agbar in Valladolid. An important incentive was the lack of investment in the water network under private ownership, although the company had consistently reaped high profits. Another motive was the relative price increase for citizens that had been 78% over 18 years under private concession, whereas the city council foresaw that for the next 15 years a relative increase of 18% would be sufficient to cover all costs of supply if the company would be brought back under public management . The proposal to re-municipalize the water company was ahead of the elections made by the RAP and strongly supported by AEOPAS and FNCA that had organized local mobilization ahead of the election and made the remunicipalization of water management an important electoral issue . The new city council organized three rounds of public consultations to ensure participation of citizens, recalling the Right2Water initiative, that in turn increased citizens' support for remunicipalization . An absolute majority in the City Council approved the remunicipalization of water services in December 2016. This was a milestone in remunicipalizations as it was the first big city in Spain that brought water service back in public hands. According to the Valladolid City Council direct management would provide the highest profitability with the lowest tariff increase and include all necessary investments . After 2 years in operation, AquaVall, the new public water company had already saved e13.3 million, equivalent to 4% of the city's total budget in 2018 . So far, the new public company has had a total turnover of e26.4 million. Much of the profits are now used to maintain and improve the city's sanitation and distribution networks . Tariffs have decreased to become among the 10 lowest in Spain . In a period of 5 years the public company invested 46 million euros in improving its services, against a 20 million euros investment in the 20 years before by the private company . The company has taken a place in society by putting emphasis on promoting responsible water use, promoting healthy habits, a commitment to respecting the environment and cooperation with cultural and sporting events and various social causes. One of the success factors of the remunicipalization of water services in Valladolid has been the public financing model. Water and sanitation services are entirely financed through tariffs and income remains earmarked for water services. This public financing model, that is comparable to the system in the Netherlands , has made Aquavall independent from banks and led to a sufficiently large cash flow to ensure economic, social, and environmental sustainability . --- . . The case of Terrassa In 2018 after years of debate and social struggle the municipality took back control over the water company when the concession contract with Agbar terminated. The water re-municipalization process has been the result of a long and intense process that has had citizens themselves as the main driving force. Laying a first stone with the signatories of the "Pact for the public management of water in Terrassa, " with the complicity of 8,000 citizen signatures. Principles that must govern the management of the city's public service are new water policy and culture in the city, improvement of efficiency and the effectiveness of the service, the improvement of water quality, the fight against climate change and the forecasting of policies for adaptation and protection of aquatic ecosystems. "We have to display everything that has to do with the human right to water and a fair price, in relation to access to water, thinking of a tariff system that integrates the concept of social justice and human rights" . https://www.aiguaesvida.org/comunicado-el-futuro-del-aguapublica-en-terrassa-empieza-hoy-con-el-esfuerzo-y-participacionde-todas/ The campaign "Aigua es vida" , a coalition of community groups, trade unions, solidarity groups, and environmentalists in Catalonia has focused on spreading information, building links with ecological movements, and pushing public debate on the issue. They linked with diverse struggles, looking at the social pact for water, such as in connection with campaigns around housing, austerity or democracy. A Water Observatory was set up as means for citizen participation and to increase transparency and accountability of the new water company toward the inhabitants of Terrassa. The new democratic public model of water management, governed democratically and according to a public ethos, such as in Terrassa, is symbolic and may offer important lessons for other struggles . --- . . The case of Barcelona In Barcelona, Agbar established a powerful, historically-grown monopoly over the urban water service provision, extending beyond the city's borders by acquisitions of competitors and smaller concessions . In 2010, a local court questioned the regularity of the contract, forcing Agbar and the government body of the Barcelona Metropolitan Area to form a PPP model and to establish the mixed-capital entity Aigües de Barcelona Empresa Metropolitana de Gestió del Cicle Integral de l'Aigua S.A . As of 2012, ABEM consists of 85% private capital where a majority of 70% is held by Agbar, another 15% belongs to the holding group Criteria and 15% to the AMB. ABEM and its subsidiaries manage and control up to 93% of the urban water supply in 23 municipalities of the AMB and thus nine out of 10 customers depend on the water supply by the company group . Activist groups declared this an abuse of power by the company in collusion with political actors to save the private management involvement and thus their profit revenues under the guise of more public control. In 2016 the Catalan high court declared that ABEM's concession acquisition without a public tender was illegal. This was a victory for the "Aigua es Vida" platform, that argued that only remunicipalization of services into a public and democratic management model could assure necessary investments, measures to improve water quality, protection of water ecosystems and simultaneously create a just and transparent tariff system. A public consultation about a referendum for remunicipalization was organized in 2018 with the help of Barcelona en Comú , the progressive municipal platform forthcoming from 15-M and other social movements, that has been governing the city since 2015. Activists collected enough signatures, but the referendum was blocked by the opposition of other political parties . Subsequently, the MAPiD was formed to further coordinate protests and alternatives to private management within the AMB. Private actors have often influenced policymaking and public debate about water services through their economic and political power. Tough legal action against any remunicipalization effort also played an important role in Agbar's strategy, but the company had to comply with the Catalan law against shut-off of water supply and accept tariff setting by the AMB. Through their appeal, in 2019, the Spanish Supreme Court annulled the Catalan high court judgment of 2016. Given this setback for social movements, the only remaining opportunity to re-municipalize water services by legal means would be to dissolve the contract at a high cost through a bailout. Currently, there is no sufficient political will to enforce such procedures and the administrative structure as well as the heterogeneous political composition of the AMB further impede what would necessarily be a "remetropolitization." However, movements' claims in the context of energy poverty and service problems during the COVID-19 pandemic increased public awareness about water rights as well as about the structural problems caused by the private management model. --- . . The "Mesa Social del Agua Andaluz" As a result of cooperation between the RAP-Andaluz, la Marea Azul del Sur, and the Red Andaluz de Nueva Cultura del Agua , the social organizations, trade unions, mayors and public water operators agreed upon the above mentioned Declaración de Cadiz, 2017. It has become a point of reference for the Spanish water justice movement . The social organizations, including also environmental organizations WWF, SEO-Birdlife, Greenpeace, and the agricultural organizations UPA and COAG, have constituted the Andalusian Roundtable on Water , with a strong and active commitment to defend and promote the HRTWS in Andalusia. In a ten-pointplan prepared by the MSAA, an Andalusian Social Pact for Water is proposed that promotes a model of active, fair and diverse citizen participation as the central axis of Integrated Water Cycle policies. It also puts the human right to water and a 100% public management of water services central to achieve socially, ecologically and economically sustainable water management to combat the impacts of the climate crisis and to face the challenges of an environmentally and socially just hydrological transition . With these proposals the Andalusian coalition goes a step further than implementing the human right to water as it addresses explicitly the needed changes in intensive agriculture and large-scale irrigation. In 2018 the Andalusian government was in a process of revision of the water supply regulation for which a group of social organizations proposed the incorporation of the HRTWS into the water regulation with special attention to vulnerable people and with supply of a vital minimum amount of water for free to all citizens. This proposal was not dealt with before the elections that year and was rejected when a new Andalusian government came into power after the elections . Currently, this struggle lingers on. . Who wins, who loses? Water struggles on the ground and in the media Right2Water landed on fertile ground in Spain as it could engage with the already present anti-privatization sentiments in the country and build on previous social movements focused on water in Spain . The Andalusian Parliament unanimously supported a motion promoted by the Spanish Association of Public Water and Sanitation Operators and by the New Water Culture Foundation for the recognition of the HRTWS and to stop liberalization and commercialization of water services . This can be considered a victory for social movements, even when the Andalusian government did not incorporate the motion in legislation. Both the Right2Water campaign and the #Initiativa2015/Social Pact for Public Water were successful because they drew together support from many social organizations and labor unions that struggled against the encroachment of privatization impacting everyday life of people. Unlike in Germany, where the European Right2Water movement achieved a lot of media attention thanks to contingency of a well-known comedian promoting the citizens' Initiative on TV , the water movement in Spain had to conquer the attention of media. They did so successfully in the wake of the 15-M movement as shown in Figure 1. Broad societal pressure appeared to be the strongest force to make governments and water utilities acknowledge water as a common good and a human right . Right2Water could capitalize on growing dissatisfaction with water policy, concerns about access to quality water and worry over increasing water privatization. The campaign strengthened the Spanish networks that campaigned against urban water services privatization processes and gave an impulse to remunicipalization processes with regards to water services in several Spanish municipalities. More than 50 municipalities and provincial councils approved in their plenary sessions that water --- FIGURE The number of times that issues "privatization," "remunicipalization," or "New Water Culture" were mentioned in Spanish newspaper articles, in relation to "water supply," and the last two cumulative with water movement organizations in the newspapers over the past years. Source: own research. is regarded as a human right and that in their political and territorial geographies it must and will remain outside liberalization and commercialization . The increasing number of re-municipalizations in Spain was an achievement of the Spanish movements. They also proved to be able to combine national pressure on government with local actions to convince and persuade municipalities and citizens that a "New Water Culture" integrating social and ecological aspects of water management is urgent. As we have scrutinized through various regional and supraregional cases, the Spanish movement was built on, among others, the three pillars of environmental justice, demanding that its voice was heard and its demands for a socially just water service provision recognized. The demand for public and democratic water services resonated widely in Spain . The pressure coming from social movements and from European governing bodies in response to Right2Water were a catalyst for local policy change in Spain. The focus in the debate on water services in the past decade shifted from privatization to re-municipalization. Remunicipalization was an unknown issue until 2013. Claims for a New Water Culture, the human right to water and remunicipalization are all part of the same struggle for water justice, reason why we can witness a cumulative number of mentions in the graph. In 2014 the number of cumulative "prowater justice" issues and actors surpass the number of "water privatization" in the newspapers . Moreover, the Spanish movements managed at the same time to turn the tide in local water management as they political-materially achieved a halt to privatization of water and a sharp increase in water remunicipalizations since 2014 . The social pact for public water is the continuation of struggles for water that emerged in times of austerity, built on local initiatives and united a diversity of organizations in Spain in a similar way as right2water had done in Europe . Examples of how the social pact goes beyond the HRTWS and the ECI Right2water demands become manifest in the level of concretization. For instance, as several movement leaders explain, ". . . when it comes to materializing the human right to water, we will demand the implementation of supply management with criteria of social equity in tariff policies. To do this, it is essential to guarantee a minimum provision... between 60 and 100 l per person per day-and the commitment not to cut off the supply in cases of socially justified non-payment" . They also make clear how the movement, more than just claiming the HRTWS, also speaks out unequivocally against privatization: "We consider that water and its associated ecosystems are common goods that cannot be appropriated for the benefit of private interests" . This was also clearly expressed during our meeting with one of the FNCA founding members, Pedro Arrojo who, as was mentioned, is now UN Special Rapporteur on the Human Rights to Water and Sanitation: "There is a direct relationship between the 2 billion marginalized people who lack safe and secure water, and the state of our rivers, polluted, depleted or monopolized by anti-democratic forms of government." Arrojo c.s. explained to support river governance approaches with a transdisciplinary and transcultural orientation toward "commoning" and "public-commons" cogovernance initiatives . The UN Special Rapporteur: "Achieving better river governance without recognizing the critical role of societal movements in reclaiming environmental justice is impossible" . Over the past decade the Spanish water justice movement has developed from a single issue alliance to a broad movement in favor of a more democratic and holistic water governance in Spain. The call for the democratic management of water services, i.e., that citizens are not just users and owners of the resource but also actively involved in the management, has been successful . This is a powerful claim, running counter to neoliberalism whilst empowering communities to take back control as well as ownership . --- . Conclusions By the end of the twentieth century struggles for water justice in Spain started with the struggle against the "top down management" of water, stemming from the time that Franco ruled and water was seen as a commodity. Water activists first used the EU Water Framework Directive and later the human right to water as tools to demand for a New Water Culture in which water is considered as a common good. The FNCA organized citizen water networks; local branches to fight for ecologically sustainable and socially just water management at the local level. Inspiration came from struggles around the world, but most from the Italian water movement and from the ECI "Right2Water." After the European Commission recognized that water is not a commodity but a public good and water was exempted from the concession directive, the European struggle slowed down, but continued in Spain with renewed energy. Struggles for water justice in Spain coincided with austerity policies in response to the Eurozone crisis and led to a confluence of social movements. An extraordinary broad and strong movement took the struggle further from anti-privatization to pro-democratization of water management and away from the mercantilist view to a holistic eco-social view on the governance of water. The Spanish movement now seems to be the voice that brings this eco-social view to the global water policy arena. Human rights principles were already brought into the debate on water management in Spain by the FNCA since 2002. It did not resonate until water supply became an intense battle ground of commodification vs. commoning and the negative effects of the "mercantilizacion" of water became more visible when austerity measures hit hard to the poor and "indignados" rose up against it. The FNCA, RAP, the PSAP, and MSAA are all examples of collective actions to promote a public management of water and working toward water justice. Their struggles are forthcoming from or strengthened by the cooperation of the European Right2Water movement with movements in Spain, with the latter continuing the struggle for water justice in Spain after 2014 in various and sometimes extended compositions led by the RAP. The Spanish water movements achieved recognition of the HRTWS from local and regional governments, awareness around water as a common good and a growing trend of re-municipalizations of water utilities: from 1 case to 14 cases in 8 years between 2006 and 2014 and from 14 to 40 cases in the following 8 years between 2014 and 2022. Remunicipalization is the concretization of water justice struggles in Spain through which the Spanish movement has shown that its powerful and resonating discourse can achieve a shift in local water policy and change the paradigm. The RAP, FNCA, and AEOPAS played an important role in the water struggles in Madrid, Valladolid, Terrassa and Barcelona. They succeeded in prevention of privatization in Madrid and the re-municipalization in Valladolid and Terrassa, but did not achieve re-municipalization of water services in Barcelona. As we have explained, the cases of Madrid, Valladolid and Terrassa are three examples of how social movements have led to the integration of societal values in water services, each one combining and balancing the 3-fold claim for recognition, redistribution and representation differently. In Barcelona corporate and financial pressure on the municipality was too big to give in to the wish of the population. However, the Catalan law against shut-offs and tariff setting by the AMB can be considered as achievements of the water movement . The MAPiD continue to mobilize for remunicipalization and plan to create a citizen parliament and a water service observatory to put pressure on private actors and on politicians. Although remunicipalization resonates with many municipalities, the financial and legal barriers to reverse water management from private to public are seen as too high by big cities where corporations control water services. Debates about more democratic water governance are continuing, especially in face of pressures on the water supply through climate change and economic crises. This bottom-up voice is proliferating, as recently was shown in -among multiple others-a side event at the UN Water conference 2023 organized by the Spanish and European water movements . The claim of the human right to water in Spain goes beyond an anthropocentric "HRTWS." It constitutes today in Spain the banner of a movement that revolves around the concept of water as a common good and aims to build a model of participatory and transparent public water management. The Andalusian roundtable on water goes a step further by including ecological aspects and the agricultural sector in the struggle for water justice. The fact that the HRTWS and the concept of water as a common resonate with so many people seems to coincide with the pursuit of sovereignty of people over their life. Simple as in a slogan "water is life" or "write water, speak democracy" it shows that people act to regain control over water and their lives. The Right2Water movement spotlighted issues arising from water service privatization and liberalization and gave a European platform to local water struggles. Advocacy at the EU-level has helped to build momentum for water mobilization efforts in Spain and reciprocally local mobilization in Spain has helped Right2Water gain success at EU-level despite of different cultural and political motives and backgrounds at the different levels. Notwithstanding social pressure on government and despite the huge efforts by social movements, implementation of the HRTWS in law has not been achieved; neither in the EU, nor in Spain, nor in Andalusia. The main victory of the Spanish water movement may be that it has overcome the differences between organizations and has built an extraordinary broad and strong movement for water democracy that serves as an example for many around the world. --- 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. --- --- --- 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/frwa.2023. 1200440/full#supplementary-material

In several cities and regions in Spain there has been a fight against privatization of water supply in the past decade. Some cities have decided to re-municipalise water supply and debates about implementing the human right to water and sanitation have been held in many parts of Spain, following the success of the Right Water European Citizens' Initiative. This paper examines how the European "Right Water" movement influenced struggles for access to and control over water in Spain from a political ecology perspective. It explores how "Right Water" fuelled the debate on privatization and remunicipalization of water services and what heritage it has left in Spain. We unfold relationships with and between water movements in Spain-like the Red Agua Publica-and relationships with other networks-like the indignados movement and subsequently how water protests converged with austerity protests. In di erent places these struggles took di erent shapes. By deploying five case studies (Madrid, Valladolid, Terrassa, Barcelona, and Andalucía), we look at how the human right to water and sanitation framework served as a tool for social and water justice movements. Struggles for water justice in Spain are ongoing and we seek to identify the temporarily outcomes of these struggles, and whether power balances in Spain's water services provision have shifted in the past decade.

Introduction Smart cities need smart citizens [14][15] [25]. The discourse on "smart learningecosystems" provides a conceptual framing to recognise the "presence of a high density of high-skilled people in a given area" [13] as an indicator of a city's "smartness". On the whole-city level such performance is influenced by factors such as the quality of education or city governance. Yet, taking a closer look at specific citizen communities reveals peer interactions among citizens as a source of new or improved skills and abilities. Rather than being driven by explicitly formulated learning goals, these peer learning activities are often motivated by citizens' basic, core needs for autonomy, relatedness and mastery [8]. Close attention to citizen communities also emphasises the plurality of citizens' skills and motivations as a key factor of the community's capacity building enabled through diffuse design [18]. In this paper we explore the opportunities Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 of leveraging learning processes happening during civic hackathon events, as catalysts for self-directed learning in citizen communities. Hackathons are fast-paced events where teams of participants "hack together" prototypes of existing or new services, apps or products [2] [22] [29]. Civic hackathons are an increasingly popular kind of hackathon, aimed at improving city services and citizen-government relationships, and promoting the use of government-published open data [17]. Unlike regular hackathons that cater mainly to software developers, civic hackathons welcome a more diverse crowd of participants. For example, Code for America [5] is an organisation running dozens of civic hackathons each year throughout cities in the US, each generating dozens of app and service concepts and prototypes. Applications developed in civic hackathons include innovative services addressing various aspects of people's lives. Examples include mapping health-related resources and services in an urban area, providing community members with legal advice, or finding vacant properties that are suitable for local businesses [23]. While there is clear value in civic hackathons as sources of such applications, in this paper we focus on another, underrepresented advantage of civic hackathons, namely their role as loci for peer learning. From this perspective, citizens participating in civic hackathons not only do so to be provided with solutions to their problems, but also to acquire the capacity to solve civic problems through the use of open data. In the presented work, we seek to answer the question of how to "hack the hackathon format", such that it can better support citizen communities in obtaining the ability to work with open data, and, thereby, how to empower groups of citizens to outsmart the "smart cities"? The premise of the hackathon format is that the participants' focus on the "making" of concrete services or products involves a learning-by-doing [10] approach, where the created solutions function as boundary objects [28], reducing communication overhead, and enabling peer-learning through showing and experiencing ways of doing things. Joint making activities, as identified by Giannakos et al. [12], a) accelerate research across involved disciplines, b) promote rigorous multidimensional and multidisciplinary, methods, experimentation strategies and metrics, c) facilitate easy starts of ambitious projects, and d) enable learning through construction activities. Such forms of learning are commonly practiced not only in brief events such as hackathons, but also over longer periods by maker communities in places such as Hackerspaces, Makerspaces or FabLabs [16]. The premise of these places is not only to allow for the collaborative making of prototypes, but also to accelerate and catalyse the learning processes of involved people. In the contexts of these places, communities of practice [32] emerge where participants share skills with each other and perform routines of collaboration. However, in civic hackathons, participants are diverse in that they don not share professional backgrounds, and they have only limited time to establish a community of practice within the event. In this context, we expect there to be a divide between "professionals" with clearly defined expertise or roles relating to the theme of the civic hackathon, and "laypeople citizens" interested in the theme itself. We hypothesise that engaging laypeople citizens in activities involving "prototyping with open data" can have a two-fold effect. On the one hand, "laypeople citizens" may not have technical coding or design skills, but are "experts in their lived experience" [26] within the local context and challenges. In that way, in the spirit of co-creation [26], "professionals" can learn from "laypeople citizens" about the intricacies of the context they develop solutions for. On the other hand, "laypeople citizens" can learn from "professionals" about the opportunities of working with open data and developing open data driven services, apps and products, empowering them in the context of smart cities [30]. The goal of the presented work has been to understand the general principles governing peer-learning in citizen hackathons, and to explore facilitation mechanisms for increasing the community's capacity to work with open data. To this end, as part of a consortium of the Horizon2020 CAPSSI project Open4Citizens , we have organised a series of civic hackathons in five large European cities. We have defined the civic hackathon as a 2-3 day "pressure-cooker" event actively involving citizens with no prior knowledge of coding or other data skills in joint making of a preliminary prototype of a technological product or service, using one or more forms of open data. By executing and analysing the learning processes of participants in these events, we aimed to understand the potential of prototyping with open data as a social learning activity [1]. Accordingly, in facilitating the process, our focus has been split between understanding a) the discovery of new opportunities of meaningful open data applications in a civic hackathon context and b) identifying ways to support bottom-up, community-driven learning and sharing of data literacy skills, such as data scraping, basic data analysis or understanding opportunities and constraints in embedding a dataset in a service or app. --- Methods Exploring learning processes in civic hackathons is uncharted territory within academia. Scarce literature on learning in regular hackathons, for example [11], takes the perspective of supporting the learning of well-defined programming skills. In civic hackathons, however, the emphasis lies on combining and improving a wide array of skills by working together across disciplines in contexts that may vary greatly across cultures, and between communities involved in the hackathons. For that reason, our studies were set up in an explorative way, leveraging semi-structured observations and interviews as the main method of enquiry, standardised across five civic hackathon cases. The five civic hackathons were organised under the auspices of the O4C project in Barcelona, Copenhagen, Karlstad, Milan and Rotterdam. The organisation and facilitation of the events was performed by local O4C consortium members in collaboration with local stakeholders, including, among others, citizen groups, local governments, non-governmental organisations and universities. The general guidelines for the O4C civic hackathons were predefined by the O4C consortium members and provided to individual pilot organisers in the "hackathon organisation handbook". The handbook specified the shared elements of the hackathon process. It indicated the actors to be involved in the process , the phases of the hackathon process, the tools to use in the various phases, as well as other practicalities in the organisation of the hackathon. The 2-3 day civic hackathon event was preceded by a "pre-hack" phase where, among other activities, the theme for each hackathon was defined, participants were recruited, and relevant datasets were collected. Each hackathon was superseded by a "post-hack" phase , where various forms of support were given to hackathon participants to continue developing solutions initiated in the hackathon and in order to sustain the community established throughout the pre-hack phase and the hackathon event. The organisational setup of the actual hackathon event has been inspired by IDEO's inspiration-ideation-implementation process [3], as summarised in Figure 1. Its steps included open data introduction, sharing inspiration, articulating needs to be addressed, brainstorming, data validation, exploration of data using a dedicated platform and toolkit, scenario design, prototype planning, prototype hacking and pitching final concepts and prototypes in a plenary session. For each of the steps, process facilitation tools were prepared . One or more data exploration and data analysis platforms were accessible to support the whole process. At the same time, the process facilitation was prescribed but with some flexibility, allowing individual hackathon organisers to adjust it to local specificities and to the interests and prior skills of participants. Throughout each of the O4C civic hackathon processes, data on these processes was collected in a standardised way across all events and recorded in a variety of forms. Methods for collecting process data included surveys of participants, carried out before and after each hackathon. The work process of all teams was observed and noted, as well as photographed and/or filmed throughout the event by designated researchers. The aim was to capture group dynamics, the process of collaborative solution development, and the mood of the participants, including 'ah-ha' moments where groups' progress towards the development of a concrete solution surged forward. Participants were interviewed during the event to further explore their motivations, and the skills and experiences brought to the event. In addition, where relevant, participants' interest in and knowledge about open data was explored further through additional interviews. The process data was collected in order to understand hackathon participants' learning experiences in light of their backgrounds and skills, motivations for participating, their individual experience of the hackathon process and their perceived gains from the event. The hackathon process data gathering was guided by the facilitation support scheme shown in Figure 1, whose standardised format was specifically tailored to the O4C hackathons and to the hackathon activities through which the participants were steered by event facilitators. Captured data was organised according to process stages, in templates standardised across all the O4C hackathon cases. Hackathon results including designs and prototypes were documented visually, catalogued and shared with the participants. Subsequently, collected data was transcribed where necessary, and analysed by teams of consortium members local to each hackathon, including coding and clustering. By distilling and capturing resulting insights in the shared format, pilot teams were able to capture and compare participants' learning processes in hackathons. The individual experiences of participants and facilitators working to create open-datarelated solutions were captured and translated using significant quotes and visuals. This standardisation of data capturing and analysis across all five hackathon cases enabled comparison of performed processes. Such comparison led to insights into opportunities and challenges in tailoring a structured hackathon process and toolkit to contexts with different thematic challenges, varying skills of facilitators and stakeholders, a range of participant motivations and variable quality of available inputs, including open data. Further, the collected data has also allowed for an assessment of how the given factors influenced the individual and collective learning processes of hackathon participants. The described data has allowed us to further outline a number of trends related to social learning in civic hackathons. --- The five O4C civic hackathons --- --- Process The process of the hackathons involved a mixture of activities. In all cases the facilitation was carried out flexibly. The facilitation scheme in Figure 1, for instance, was considered as an orientation map, rather than a prescriptive sequence of phases, and the facilitators were free to use other tools that might be more familiar to them. The flexible facilitation resulted from pragmatic decisions by the pilot teams about how best to support a good experience for participants as well as learning about the process and the development of solutions to challenges posed within the time available. Table 3 summarizes the resulting key differences between hackathons. As a result of facilitation adjustments, the hackathon processes in Barcelona, Copenhagen and Karlstad resembled fast design sprints, with concepts being designed on paper during the implementation phase. The Milan hackathon process was the closest to a typical coding hackathon, while the Rotterdam hackathon involved mainly conceptual and open-ended designs, albeit converging in concrete sketched design concepts at the end. In most cases, use of open data was limited to finding inspiration in descriptions of other open-data-driven solutions and in descriptions of the available and possibly useful open data sets. In some cases, inspiration was found in performing ad-hoc data visualisations and analysis, or simply browsing through available datasets and discovering intriguing facts. The post-hack phase was in all cases considered as the phase in which the concept could be implemented, using the most appropriate datasets. Among participating teams, two distinct types of activities were observed in all hackathons. Many teams engaged in extended "discussing" of topics related to the hackathon theme and their design response to it, where views and ideas were shared and the team's direction determined. This type of activity typically took place with all participants assembled around a table, whiteboard or flip chart, as shown in Figure 2. Plenary presentations of teams' work, performed in some cases during, and in all cases at the end of the event, were a special form of "discussing". During plenary presentations teams had an opportunity to exchange views and ideas, as well as share technical approaches or data findings, and in this way influence each other's work. Team members also performed "focused" activities. These activities typically entailed some form of making, either performed as drawing, writing or coding, but always directed towards producing a form of concrete representation or prototype of a designed solution or its part. During such activities, participants focused on specific tasks, such as drawing a screen layout, a diagram or a scenario, browsing through data, or writing a snippet of code. During some moments of "focusing", every team member performed an individual activity. At other times, one person would work on a task and one or more of the other participants would look over that person's shoulder, trying to understand and follow that person's activity, making comments and asking questions, and in this way familiarise themselves with the skills and know-how involved. The "discussing" and "focusing" activities were performed iteratively in hackathons. The organisation of the hackathon influenced some of these iterations by limiting their time, although the number of iterations between forced presentation moments or breaks varied by team and hackathon. The time duration ratios between "discussing" and "focusing" varied significantly between hackathons. Although no exact account of the duration of these activities was recorded per team, a general comparison is possible based on our observations. For example, in the Rotterdam hackathon mainly "discussing" took place with only short intervals of "focusing". In Milan, "focusing" was the dominant activity in the middle of the hackathon, with "discussing" mainly taking place at the outset of the event. In the Copenhagen hackathon, the "focusing" elements were integrated around plenary presentations of tools by the facilitators and organisers or of work in progress by the teams of participants. Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 --- Outputs Assessment of artifacts generated by hackathon participants revealed the intricacies of the performed processes. On the one hand, the teams generated various forms of written notes throughout their process, typically using post-its, at times without structure, and at times using some of the prescribed facilitation formats. This documentation, mainly intended for internal communication within teams, showed the dynamic nature of the processes and the mix of ideas from various domains intertwining with each other. On the other hand, the teams also generated various forms of service design representations and prototypes, which provided a more legible, albeit indirect, account of their processes. Design representations and prototypes varied significantly in their forms across hackathons, which can be attributed to differences in team compositions and facilitation. However, we could generally discern two types of prototypes and design representations, following their correspondence with earlier discussed "discussing" and "focusing'' activities of participants. The first type of representations and prototypes was typically made together by entire teams during the "discussing" activity, and can be characterised by their low fidelity. These representations and prototypes were made quickly using ad-hoc techniques. Figure 3 shows one example, a "paper prototype" [27], developed during the Rotterdam hackathon. Such prototypes are make-believe sketches of an app interface made on post-its attached to a mobile-phone cardboard mockup. Paper prototypes are almost instantaneous to make and can be used to crudely explore app functions and communicate them within and outside of the team. Figure 4 shows another example of the same type, where a user scenario was explored and enacted using lego blocks during the Copenhagen hackathon. Similarly to a paper prototype, such a representation of a service allowed the team to both explore and rapidly validate the intricacies of the designed service as used in various contexts. The second type of representations were high-fidelity drawings, models and prototypes, made during "focusing" activities, typically not involving extended cooperation in the process of their making. Such representations and prototypes had different roles depending on the phase of the hackathon. Initially, they could only address a narrow aspect of a possible solution, for example, by making a visualisation of a subset of available data using a spreadsheet program. Later in the process they were combined into high-fidelity prototypes and detailed representations of designed services, providing a believable outlook of how a designed service might function and feel. Figure 5 shows an example of a website mockup generated during the Barcelona hackathon that fits in this category. Here, various parts created by different participants were integrated in a way that could provide an experience of using the designed service. Such prototypes were used during the final round of hackathon presentations in order to show how attainable the concepts were and that their further development would be useful. In all cases, participants used the final prototypes and design representations as tools to tell the story of the specific city problems on which they had worked during the hackathon and to explain their visions of how these problems could be addressed. However, the way in which the different types of prototypes were integrated in teams' activities clearly differed; they either dynamically elicited discussions and explorations of a topic, or consolidated teams' conclusions, individual work and individual expertise to represent and communicate a single design solution. --- Results The twofold goal initially formulated for the O4C civic hackathons was to support citizens in a) discovering new opportunities of meaningful open data applications and b) bottom-up, community-driven learning and sharing of data literacy skills. The two goals have proven to be tightly interlinked. The open data applications, whether instantiated as low-fidelity sketched ideas, or high-fidelity working app prototypes, functioned as boundary objects in the hackathons. They supported the consolidation of participants' skills and ideas, confronting these with the world, as well as communicating and reflecting on them. Consequently, the open data applications have also supported the learning and sharing of data literacy skills. As such, they helped reveal different competences involved in the hackathons and participants' different expectations. The learning processes at hand have also proven to be more complex and difficult to measure than initially expected, as they differed considerably between participants and cases. This complexity of measurement was also reflected in the different criteria proposed for measuring such goals. The O4C team had clear and quantitative criteria regarding the number of viable concepts developed in the hackathons and the number of startups or entrepreneurial initiatives triggered by each hackathon. On the other hand, the learning objectives were only qualitatively defined, Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 on the basis of the pre-and post-hack interviews that provided information about initial skills and expectations and about new skills and knowledge acquired during the hackathon. Such qualitative evaluation is also influenced by the different cultural backgrounds and on the diversity of skills among participants in the same team. --- The learning outcomes The heterogeneous mix of knowledge brought by the participants to the hackathons and the varying conditions of accessibility of open datasets has made it difficult to perform a quantitative evaluation of the learning outcomes of this process. The participants' learning expectations were in fact related to their motivations for participating. In addition, different types of participants had different motivations, as summarised in Figure 6. Interviews with the participants were not carried out assuming that these motivations corresponded to learning objectives. Instead, the interviews aimed to define the effectiveness of the hackathon as a tool to address the participants' expectations regarding open data. Aspects considered in the interviews were not only related to the use or disclosure of the data, but also to the systemic effects that the hackathon can trigger in the process of innovation. The interviews revealed two ways in which participants experienced learning during the hackathon process. First, the majority of participants confirmed that in some ways they have acquired new knowledge regarding the expected learning domain. For example, a participant interested in understanding the potential of open data, learned about the existence of several relevant open data sources, and their application to his area of interest. Second, most participants indicated that their most valued learning experiences lay in learning unexpected things. For example, one of the civic activists indicated that she came to the hackathon to find resources for her project in the city budget, and instead she learned how to use open data to build a business case for the community project independent of city funding. Yet, above all, the majority of participants have indicated that the main benefits they saw in joining the hackathon was in the social connections they made, developing mutual understanding of each other's skills and goals, and an expectation that the team, working together, will be able to continue their joint work after the hackathon. In this way, the increased capacity of the community to work with open data is larger than the sum of individual learnings of the participating community members, and is evidenced by open data applications generated in the hackathons. --- Open data applications The hackathon activities across the various pilots generated a number of concepts, with different levels of maturity, ranging from sketched ideas to working prototypes. Table 4 synthesises the results of the hackathons. The number of start-ups created after the hackathon is an additional metric, indicating where the relevance and maturity of developed applications was high. Within the distinction of 'level of development' between 'concepts', 'prototypes' and 'mock-ups' reported in Table 4, there were some further disparities. For instance, some hackathons only developed paper prototypes, whereas others developed high fidelity prototypes. Also mock-ups had varying levels of detail. Some of them were used as proof of concept for presentations in the post-hack phase, whereas others were the starting point for the construction of apps that were further developed up to the commercial stage. The further qualitative analysis of open data applications not developed as part of the O4C project has revealed the potential of these applications to increase the capacity of involved citizen groups to leverage open data to learn and meaningfully transform their city outside of the time-frame of hackathon events. For example, a prototype app for mapping valuable trees across the city promised to further spread the knowledge and awareness about the importance of natural ecosystems in cities and gives citizens a tool to actively protect that ecosystem by taking collective action to protect trees when needed, and based on facts rather than emotions. The diversity and richness of the opportunities coming from such applications prompts a more open discussion of the opportunities of civic hackathons, not just as one-time learning and capacity building events, but also as parts of learning ecosystems at a larger timescale. --- Discussion The results of the presented civic hackathon case studies have enabled us to derive considerations about the nature of the learning processes involved. We have structured these considerations based on the four perspectives of a) individual learning, b) community capacity building, c) learning through prototyping, and d) civic hackathon facilitation. --- Individual learning perspective The individual learning perspective was evaluated through interviews conducted before and after each hackathon, which gave the O4C project team a chance to compare initial expectations, skills of participants and the actual experience of participants in the hackathon. The five hackathons took different approaches towards recruiting participants and responding to the hackathon themes. As a result, the distribution of participant types has varied considerably across the hackathons. This has led to different hackathon profiles and differences in the characters and roles of created prototypes. For example, the Milan hackathon had the highest turnout of coders, and prototypes included snippets of working code and detailed app mockups. On the other Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 hand, the Rotterdam hackathon mainly involved citizen activists and municipality employees, no code was developed and prototypes were limited to paper app mockups. In the first case, participants were able to improve coding skills, while in the latter the focus was mainly on developing conceptualisation skills. The interviews with participants and post-hackathon surveys have revealed a different perspective than analysis of participant activities and prototypes alone would indicate. The first trend was for participants to question the non-nuanced classification of their own expertise. For example, several municipality employees emphasised that they are also regular citizens, one indicated that coding is a hobby of his. To give another example, the differences between designers and coders were often difficult to clearly articulate, as many designers had some coding skills, and many coders had experience in app design. Most participants also demonstrated design and creative thinking capabilities, even if they indicated they did not have any. The other trend was for several participants to indicate that their perception of the social or institutional group they represent is that it is composed of individuals of very diverse skills, motivations, characters, and underpinning values. For example, a municipality worker in the Rotterdam hackathon said "Citizens see municipality as one thing, while there are many very different people working there. At the same time, municipality workers see citizens as one mass of people, without understanding the differences between individuals." The third trend was that there was rarely a single, clearly articulated motivation for each participant to participate in a hackathon. Participants typically stated only very general motivations such as a) improving one's skills by applying them to a practical case, b) learning new skills from others, or c) only indicating general curiosity in the hackathon's theme and/or open data. The above trends allow us to conclude that strong profiling of participants is to a large extent futile in the context of a civic hackathon. Participants generally shared a lack of explicitly articulated motivations to participate in the hackathon, while they exhibited multiple combinations of prior skills and abilities during the hackathon itself, and the roles that individuals eventually assumed in their teams were dynamic and organic throughout the processes. The above insights have inspired us to use the T-profile of a designer popularised, by Tim Brown [3], to summarise these insights for further reflection. In the T-profile, the vertical leg of the letter T symbolizes one person's core expertise and the horizontal bar represents their general overview of other knowledge. Figure 7 uses the T-profile metaphor to indicate three learning opportunities that we have observed among hackathon participants, independently of the specifics of their core expertise. Red arrows indicate enhancing existing core expertise. Green arrows indicate developing new expertises during a hackathon. The blue arrows indicate gaining a broad view of opportunities coming from the skills of others, i.e., getting "a taste" of diverse skills and knowledge involved in the design and development of an app or service. The advantage of this model is that it does not specify what the core competence of the hackathon participant is, only that there is one, while differentiating between learning that may occur within that competence, regarding other expert competences, or on a very broad level. --- Community capacity building The formulation and evaluation of the community capacity building perspective is based on pre-hack and post-hack interviews. The hackathon pilots were organised with a premise of supporting citizens to learn data-related skills, while developing meaningful solutions to problems found in their city. However, in the post-hackathon feedback, participants indicated that learning about other participants' views, improving their ability to work in a diverse team, and gaining new perspectives on the city's problems were the learning outcomes that they valued the most. Furthermore, we have observed that while jointly working on solutions, participants' views of the problem they were facing, and ways in which they were articulating their city-related values were also transforming. For example, in the Rotterdam pilot, one participant changed the attitude towards the city from "it's their job [as a municipality] to support our park financially" to "we should come up with a business plan for our park that we could pitch to the city [officials]". Consequently, understanding and taking advantage of involved social learning [1] happening between participants, and resulting alignment of their values and perspectives, called for a deeper investigation in an attempt to explain its root causes and mechanisms. Another aspect concerning social learning was the growing perspective on open data as a new resource for public innovation or a new commons [21]. The elaboration of this perspective was purely conceptual in Rotterdam, where the hackathon produced concept prototypes, whereas it brought about a concrete application in Milan, where a specialised prototype was further developed and integrated in the existing web portal that informs citizens about the construction of the new metro line. In their discussion on constructivist and sociocultural perspectives on learning, Packer and Goiocoechea [24] conclude that "acquiring knowledge and expertise always entails participation in relationships and community and the transformation of the person and of the social world" where "the person is constructed in a social context formed through practical activity ". We can expand this general perspective on learning in a community by considering the notion of shared cognitive models. As Cooke et al. [6] argue, shared cognitive models develop through interactions between team members, and enable teams to address problems in synchrony. We will further refer to such synchrony as "collaborative learning alignment", as it can be applied to situations where not only teams, but also other, more fluid forms of collaboration can Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 take place. The above view of the intrinsic nature of "social learning" matches our need to consider hackathon participants' identities not only as being unique, but also as actively evolving, aligning with each other, and stimulating each other's growth throughout the execution of the making activities during hackathons. This kind of learning contributes not only to individual abilities or skills, but especially to the capacity of the community as a whole to perform certain actions, connecting to the notion of "community capacity building" [7], especially popular in the discourse on governance. Collaborative learning alignment in the social learning context expands on the considerations that the previous section accounted for regarding individual learning. Figure 8 indicates how we can use the proposed participant knowledge profile to address how participant profiles overlap and how the values and perspectives of participants align, while specialised learning occurs. Notably, the diagram also suggests that the specialised learning of one participant can trigger other participants to follow in obtaining knowledge in the same or related areas of expertise. --- Prototyping perspective Having gained a better understanding of individual learning and community capacity building happening in civic hackathons, the third perspective that we take focuses on the hacking of a prototype as an activity that differentiates a civic hackathon from other peer learning activities. What appeared significant is the role played by prototypes and other forms of intermediate design representations. We have observed that talking about the application of individual skills to an app, service or product gave participants a way to both align their general views on addressed problems and to position their individual skills in the context of what would need to be done to bring such an app, service or product into being. This supports the role that Carlile [4] gives to prototypes and design representations as boundary objects "representing, learning about, and transforming knowledge to resolve the consequences that exist at a given [community] boundary". Articulating this role of prototypes in the learning process during civic hackathons has given us ground to extend our enquiry into the way in which prototyping tools introduced to participants support or inhibit the prototype's boundary object role. For example, a code snippet written by one of the participants extracting and processing data from one of the data repositories was a specialised, and functionally useful, ingredient of a potential app. However, in and of itself, it had a very limited role in communicating to other hackathon participants about the merits of the problem being addressed. Conversely, for example, a "paper prototype" jointly developed by participants could be easily understood by all of them on the conceptual level and supported discussion and team alignment. As illustrated in Figure 9, both such types of prototypes can be related to the different types of learning that we have previously identified. "Conceptual prototypes" overlap with team alignment, while "specialised prototypes" support specialised skill learning. The proposed distinction of prototypes raises numerous new questions regarding process organisation and the facilitation of civic hackathons. Different forms of process facilitation, and various selections of prototyping and design tools provided to participants in the civic hackathon can effectively guide the learning process towards either alignment of participants or specialisation in given knowledge areas. --- Facilitation perspective The organisation of the O4C civic hackathon pilots was inspired by the inspire-ideateimplement approach of IDEO [3]. Considering our findings regarding the different forms of learning and different roles that prototypes played in this process, we can attempt to gain additional understanding of how participants' learning process is affected by the organisational context of civic hackathons, and how this context can be expanded beyond the physical and temporal scale of the hackathon event. In this investigation, Von Hippel's concept of "sticky information" [31] sheds particular light on the role of context and locality in learning-by-making communities. It also draws attention to the notion of a community of practice [10][32] that hackathon participants are stimulated to form, within, and beyond the hackathon event. Those perspectives provide us with a rich framing for the capacity building process occurring in the civic hackathons, where individuals' learning processes align in respect to one another, around the prototype and within the nested loci of the hackathon event and the city in which the hackathon team operates. The above theoretical framing brings two facilitation problems to the foreground, which open further areas of enquiry for the next stages of our research. First is the question of the role of data prototyping tools. These tools need to accommodate both the specialised and conceptual role and, in this, may be used throughout the stages of the entire process, not exclusively in the final implementation stage. Second is the question of sustaining the capacity building locus beyond what the hackathon temporarily provides. The civic hackathon processes we have facilitated had an apparently linear structure. They started with pre-hackathon preparations, proceeded through inspire, ideate, implement stages during a hackathon and were followed with post-hackathon activities aiming to follow up on designed and prototyped ideas. Taking the learning perspective on understanding this process gives us a framing that allows us to conceptually expand our understanding of the learning process beyond the spatio-temporal scale of a single event. The above consideration can be translated to a three-step framework for facilitation of civic hackathons as a means to support smart learning ecosystems in smart cities, illustrated in Figure 10. The three steps correspond to the three discussed perspectives for understanding learning in civic hackathons. Individual learning, community capacity building and generation of prototypes are, consequently, three different aspects of civic hackathons that have different key performance and key behaviour indicators, yet are all essential ingredients of a civic hackathon event. Performed together, they constitute a positive learning loop, where individuals' skills contribute to the capacity of the community, lead to applications of these skills in a complementary fashion to prototypes, and enable individuals to learn from these prototypes. --- Conclusions Research presented in this paper has explored the opportunities of leveraging civic hackathon events as catalysts for learning in citizen communities. The aim behind the five civic hackathon cases involved in our research has been to improve citizens' capacity to use open data and to enable citizens to participate in the process of developing open data applications. It was our key assumption that "making with data" Interaction Design and Architecture Journal -IxD&A, N.43, 2019-20, pp. 8 -29 by jointly designing and prototyping apps, services or products during a pressurecooker civic hackathon event creates conditions in which citizens develop data literacy and apply it for the common good. Performing and analysing five civic hackathon processes has given us an opportunity to confirm this assumption. It has also expanded our understanding of how such learning processes happen and how they can be supported. In evaluating collected data, we have reached out to theories originating from a variety of disciplines, providing us with a cross-disciplinary perspective on the civic hackathon learning phenomenon, its challenges and opportunities. In our findings we have identified and scrutinised four perspectives of the civic hackathon learning processes, namely, a) individual learning, b) community capacity building, c) learning through prototyping and d) process facilitation. Across these perspectives, a view of citizens as individuals with a unique composition of prior skills and knowledge prevailed. Next to our initial goal of supporting these citizens in obtaining data-related expertise, we recognised the relevance of their different skills and abilities for the development of valuable open data applications. During the civic hackathons, we have observed the alignment of involved citizens' values and motivations, including learning about each other's viewpoints and expertises, and developing shared visions, as an essential component of their learning process. In fact, without a shared understanding among hackathon team members of the merit of other participants' viewpoints, skills and knowledge, successful collaborative work on applying open data to solving complex social challenges would have been very difficult. In keeping with van Waart et al. [30], we found the value of participatory prototyping in strengthening such shared understanding among the local citizen activist networks. Joint prototyping allowed participants to make their own specific expertise and its usefulness for the job at hand explicit to themselves and to other team members. This encouraged both individual and social learning during civic hackathon events. It also formed bonds and shared goals among participants to continue their joint activities beyond the hackathon event. The collected insights have brought us to a general framework defining individual learning, community capacity, and learning through prototyping as three mutually enforcing learning activities in civic hackathons. This framework promises to serve as a guideline for defining appropriate key performance and behavioral indicators for assessing learning in future civic hackathons. We also hope this framework will be of use to civic hackathon facilitators, by emphasizing the relevance of capacity building, individual learning, and the knowledge-generating role of prototypes.

This paper investigates the opportunities of leveraging a hackathon format to empower citizens by increasing their abilities to use open data to improve their neighbourhoods and communities. The presented discussion is grounded in five civic hackathon case studies organised in five European cities. The research revealed specialised learning and collaborative alignment as two mutually complementary aspects of the involved learning processes, which were achieved with the help of high-fidelity and low-fidelity prototypes, respectively. Consequently, the paper identifies and discusses three main factors required to sustain social learning ecosystems beyond hackathon events, and with the purpose of democratising smart city services. These factors include a) supporting individuals in obtaining specific expert knowledge and skills, b) nurturing dataliterate activist communities of practice made up of citizens with complementary expert skillsets, and c) enabling members of these communities to generate prototypes of open-data services of varying fidelity.

INTRODUCTION Available research suggests potential associations between community characteristics and physical and economic availability of cigarettes to youth. However, much of this research is limited by the number of community characteristics that have been included and the lack of diversity of the communities that have been studied. Moreover, although some studies have modeled characteristics related to tobacco sales to youths , none has modeled a full range of contextual characteristics together with community characteristics. Since illegal tobacco purchase attempts by minors occur in an ecological environment that includes various contextual and community characteristics, this approach can highlight ways to tailor policies to more effectively reduce youth access to cigarettes via commercial sources. The present study investigates the associations between a wide range of contextual and community-level characteristics and retailer compliance with underage tobacco sales laws and with cigarette pack prices in 50 non-contiguous mid-sized California communities. Reducing youth access to tobacco products from commercial sources through implementation and enforcement of policies that target retailers and clerks who sell such products to young people and increasing taxes, and therefore cigarette prices are recommended to control tobacco use by youth [1][2][3][4]. A comprehensive review that evaluated efforts to prevent the sale of tobacco to youth concluded that every intervention that has successfully disrupted the sale of tobacco to minors has been associated with an observed reduction in tobacco use by youth [5]. Also, higher cigarette prices decrease cigarette smoking [6][7][8], particularly among youth, who are more price sensitive than adults [9,10]. Contextual factors associated with youth access to tobacco products from commercial sources include characteristics of youth buyers and retail clerks. Research has shown that girls and older minors are more successful in their purchase attempts [11][12][13][14][15]. Findings related to characteristics of store clerks, including gender and age, are mixed. For example, a study which used federal compliance data from 36 states found that female clerks were more likely to sell tobacco to minors [13]. In contrast, other studies have found that illegal sales are more likely when a clerk is male [12,16] or did not find any relationship between clerk gender and tobacco sales to minors [11,17]. Similarly, while studies by Pearson et al [15] and DiFranza et al [18] found that clerks judged to be less than 18 or 21 years of age were more likely to make an illegal sale, no such relation was reported in other studies [11,17]. A few studies that examined youth and clerk ethnicity have shown that White clerks were more likely to sell cigarette to youth [12,19], and that cigarette sales to minority youth were higher than to White youth [12,20]. Factors related to the retail environment have also been examined. For example, a lower likelihood of illegal sales rate was associated with chain stores compared to independent stores in California [21]. In another study, sales rates were highest in gas stations [13], and a study in a metropolitan county in Washington State found that convenience stores selling gas were more likely to sell tobacco products to minors than restaurants, bars, and tobacco discount stores [15]. Age-of-sale signs, asking about the purchaser's age, and requesting purchaser IDs were also associated with lower rates of underage sales [13,16,17,22]. Less is known about the relationship between youth access to tobacco products from commercial sources and community characteristics. In one study, rural location was associated with increased underage sales [13]. A study by Lipton et al [23] found that areas of higher sales to minors in Los Angeles had lower mean family income, a higher percentage of foreign born residents, and greater population density. In southern California, Klonoff et al [24] found that minors were able to purchase single cigarettes in 71.2% of visits to minority neighborhoods, but could do so only in 34.4% of visits to predominately white neighborhoods. No neighborhood differences in illegal sales of cigarette packs were found [12]. Similarly, in a case study in Miami, Florida, underage tobacco sales were significantly more concentrated in Hispanic-majority neighborhoods [25]. These investigations are illustrative but omit other important community characteristics, such as tobacco policies and adult smoking prevalence, that might also be associated with higher or lower sales rates. Focusing on economic availability, a Minnesota study of pack prices in one metropolitan area observed that the maximum price was 1.7 to 1.8 times higher than the lowest price for the same brand [26]. A positive association between percentage of nonwhite residents and the price of discount and premium cigarettes, as well as the overall mean price, was found. No such relation was found with the price of menthol cigarettes. Prices of premium cigarettes and the overall mean price were lower in neighborhoods with a higher percentage of youth. Going beyond previous research, the present study examines the associations between a wide range of contextual and community-level characteristics and retailer compliance with underage tobacco sales laws and with pack prices in 50 non-contiguous mid-sized California communities. --- METHODS --- Study sample and survey methods This study used data from access surveys conducted by 4 confederate buyers , who were over age 18 but judged to appear younger by an independent panel. Purchase attempts were made at 997 tobacco outlets in 50 midsized California cities. To select 50 non-contiguous California cities, the initial sample frame comprised all 138 California cities with populations between 50,000 and 500,000. We randomly sampled one city and then eliminated all contiguous cities, all cities contiguous to those cities and those that were within a one-mile radius of the selected city. This process was repeated until 50 cities were selected. The resulting sample of 50 cities is a purposive geographic sample intended to maximize validity with regard to the geography and ecology of the state [27]. There were no significant differences between the sampled and the unsampled cities in relation to population size, ethnic diversity, household size, and median household incomes. Twenty randomly selected tobacco outlets in each city were surveyed to collect information on compliance with underage tobacco sales laws and cigarette prices. Because comprehensive address lists for tobacco outlets in California are not available, outlets were randomly sampled from lists created specifically for the current study. Shape files of parcel or zoning areas with recent zoning attributes were obtained from each of the 50 cities. For 5 cities without usable Geographic Information System files, a zoning map was obtained. Zoning code definitions were reviewed to indicate which areas could include tobacco retailers. Map books were made for field study coordinators. For all but the five largest cities , the map books included all retail/ commercial areas within the city. For the five most populous cities, retail/commercial areas were randomly selected until a minimum of 124 outlets were identified. Observations in the 50 cities then documented retailer addresses. Randomly selected tobacco outlets in each city were surveyed by a team of two buyers. At each outlet a single buyer attempted to purchase a pack of Marlboro or Newport cigarettes, which are the most popular cigarette brands among high-school-aged students [28]. Each buyer asked for Marlboro in one outlet and Newport in the next one. If asked about their age they stated that they were over 18 years old and if asked for age identification they indicated they had none. If a sale was refused, the buyers left without attempting to pressure the clerk. After leaving the outlet, the buyer recorded the purchase outcome and outlet data on a standardized form. These protocols have been successfully used in studies of youth access to alcohol [29][30][31] and tobacco [32,33]. Institutional review board approval was obtained prior to study implementation. --- Measures --- Retail-level measures --- Independent variables: The following data, documented by the confederate buyers after leaving the outlets, were used: age requested , clerk gender , approximate age of the clerk , the number of customers in line at the time of the purchase attempt, the presence of warning signs about tobacco sales to minors in the store , and the type of outlet . Buyers' gender and actual age were also included as variables in the analyses. --- Outcome variables: The confederate buyers documented whether the sale was made , whether asked to show an ID , and the price of the pack of cigarettes they purchased or attempted to purchase. Outlets where buyers could not ascertain the price of a pack of cigarettes during the purchase attempt were later telephoned. We excluded three surveys in which the confederate buyer attempted to purchase a pack of cigarettes other than Marlboro or Newport. --- Community-level measures City demographics: Measures of city demographics were obtained from 2010 GeoLytics data [34]. City characteristics included population density , percentage of population under 18 years old, median family income, percentage of population that was African-American, percentage of population that was Hispanic, and percentage of population with a college education. All city-level demographics were standardized. Adult smoking prevalence: Adult smoking prevalence in each city was ascertained from a survey of 8,918 adults over the age of 21 years old conducted in the same 50 cities as a part of another study. Respondents were surveyed through a random digit dial computer-assisted telephone interview and were asked whether they currently smoked cigarettes every day, some days, or not at all. Adult smoking prevalence was computed as the percentage of every-day and some-day smokers in each city. --- Tobacco outlet density: The total number of licensed tobacco retail establishments in each city was obtained from State of California Board of Equalization data-files for September 2011. These data include the total number of licensed tobacco outlets by city and zip code, but not outlet names or addresses. Outlet density in each city was calculated as the number of outlets per 10,000 persons. Local tobacco retailer licensing: Data about localities requiring tobacco retailer licensing were obtained from the California American Lung Association website [35]. Cities were coded as having or not having a local tobacco retailer licensing ordinance. Cigarette tax: Cigarette sales tax information was obtained from State of California Board of Equalization data available online [36]. This tax information represents the local tax in each city at the time the access survey conducted. --- Data analysis Multilevel logistic and linear regression analyses were conducted with HLM version 6.04 software to adjust for clustering of observations within cities [37]. Intraclass correlations were .15, .30 and .08 for whether the sale was made, whether retailer requested ID, and cigarette pack price, respectively. ICC results indicate that the between city variation is less for cigarette price than it is for whether the sale was made and whether retailer requested ID. These intraclass correlation values suggest that observations within the cities were not independent and indicate the value of including cities as a random second-level unit. In predicting retail compliance with underage tobacco sales and clerk request for ID, all retail-level variables were included as well as all community-level variables. In predicting cigarette price only the type of outlet was included at the retail-level together with all community-level variables. Additional analyses were conducted to examine the relationships between cigarette price and contextual and community characteristics by brand . In each model, variables at both levels were entered simultaneously. The models assumed that slopes were fixed and only the retail-level intercept coefficient was allowed to vary across cities. Unit-specific models were used to evaluate results of the logistic regression analyses. --- RESULTS --- Retailer compliance with underage tobacco sales laws Overall rate of retailer noncompliance with underage tobacco sales laws in the 997 selected outlets was 14.3%. Descriptive statistics from the access surveys are in Table 1. A multilevel logistic regression analysis was conducted to examine the relationships among retail and community level characteristics and retailer compliance with underage tobacco sales laws. At the retail level, buyer's actual age, the clerk being male, and asking young buyers about their age were each positively associated with selling a pack of cigarettes . None of the other retail level characteristics were related to non-compliance. At the community level, higher percentage of minors in the population, a greater percentage of residents with at least a college degree, and a greater percentage of African Americans were associated with increased likelihood of non-compliance. An additional multilevel logistic regression analysis was conducted to examine predictors of clerks requesting identification . At the retail level, the presence of age-of-sale signage was positively associated with clerks requesting ID. Buyer's actual age decreased likelihood of clerks requesting ID. At the community level, lower percentage of minors, lower percentage of residents with at least a college degree, lower percentage of African Americans, and local tobacco retailer licensing were associated with retailers asking for ID. [Insert Table 2 here] --- Cigarette price Pack prices ranged from $4.31 to $8.72 and varied greatly across outlets. For the same brand, the maximum price was 1.9 to 2 times higher than the lowest price. Results of a multilevel linear regression analysis to examine the relationships among retail and community level characteristics and cigarette prices are presented in Table 3. Results indicate significant associations between the type of the outlet and prices of a pack of cigarettes. More specifically, supermarkets charged significantly more for a pack of cigarettes than small markets whereas smoke/tobacco shops and drug stores/pharmacies charged less on average. Turning to community level characteristics, higher cigarette prices were associated with higher median household income and a greater percentage of Hispanics in the community. Similar results were found with respect to outlet type in the brand-specific models . Supermarkets charged more for both Marlboro and Newport and smoke/tobacco shops and drug stores/pharmacies charged less. Differences, however, were observed in the community level variables. While higher median household income, lower prevalence of adult smokers in the community, and not having local tobacco retailer licensing policy were associated with higher prices of Marlboro, lower percentage of minors and increased percentage of Hispanics were associated with higher prices of Newport. None of the other community level variables that predicted either general cigarette prices or Marlboro prices were related to prices of Newport. --- DISCUSSION We investigated the relationships of a wide range of retail and community factors with youth access to cigarettes through commercial sources. Results of our investigation suggest that, after controlling for a range of community level characteristics, only a few retail level characteristics are important. Whereas results of previous studies indicated associations between sales of cigarettes to minors and various characteristics of youth buyers, store clerks, the retail environment and factors related to the purchase attempt [11-17, 19, 22], we only found buyer's actual age, being a male clerk, and clerks asking young buyers about their age related to successful cigarette purchases by underage appearing buyers. This supports our argument about the importance of studying a full range of both contextual and community characteristics in order to highlight ways to tailor policies to reduce youth access to cigarettes via commercial sources more effectively. Our finding that asking young buyers about their age was positively associated with successful purchases is contrary to findings from Arday et al. [22], who reported the opposite. Our finding suggests that retailers who do ask about the age may sell cigarettes to youth who state they are over 18 years old without requesting proof of age. Interventions with retailers should emphasize the importance of requiring clerks to request proof of age regardless of if the person states that he or she is over 18 years old. Procedures to verify that clerks consistently check IDs may be critically important if such policies are to be effective [38]. The importance of store policies is further highlighted by our results showing that visible age-of-sale signs were associated with an increased likelihood that the clerk requested proof of age. Similar findings have been reported regarding the presence of signage sales of alcohol to minors [30]. It is possible that the presence of such signs reminds clerks to be vigilant or that posting signs is simply one indicator that a retailer has implemented consistent and effective policies regarding checking ID. Our findings also suggest that youth access to cigarettes from commercial sources is easier in some communities than in others. Youth in communities with higher educational levels, greater percentage of minors, or greater percentage of African Americans may have easier access to cigarettes from retail stores. It is possible that retailers in such communities might have more trust that minors will not try to illegally purchase cigarettes, are less concerned with youth smoking in their communities, or believe it is less likely they will be the target of a compliance check. This result supports the importance of enforcing laws that prohibit selling tobacco to minors in all communities. The likelihood that a retailer will ask for proof of age is greater in communities with local tobacco retailer licensing. Strong licensing policies usually include a fee set to fund enforcement of tobacco sales to minor laws, a provision that a violation of existing local, state, or federal tobacco regulation results in a suspended or revoked license, and financial disincentives through increased fines and penalties [35]. Our results suggest that requiring a license to sell tobacco may help reduce youth access to cigarettes through commercial sources. Similar to previous studies [26,39], our findings suggest that drug stores/pharmacies and smoke/tobacco shops may provide greater economic access to cigarettes for youth than do large supermarkets. Because previous studies have shown consistently that cigarette prices decrease cigarette smoking [6][7][8] and that youth are more price sensitive than adults [9,10], control over cigarette prices in such stores may help to reduce youth smoking. Establishing minimum prices for tobacco products may be one effective policy that can be implemented locally. Interestingly, the relationships between community characteristics and cigarette prices varied by cigarette brand. Higher median household income, lower prevalence of adult smokers in the community, and not having a local tobacco retailer licensing were all associated with higher prices of Marlboros. The finding about the inverse relation between prevalence of adult smokers in the community and price supports other studies which consistently have shown negative relation between cigarette prices and cigarette smoking [6][7][8]. Also, it is very likely that retailers charge more for cigarettes in communities with wealthy households. The positive association between local tobacco retailer licensing and Marlboro price may be explained by the other community-level characteristics identified in this study. It is possible that wealthiest communities with fewer smokers tend to adopt local tobacco retailer licensing ordinances which may also reflect broader community norms that are less supportive of smoking and therefore price promotion. The findings that the price of Newport was lower in communities with higher percentage of minors and higher in communities with higher percentage of Hispanics may explain why Newports are popular among high school students and less among Hispanics [28]. In another study in California, Henriksen et al [39] found that the price of Newport was lower in high school neighborhoods with higher percentage of Black students. Finally, different results regarding the associations between some community characteristics and Marlboro versus Newport prices may be explained by other factors which were not included in the current study. For example, it is possible that more extensive Newport price promotion in communities with local tobacco licensing diminishes any potential effect such policies may have. --- Limitations The results of this study should be considered in light of some limitations. First, only two confederate buyers conducted the surveys in each city, which limits our ability to consider characteristics of the buyers other than gender and age. Other studies, for example, have identified buyer ethnicity as significantly associated with increased youth tobacco sales [19,20]. Second, we only collected price data for Marlboro and Newport cigarettes. Although they are the most popular cigarette brands among high school students [28], they may not represent the range of prices among all cigarette brands. Third, the cross-sectional design of the study limited our ability to make directional inferences about relationships between contextual and community characteristics and outcome variables. For example, in understanding the relationships between the presence of age-of-sale signs and clerk requesting for ID, it is possible that posting these signs is simply an indicator that a retailer has implemented policies regarding checking ID rather than serving to increase clerks' awareness about legal age. Despite these possible shortcomings, this investigation enhances our understanding of the associations between contextual and community characteristics and youth commercial and economic access to tobacco. Such understanding can help policymakers to identify and target at-risk communities and outlets to decrease youth access to tobacco. --- What This Paper Adds Going beyond previous research, the present study examines the associations between a wide range of contextual and community-level characteristics and retailer compliance with underage tobacco sales laws and with pack prices in 50 non-contiguous mid-sized California communities. Since any illegal tobacco purchase attempt by minors occurs in an ecological environment that includes both contextual and community characteristics, this approach may help to highlight ways to tailor policies to reduce youth access to cigarettes via commercial sources more effectively. --- Acknowledgments

Objectives-This study examines contextual and community level characteristics associated with youth access to tobacco through commercial sources in 50 non-contiguous mid-sized California communities.The study is based on data from access surveys conducted by 4 confederate buyers (2 males and 2 females) in 997 tobacco outlets. City demographics, adult smoking prevalence and measures of tobacco outlet density, local tobacco retailer licensing and cigarette tax were included. Results-Multilevel regression analyses indicated that buyer actual age, a male clerk and asking young buyers about their age were related to successful cigarette purchases. Buyer actual age and minimum age signs increased the likelihood that clerks will request an ID. At the community level, higher percentage of minors, higher education, and a greater percentage of African Americans were associated with increased likelihood of a successful purchase. Lower percentage of minors, lower education, lower percentage of African Americans, and having a local tobacco retailer licensing were associated with retailer asked for ID. Additionally, supermarkets charged significantly more for a pack of cigarettes than small markets whereas smoke/tobacco shops and drug stores/pharmacies charged less. Higher prices were associated with higher median household income and greater percentage of Hispanics. Findings about community characteristics, however, differed by cigarette brand. Conclusions-This study enhances our understanding of the associations between contextual and community characteristics and youth access to tobacco through commercial sources which can help policymakers to identify and target at-risk communities and outlets to decrease youth access to tobacco.

Introduction Around 850 000 people live with dementia in the UK. 1 By 2040, this number is likely to increase by 57%. 2 People with dementia have high rates of physical morbidities that are primarily managed in general practice. 3 Reducing physical morbidity improves quality of life for people with dementia 4 and may slow cognitive decline. 5,6 The Dementia: Good Care Planning framework recommended an annual GP review as a minimum standard of care, as a Quality Outcomes Framework measure in 2015. 7 Routine physical health monitoring should include tailored management and prevention of comorbidities, with medication, nutrition, and hydration reviews. 7 UK policies prioritise fair access to health care, 8 but inequalities exist. In the UK, people from black ethnic groups living with dementia are less likely to be diagnosed, and people from Asian backgrounds less likely to receive symptomatic treatments post-diagnosis, compared with people from white ethnic groups. 9,10 Rates of physical morbidities are higher among people from black and Asian backgrounds compared with the white population, 9,11 so it would be concerning if health inequalities extend to preventive care for physical disorders. The study aimed to examine routine physical health monitoring for people with dementia in UK primary care and investigate whether this differs between white, black, and Asian ethnic groups as the main objective. As a secondary objective, the study investigated physical health monitoring for the overall population by comorbidity status. Informed by guidelines 7 and expert consensus, routine physical health monitoring was defined for the study . --- Method This cohort study used electronic primary care records from UK GP practices contributing to The Health Improvement Network database, 12 which is broadly representative of the UK population. At the time of data collection, THIN included 744 general practices and 15.6 million patients in 2016. In the UK, most physical health monitoring of people with dementia happens in primary care. 13 Clinical data are coded using Read codes. 14 THIN captures demographic information on: sex, ethnic group, birth year, and patient-level Townsend score . 15,16 --- Study population Individuals included were aged 50-105 years contributing to THIN between 1 April 2015 and 31 March 2016 to coincide with the QOF recording year. Individuals with dementia were defined by a Read code indicating a dementia diagnosis or an anti-dementia drug prescription , as in previous studies. 9,10 Individuals who were registered with a GP practice in the relevant period and had received a dementia diagnosis or been prescribed anti-dementia medication prior to that date were included. All GP practices met standard criteria for acceptable mortality reporting and computer usage. 17,18 Outcome and patient characteristics Good routine physical health care was defined as, over 1 year, a record of the following being received: a dementia review ; a BP measurement ; a GP consultation ; weight and/or BMI recorded ; and an influenza vaccination. Ethnic groups were catergorised based on Office for National Statistics classification: 9 white ; Asian ; black ; and mixed or other ethnic groups . 9 Individuals from mixed or other ethnic groups were excluded from analyses owing to small numbers . For comorbidity status, people were identified with a diagnosis of hypertension, myocardial infarction , stroke, diabetes, or chronic kidney disease . Covariates were age, sex, Townsend score, comorbidities, and prescribing index . Age was grouped in 10-year categories; the last category was 90-105 years owing to few individuals being aged >100 years. --- Analyses Multivariable Poisson regression was used to investigate the relationship between ethnic group and the outcomes. Models were adjusted for age, sex, Townsend deprivation, comorbidities, and prescribing index. IRRs were calculated for outcomes comparing black and Asian ethnic groups with the white ethnic group. Comorbidities were also stratified comparing the cohort with and without comorbidity. Characteristics of individuals were compared with and without the ethnic group recorded, and complete case analyses were conducted for those with a record of ethnic group and Townsend score. Missing ethnic group data were then imputed using multiple imputation by chained equations. 20 Compared with complete case analysis, multiple imputation can greatly strengthen results by collating all the information available from individuals with observed data. 21 This way, information was used from the full dataset in the analysis to provide more precise estimates of the outcomes . The multiple imputation results presented in this article are considered to be the primary results. They were consistent with the complete case analysis. Stata was used. --- Results --- Included individuals Of the 20 821 individuals aged 50-105 years in the study, 10 570 had their ethnic group recorded. Of 10 570 individuals with a recorded ethnic group, 96.6% were recorded as from a white ethnic group, 2.1% from an Asian ethnic group and 1.3% from a black ethnic group. After multiple imputation of ethnic group, 96.6% were from a white ethnic group, 2.1% from an Asian ethnic group and 1.2% from a black ethnic group . The mean age was 79.8 years. Among this cohort, 59% had diabetes, 55% had hypertension, 44% had CKD, 23% had a stroke diagnosis, 9% had an MI diagnosis, and 13% had no comorbidity recorded. There were some slight differences in age and sex distribution between those with and without ethnic group recorded . --- Annual dementia review An annual dementia review was recorded for 14 105 individuals with dementia. There was no significant difference in the proportion of people with dementia from black, Asian, or white ethnic groups receiving a review . Neither was there significant difference in the proportion of people with dementia receiving a review according to comorbidity status . --- Annual blood pressure monitoring At least one BP check was recorded in 16 611 individuals with dementia, with an average of 2.9 BP recordings per person per year. There was no significant difference in the proportion of people with dementia from black, Asian, or white ethnic groups receiving at least one BP check. Those without comorbidities were less likely to have at least one BP check, compared with those with at least one comorbidity . --- Number of GP consultations At least one GP surgery, telephone or home visit consultation was recorded in 20 231 individuals with dementia. There was no significant difference in the proportion of people with dementia from black, Asian, or white ethnic groups receiving at least one GP consultation. There was also no significant difference in the proportion of people with dementia according to comorbidity status receiving at least one GP consultation . --- Annual weight and/or BMI recording A weight and/or BMI was recorded in 9927 individuals with dementia. Compared with white groups, black people were 23% less likely and Asian people 16% less likely to have their weight recorded. People with dementia without any comorbidities were less likely to have at least one weight and/ or BMI check within a year, compared with people with at least one comorbidity . --- Annual influenza vaccination Receipt of the annual influenza vaccination was recorded in 16 781 individuals with dementia. There was no significant difference in proportions of people with dementia from black, Asian, or white ethnic groups receiving the vaccination. There was also no significant difference in the proportion of people with dementia according to comorbidity status receiving the vaccination . --- Discussion --- Summary In this large UK primary care study, only two-thirds of people with dementia received an annual review, despite guidance recommending this as a minimum standard. 7 Less than half of people with dementia in the study had a weight and/or BMI recorded; this happened less frequently in people from black and Asian ethinic groups compared with white ethnic groups, and less frequently in people without comorbidities compared with those with ≥1 comorbidity. A fifth of people with dementia did not receive an influenza vaccine or a BP check within a year, and only 57% of people with dementia without other comorbidities had a BP check. Ethnic group was not associated with differences in BP monitoring, GP consultations, influenza vaccination, or dementia annual review. --- Strengths and limitations THIN comprises around 6% of the UK population, and is broadly representative in terms of demographic and health variables. 13 Ethnic group was imputed to address missing data, and results from complete case and multiple imputation analyses were similar . Imputing missing data meant the full sample of individuals with dementia could be used to provide more precise estimates of outcomes. People living with dementia were identified through a coded dementia diagnosis or anti-dementia drug repeat prescription, which may exclude a small number diagnosed with dementia within secondary care and not captured in GP records. The study focused on ethnic group as black, Asian, or white as data at more detailed levels were not well recorded. However, this may have obscured smaller intra-group differences. Ethnic groups are considered to share a common ancestry, culture, and feeling of solidarity with one another. 22 There is wide variation within minority ethnic groups in country of origin, language, religion, socioeconomic characteristics, and experiences, but enough shared culture with regards to family structures, identity, and health beliefs to make ethnic group relevant to health behaviours. 22 While some comorbidities were accounted for, analyses may have been confounded by unmeasured variables affecting service use. This study also had relatively low numbers of people from black and Asian backgrounds, which may have obscured smaller differences between groups. The study investigated the likelihood of receiving at least one BP or weight and/or BMI measurement, which is a crude binary outcome and it is not known whether those with abnormal values received adequate treatment or follow-up. While Read codes were used to capture influenza vaccination delivered in other settings, some may have been missed if they were not documented in GP records. --- Comparison with existing literature To the authors' knowledge, this is the first study to explore routine GP physical health monitoring and ethnic group in people with dementia. The finding that black and Asian ethnic groups are less likely to have their weight and/or BMI recordedx adds to other known barriers to good dementia care for these groups. 9,10 Contextual barriers to help-seeking among these groups include negative experiences of health services and perceptions that caring is a family responsibility. [22][23][24] However, for most outcomes, no differences between ethnic groups were found. This may indicate that among people with dementia, barriers to accessing physical health care are shared among all ethnic groups, or that smaller differences between groups were undetected owing to small numbers of minority groups. The study found weight and/or BMI had the lowest rates of completion of all the component of monitoring studied, so perhaps inequalities emerge in less routine care activities, because clinicians are more vulnerable to unconscious bias when exercising more discretion. Lower levels of overall comorbidity were identified among white groups compared with black and Asian groups, as is consistent with the literature. 9,11 The findings contrast with those investigating mental health care, where minority ethnic groups are less likely to receive anti-dementia medication where potentially indicated and receive antipsychotic medication for longer. 10 The impact of stigma on mental health care access compared with physical health care may explain this difference. Between 2002-2013, dementia annual review uptake was under 50%. 25 The results of the present study show an improvement in 2015-2016 , 7 but still a third of people with dementia had no review recorded, despite most having at least one annual GP contact. This lack of review is important as pressures on secondary mental health services are rising, and responsibility for routine monitoring may increasingly fall to primary care. Guidelines for the annual dementia review are extensive; they emphasise a tailored review to consider physical health, mental health, nutrition, optimising polypharmacy, care needs, functional ability, and end-of-life discussions, among other areas. 7 Such broad guidance may challenge GPs facing workload and time pressures. 26 Some studies suggest that GPs may believe they have little to offer patients with dementia. 27 The rationale for the dementia review needs to be clearly explained, including the importance of routine physical health monitoring to dementia outcomes. Other factors explaining low uptake may include varying QOF renumeration 28 and stigma related to dementia; 29 GPs may be undertaking the dementia review tasks, but not recording them as such. While dementia detection and diagnosis rates in primary care are rising, 30 the findings suggest optimal care post-diagnosis lags behind. The study found less than half of people with dementia had a weight and/or BMI recorded, consistent with another THIN study reporting 47% had this recorded in 2010-2013, 25 and other reports of poor weight recording in primary care. 31 Eating and drinking difficulties are common in dementia. The risk of malnutrition increases as dementia progresses and people with dementia have 10 times more malnutrition or dehydration-related hospital admissions compared with age-matched controls. 32 Anti-dementia drugs causing side effects, such as nausea, can also exacerbate difficulties. Other research has highlighted concerns that nutrition support for older people is insufficient 33 and weight loss may not be recognised as a problem. 34 The study found over 80% of people with dementia had at least one BP check, one GP contact, and the influenza vaccination, confirming previous studies. 25,35 An annual BP check is not specified in guidelines, although this was used as a marker of good preventive health care. While some studies suggest that good BP control can reduce risk of further cerebrovascular changes, others describe a complex relationship between optimal BP and dementia. 36 People with dementia are more likely to experience postural hypotension and adverse effects of low BP, 37 which is an argument for regular monitoring, regardless of comorbidities. However, standard BP monitoring may cause anxiety, 38 so should be considered carefully. --- Implications for research and practice The results suggest routine physical health monitoring for people living with dementia, particularly nutrition monitoring, is inconsistent and requires improvement. Much emphasis in primary care has been on dementia detection, where meaningful improvements have been achieved and sustained. Priorities should now shift to post-diagnosis support. 39 Dementia should be considered a long-term health condition requiring routine monitoring in its own right. At present, guidelines for the annual dementia review are extensive. Prioritising components of the review that are realistic and achievable in general practice should be central to any revisions. Improved weight recording and management of nutrition should be prioritised, promoting ethnic and sex equalities in access. The role of BP monitoring, particularly for people living with dementia without other cardiovascular risk factors, should be considered by researchers. Defining best practice in physical healthcare monitoring and management of comorbidities could shape future dementia guidelines. --- Ethical approval The NHS South-East Multi-Centre Research Ethics Committee approved the use of THIN for scientific research in 2003. The IQVIA World Publications Scientific Review Committee granted scientific approval in March 2017 . --- Provenance Freely submitted; externally peer reviewed.

Background: Good physical health monitoring can increase quality of life for people with dementia, but the monitoring may vary and ethnic inequalities may exist. Aim: To investigate UK primary care routine physical health monitoring for people with dementia by: (a) ethnic groups, and (b) comorbidity status. Design & setting: A retrospective cohort study was undertaken using electronic primary care records in the UK. Method: Physical health monitoring was compared in people with dementia from white, black, and Asian ethnic groups and compared those with ≥1 comorbidity versus no comorbidity, from 1 April 2015 to 31 March 2016. Using the Dementia: Good Care Planning framework and expert consensus, good care was defined as receiving, within 1 year: a dementia review; a blood pressure (BP) check (at least one); a GP consultation (at least one); a weight and/or body mass index (BMI) recording (at least one); and an influenza vaccination. Results: Of 20 821 people with dementia, 68% received a dementia review, 80% at least one BP recording, 97% at least one GP contact, 48% a weight and/or BMI recording, and 81% an influenza vaccination in 1 year. Compared with white people, black people were 23% less likely and Asian people 16% less likely to have weight recorded (adjusted incidence rate ratio [IRR] = 0.77, 95% confidence interval [CI] = 0.60 to 0.98/0.84, 0.71 to 1.00). People without comorbidities were less likely to have weight recorded (adjusted IRR = 0.74, 95% CI = 0.69 to 0.79) and BP monitored (adjusted IRR = 0.71, 95% CI = 0.68 to 0.75).Ethnic group was not associated with differences in physical health monitoring, other than weight monitoring. Comorbidity status was associated with weight and BP monitoring. Physical health monitoring in dementia, in particular nutrition, requires improvement.Good physical health supports quality of life for people living with dementia. Routine physical health monitoring in general practice can support this and should be accessible to all. This is the first study to explore routine GP physical health monitoring and ethnic group in people living with dementia. Findings can inform policies, promoting access to good post-diagnostic support.

Background Alcohol consumption among adolescents is a common concern that is growing in most countries and, notably, hazardous and harmful drinking patterns seem to be on the rise [1,2]. There is extensive evidence that alcohol and physical aggression are associated. It has been shown that adolescents who display violent behaviour are more likely to exhibit problematic alcohol consumption than other adolescents [3,4], excessive alcohol use or drunkenness [5]. Similarly, it has been demonstrated that adolescents who misuse alcohol have higher rates of violent behaviours [6][7][8]. According to Fagan [9], alcohol consumption provides a "provocative context" for violence, rather than a direct cause, whereas another study proposes alcohol consumption as a moderating variable with a conditioning and reinforcing role in explaining aggressive behaviour [10]. Results from a longitudinal study [11], focusing on the relationships between alcohol misuse, antisocial behaviour and alcohol-related problems at particular ages, strongly support the reciprocal hypothesis. Alcohol misuse and antisocial behaviour establish a "feedback loop" in a joint-effects model , whereas the susceptibility hypothesis is prevalent in the shorter term model. Similarly, results from a study on alcohol and violence [12] suggest that alcohol has a "magnifying" effect, amplifying underlying aggressive tendencies. Even experimental studies support a strong relationship between acute alcohol consumption and aggressive behaviour, confirming the idea that acute alcohol consumption facilitates or increases the expression of aggressive behaviour . In the body of research on the relationship between alcohol consumption and aggression, aggressive behaviours have been investigated in many respects. Our study deals with alcohol-related aggression, defined as the co-occurrence of drinking and physical fighting within a single episode [14]. In this instance, questions remain concerning which alcohol consumption indicators are most suitable to measure the impact on aggressive behaviours. Although the "average amount of alcohol consumed" is consistent with a range of physical and social consequences, a growing number of behavioural studies provide evidence that not only quantity but also patterns of drinking are measures that relate to drinking outcomes [1,15,16] and, notably, to alcoholviolence association [4]. Besides frequency of use, a basic parameter indicating the regularity of drinking, one of the most studied characteristics is binge drinking , which exhibits high prevalence among youth [1,17]. Other studies, focused on alcohol's negative effects, report that the extent of drunkenness rather than total volume of alcohol consumed relates to acute consequences such as various types of aggression and violence [4,[18][19][20]. All the indicators, above, of alcohol use, if combined, can capture the diversity of drinking customs. However, a difficulty arises when they are used concurrently in multivariate analyses, due to multicollinearity that may result from the high functional correlation among them. To overcome this difficulty, we propose a different characterization of drinking patterns that constructs a composite metric that combines standard alcohol use parameters such as frequency of consumption, frequency of binge drinking, and frequency of perceived intoxication. This new analytical approach could provide more detailed information on the relationship between alcohol consumption and aggression. Moreover, since it has been demonstrated [21,22] that preferences for alcoholic beverage may reflect different attitudes towards alcohol consumption, such preferences were also taken into account. The connection between alcohol consumption and aggressive behaviour operates at multiple levels and is the result of a dynamic interplay among personal and socio-environmental systems [23]. This comprehensive approach, incorporating concepts derived from problembehaviour theory [24], offers a theoretical framework for better understanding underage alcohol use. A broad array of factors potentially affecting alcohol-related aggression in adolescents has been identified in previous research. For this reason, a number of individual, environmental and behavioural variables that may be either "protective" or a "risk" factor for adolescent behaviour problems were also evaluated. Both violent behaviour and alcohol drinking undergo significant changes during adolescence and age and gender can characterize both behaviours. The same applies to substance use and leisure time activity. Another set of factors, whose influence has been repeatedly assessed, involves the socio-environmental system. In this context, family function vs dysfunction, as well as school performance, and peer influence are welldocumented confounders that can impact involvement in alcohol-related aggression [13,18,23]. Thus, the purpose of the present study was threefold: a) to test the ability of observed drinking patterns for association with alcohol-related aggression, b) to determine if drinking patterns, as evaluated in this study, add value to common indicators of alcohol use, and c) to verify the role of a number of factors as mediators between drinking patterns and alcohol-related aggression. --- Methods A full description of sampling and data collection procedures has been reported in the 2011 European School Survey Project on Alcohol and Drugs Report [25]. Briefly, standardized data collection was performed using an anonymous self-administered questionnaire completed on a voluntary basis in the classroom setting. The authorization of the school head to fill in the Italian ESPAD questionnaire by the students was required. --- --- Alcohol use indicators Indicators of alcohol consumption were assessed for the month prior to the survey . Three questions from the 2011 ESPAD core questionnaire were used as screens: a) "During the last 30 days, on how many occasions have you had any alcohol beverage to drink?" b) "Think back over the last 30 days. How many times have you had five or more drinks on one occasion?" c) "During the last 30 days, on how many occasions have you been intoxicated from drinking beverages, for example staggered when walking, not being able to speak properly, throwing up or not remembering what happened?" Questions a) and c) had 7 response categories: "0, 1-2, 3-5, 6-9, 10-19, 20-39 and 40+ occasions", while question b) had 6 response categories: "none, 1, 2, 3-5, 6-9, 10+ times". Alcoholic beverage preferences and their frequency of use were also considered using the question "Think back over the last 30 days. On how many occasions have you had any of the following to drink?" Possible choices were beer, alcopops, wine, and spirits with response categories "0, 1-2, 3-5, 6-9, 10-19, 20-39, 40+ occasions". --- Alcohol-related aggression Alcohol-related aggression was evaluated on the basis of the following question: "Because of your own alcohol use, how often during the last 12 months have you experienced physical fighting?" The response categories were "0, 1-2, 3-5, 6-9, 10-19, 20-39 and 40+ occasions". Due to the low number of observations reported in the upper ranges, the response was dichotomized . "Experienced physical fighting" is intended as direct involvement in a fight. We use the term "aggression" in place of "physical fighting" in the current text. --- Other variables Some other variables potentially affecting adolescent behaviour were also considered. These parameters were grouped as follows: a) family: parental monitoring ; family structure ; b) substance use: use of substances at least once during the last year vs none; having smoked cigarettes daily during the last month vs less than one cigarette per day ; c) school: having missed school, without a valid reason, for 3 days during the last month vs less than 3 days; having obtained high marks in the last term vs low/ medium marks; d) frequent vs infrequent leisure time activity: sports practice, going out in the evening , slot machines gambling ; e) friends' behaviour with alcohol: categorized as nondrinkers, regular drinkers but few get drunk, regular drinkers and most get drunk. --- Statistical analysis Principal component analysis [26] was applied to the three indicators of alcohol consumption, expressed as frequencies, to obtain three independent factors representing different drinking patterns and used simultaneously in the regression model. PCA extracts a set of principal components obtained as a linear combination of the original indicators. No rotation procedures were required to facilitate the interpretation of the factors. The contribution of each indicator is the loading derived from the analysis. A positive loading means that higher levels of an indicator are associated with higher levels of that factor and a negative loading means that lower levels of an indicator are associated with higher levels of that factor. Each principal component represents a certain amount of total variance in the data: by using all components the total amount of variance is conserved. The components obtained were interpreted in terms of different alcohol drinking patterns. Each pattern can be treated as a numerical variable similar to an assessment scale, with a minimum and a maximum value. In each pattern, an increase of one unit must be interpreted in terms of the composite indicators and their specific contributions. For example, a pattern would result from a linear combination of the three indicators with positive loadings , characterized by a scale that increases with frequency of alcohol use or binge drinking or intoxication, separately, or in combination. In this pattern, the minimum drinking pattern value is equal to zero and means no alcohol use in the last month and the maximum value is equal to the highest frequency of intoxication and binge drinking . Pearson's correlation was used to explore the relationship among the three indicators and between drinking patterns and frequency of use of specific alcoholic beverages . Logistic regression analysis was performed to verify the association between aggressive behaviour and alcohol consumption, evaluated both as individual indicators and as drinking patterns. Individual indicators and drinking patterns were treated as continuous variables. Three models were evaluated: univariate logistic regression using indicators representing alcohol use , multivariate logistic regression using indicators indicating alcohol use and multivariate logistic regression using drinking patterns . Results are reported using beta coefficients and standard errors, odds ratios and 95% confidence interval . Univariate and multivariate logistic regression models were used to control for potential confounding effects. Alcohol indicators were tested for confounding effects by jointly introducing them into the model and examining beta coefficients: a change in beta coefficient greater than 10% was considered to be a source of confounding. In Model 2, multicollinearity among independent indicators of alcohol use was also evaluated using the variance inflation factor : a VIF that exceeded 5 was taken as an indication of multicollinearity [27]. All other personal and behavioural variables had been previously tested using univariate analysis and those that appeared statistically significant were included in the multivariate regression model along with drinking patterns. All the analyses were performed separately against gender. Statistical significance was set at p < 0.05 . All the analyses were performed using Stata software, version 10.1. --- Results --- Descriptive statistics Overall, the majority of students who have consumed alcohol during the last year were also current drinkers: in fact, 82% of them consumed alcohol at least once in the last month. Among them, alcohol use without binge drinking or perceived intoxication was a common habit ; binge drinking without any experience of perceived intoxication was also significant , whereas perceived intoxication alone occurred infrequently . Finally, 14% of students reported at least one experience of binge drinking and at least one of perceived intoxication. Significant correlations were found among the three indicators: for alcohol use and binge drinking r = 0.57, for alcohol use and perceived intoxication r = 0.40, for binge drinking and perceived intoxication r = 0.47. Regarding alcohol-related aggression, about 12% of students have been involved, predominantly males , but only 3.7% of adolescents reported involvement more than twice. --- Drinking patterns Drinking patterns were calculated on the basis of current alcohol consumption among adolescents who had consumed alcohol at least once in the last year. There was a portion of adolescents who did not use alcohol in the month prior to the survey. Drinking patterns identified by PCA were defined as 1) Drinking to Excess ; 2) Drinking with Intoxication ; and, 3) Drinking but Not to Excess . Table 1 summarizes the PCA results. DE pattern accounted for 65% of the total variance and was characterized by a positive correlation with the frequency of all individual indicators and is interpreted as the pattern of those who 1) drink frequently, 2) report perceived intoxication and 3) do binge drinking. When using a composite numerical variable, the minimum value was zero and the maximum value was 10 . Thus, an increase of one unit in DE pattern indicated an increase in frequency of excessive alcohol use. DE pattern showed that experience of excessive drinking alcohol among adolescents is both related to binge drinking and to experiencing intoxication. DI pattern exhibited 21% variance and also a negative correlation with frequency of alcohol consumption, a positive high correlation with perceived intoxication and a small negative correlation with frequency of binge drinking and it is interpreted as the pattern of those who 1) drink infrequently, 2) report perceived intoxication, but 3) do not binge drink. The corresponding minimum composite value was -4 and the maximum value was 2 . An increase of one unit in DI pattern indicated an increase in frequency of perceived intoxication during every drinking experience and a decrease in frequency of binge drinking. DI pattern showed that perceived intoxication among adolescents was not necessarily linked to binge drinking, and there was a proportion of them who experienced intoxication even if they drank infrequently. DNE pattern exhibited a variance of 14% and a positive correlation with frequency of alcohol consumption, a positive, but low correlation with frequency of perceived intoxication and a negative high correlation with frequency of binge drinking and it is interpreted as the pattern of those who 1) drink frequently, 2) do not report perceived intoxication, and 3) infrequently binge drink. The corresponding minimum ordinal value was -1 and the maximum value was 5 . An increase of one unit in DNE indicated an increase in frequency of alcohol consumption without binging. DNE pattern showed that there was a portion of adolescents who experienced moderate alcohol use, drinking frequently but without experiencing binge drinking or perceived intoxication. Figure 1 shows a three-dimensional representation of the aforementioned drinking patterns: points represent all the theoretical values that the specific pattern could assume in correspondence of the combination of frequencies of the three selected indicators. The points are presented in shades of grey, from light grey at the lowest value to dark grey at the highest. As shown, ED pattern had higher values at increasing frequencies of all indicators, DI pattern had higher values at increasing frequency of both alcohol use and intoxication, DNE pattern had higher values in correspondence with higher frequency of alcohol use and lower frequency of binge drinking. --- Correlation between drinking patterns and alcoholic beverages Pearson correlation between the DE pattern and frequency of the specific alcoholic preferences resulted in significant correlations with all beverages , whereas significant negative correlations were reported for the DI pattern . For DNE pattern, weak significant positive correlations were found with all beverages . --- Association between alcohol consumption and alcohol-related aggression The association of individual alcohol use indicators and drinking patterns with alcohol-related aggression was tested using logistic regression analysis. Table 2 shows results from several models. First, individual alcohol indicators were analysed in a univariate model : the higher the frequency of the three indicators, the higher the likelihood of being involved in alcohol-related aggression and in both genders. Second, these alcohol use indicators were analysed using a multivariate model : although, as expected, a positive association with alcohol-related aggression was found for all the indicators, changes in beta coefficients of more than 30% denoted a substantial confounding effect. Mean value of VIF was greater than 5 , with the highest value for frequency of alcohol consumption . Finally, in Model 3 drinking patterns, only, were examined: in both genders, all were significantly associated with alcoholrelated aggression, but while DE and DI patterns correlated positively, DNE pattern had a negative correlation. Table 3 reports results from logistic regression between alcohol-related aggression and drinking patterns, controlling for the other variables. No changes in drinking patterns' association were observed in males, whereas in females DI pattern was no longer significantly associated with alcohol-related aggression. Age was negatively associated with alcohol-related aggression only for males. Investigating the influence of the drinking habits of peers, having many friends who become intoxicated was more likely observed in male adolescents who exhibited alcohol-related aggression. Overall, illegal drug use, truancy, and frequent evenings spent outside of the home were habits strongly associated with alcohol-related aggression and without distinction of gender. Cigarette smoking showed a strong association but only for males. Other factors such as achieving high marks at school and high level of parental monitoring were negatively associated with alcohol-related aggression, whereas frequent participation in sports, evaluated only in males, was more likely associated with alcohol-related aggression. --- Discussion Regardless of the minimum legal drinking age , the experience of drinking alcohol is widespread among young people in Italy. Our findings have shown that alcohol consumption was higher in male students and that, for both genders, alcohol use without major consequences predominated even if binge drinking was observed as a widespread pattern . It can be also observed that binge drinking and perceived intoxication occurred jointly , supporting the idea of identifying and using patterns of drinkingresulting in composite relationshipsrather than evaluating alcohol habits using a single indicator or by type/ quantity of beverage consumed. In addition, although the ESPAD questionnaire clearly defines a "drink" , a glass/bottle can of cider , a bottle of alcopops , a glass of wine , a glass of spirits or a mixed alcoholic beverage), it remains quite difficult to estimate the actual amount of ethanol consumed as well as the individual sensitivity to any specific alcohol volume. For this reason, using PCA results, we have examined the impact of three styles of alcohol consumption on alcohol-related aggression: 1. alcohol consumption leading to DE pattern, the greatest risk; 2. consuming alcohol infrequently but leading to perceived intoxication , popular behaviour among adolescents in the so called "dry" countries, but now also increasing in Italy; 3. drinking but not to excess , a "moderate" approach that is poorly studied. Regarding alcoholic beverage preferences, the DE pattern appeared to correlate with consumption frequency of spirits more than other alcoholic beverages, while DI negatively correlated with consumption frequency of beer to a greater extent. Lastly, the DNE pattern correlated poorly with all beverages. It would be important to explore this in future studies using the ESPAD data, particularly comparing different drinking cultures to assess whether drinking patterns are connected to new drinking habits . In investigating the relationship between aggressive behaviours and alcohol use, some studies have considered single parameters to assess drinking [23,28], whereas other studies have evaluated the interaction of several parameters to characterize different classes of drinkers [29] or have utilized a single drinking pattern score obtained by combining several indicators [30]. The analysis of drinking patterns is a key factor in alcohol-related aggression as it draws information which otherwise cannot be inferred. In our study, the main advantage of this approach was to highlight a drinking pattern that has so far been little studied: in fact, our study not only confirms what is already known, i.e., excessive drinking is associated with an increased likelihood of alcohol-related aggression but it has also identified a negative association with moderate drinking: in other words, moderate drinking significantly decreases the likelihood of being involved in alcohol-related aggression, a finding confirmed in both genders. It should be further investigated as to whether this is due to a "protective" role of this drinking pattern or to a more general moderate behaviour. Actually, little is known about young people consuming alcohol at "lowrisk" levels: only recently, research has addressed this issue in an attempt to establish drinking guidelines for youth [31]. For this reason, the DNE pattern deserves greater attention in future studies because it is a drinking style adopted widely among young Italian drinkers and a thorough understanding of this pattern may provide additional perspectives on other behaviours. In addition, DI pattern also merits comment: although less strongly associated with alcohol-related aggression compared to DE pattern, it provides an insight into the risky behaviour of those who drink infrequently without experiencing binge drinking, reflecting the fact that alcoholrelated aggression is not exclusively related to frequent or compulsive drinking. Many other factors enter into the relationship between aggressive behaviour and alcohol use among adolescents. Some gender differences were detected by the logistic regression model: in fact, while in males the relationship between alcohol-related aggression and all the three drinking patterns was not modified by the potentially confounding variables, in females the association between alcohol-related aggression and DI pattern was no longer present. This finding suggests that, in females, socio-environmental characteristics underlying alcoholrelated aggression and DI pattern are similar and therefore responsible for the association. Alcohol-related aggression varied by age only in males, demonstrating that involvement in alcohol-related aggression was more likely in younger males and that this behaviour gradually changes through the teenage years. The effect of age appears therefore relevant, especially in males, as shown also in previous studies [18,28]. Furthermore, the drinking behaviour of peers has often been considered influential. Compared to other studies that have evaluated only the number of peers who drink, in our study we have considered different friends' habits: non-drinkers, regular drinkers but few get drunk, regular drinkers and most get drunk: from our results, associating with friends who drink alcohol doesn't appear influential as long as they do not consume alcohol in excess. As observed by others [18], the gender-specific analysis reveals significant differences and underlines the differential impact, for males and females, of drinking patterns and of the other factors that play a role in alcohol-related aggression. Other correlates were equally relevant for both genders. As already shown [32], the use of illegal drugs resulted as always positively associated with alcoholrelated aggression, indicating a tendency to concurrent problem behaviours. In addition, truancy and simply spending many evenings outside the home environment represent attitudes that can contribute to problematic behaviours, and, as found in our study, to alcohol-related aggression involvement. Overall, from the outcomes of the analysis, it can be argued that alcohol-related aggression in young people is more commonly associated not only with drinking to excess, but also with a number of features that express discomfort, converging towards an overall risk-taking behaviour. Some limitations of the study should be mentioned. First, data were derived from a school-based sample of adolescents, thus excluding school dropouts, and were self-reported. Second, the definition of alcohol-related aggression was based on a question that asked participants if they had experienced physical fighting "because of your own alcohol use". In order to answer to this question, the participants must attribute their fighting behaviour to their alcohol use. Since the question specifically asks participants only about fighting behaviours that were attributed to drinking, our analyses may underestimate the number of all fighting and drinking that co-occur, since a portion of these occurrences may not have been attributed to the drinking. Third, conclusions on the causal relationships cannot be drawn as the data were cross-sectional. Moreover, we recognize the lack of other important indicators such as the volume of drinking or the drinking context that could provide more comprehensive information regarding alcoholrelated aggression. --- Conclusions Our results suggest that alcohol consumption, alcoholrelated aggression and their relationship are the result of a more complex system in which many other factors play important roles, leading to overall risk-taking behaviours. Therefore, for long-term impact, efforts to reduce aggressive behaviour in youths and policies aimed at curbing alcohol use need to adopt a "whole system" approach that should include both regulatory interventions and concomitant strategies for reducing the negative consequences of problems once they have emerged. With specific regards to alcohol consumption, interventions involving education and returning to Mediterranean cultural traditions , could possibly encourage a more responsible approach to alcohol consumption and a greater awareness of the consequences of excessive drinking. --- Competing interests The authors declare that they have no financial competing interests. Authors' contributions SM obtained the funding for the study. VS and SM developed the study design. LM managed the literature searches and summaries of previous related work. VS and VL undertook the statistical analysis. VS, SM, SP, and LM interpreted the data. LM and VS wrote the initial draft of the manuscript. All authors contributed to and have approved the final manuscript.

Background: Although there have been a wide range of epidemiological studies examining the impact of patterns of alcohol consumption among adolescents, there remains considerable variability in both defining these patterns and the ability to comprehensively evaluate their relationship to behavioural patterns. This study explores a new procedure for defining and evaluating drinking patterns and integrating well-established indicators. The composite measure is then used to estimate the impact of these patterns on alcohol-related aggressive behaviour among Italian adolescents. Methods: Data were collected as part of the 2011 European School Survey Project on Alcohol and other Drugs (ESPAD). A national sample of 14,199 students aged 15-19 years was collected using an anonymous, selfadministered questionnaire completed in a classroom setting. Drinking patterns were established using principal component analysis. Alcohol-related aggression was analysed as to its relationship to patterns of drinking, behaviour of friends towards alcohol use, substance use/abuse, school performance, family relationships and leisure activities. Results: Several specific drinking patterns were identified: "Drinking to Excess" (DE), "Drinking with Intoxication" (DI) and "Drinking but Not to Excess" (DNE). A higher percentage of males were involved in alcohol-related aggression compared with females. In males, the DE and DI patterns significantly increased the likelihood of alcohol-related aggression, whereas the DNE pattern was negatively associated. Similar results were found in females, although the DI pattern was not significantly associated with alcohol-related aggression. Overall, cigarette smoking, illegal drug use, truancy, limited parental monitoring, frequent evenings spent outside of the home and peer influence associated strongly with alcohol-related aggression. Conclusions: Our findings suggest that drinking patterns, as uniquely monitored with an integrated metric, can: 1) explain drinking habits better than commonly used indicators of alcohol use and 2) provide a better understanding of behavioural risks such as alcohol-related aggression. Environmental background also appears to strongly associate with this type of aggressive behaviour.

Introduction School culture is one of the oldest, most complex and important concepts in education. In relation to school improvement, it has also been one of the most neglected. School culture considers the basic essence of an organization's culture to be, "the deeper level of basic assumptions and beliefs that are shared by members of an organization, that operate unconsciously, and that define in a basic 'taken-for-granted' fashion an organization's view of itself and its environment" . Culture describes the way the institution operates and acts as a screen or lens through which the world is viewed. In essence, it defines reality for those within a social organization, gives them support and identity and creates a framework for occupational learning . Each school has a different reality or mindset of school life, often captured in the simple phrase "the way we do things around here". It also has its own mindset in relation to what occurs in its external environment . Culture is, thus, "situationally unique", as can be seen in the example of two ostensibly similar secondary schools, located in the same area and drawing from the same population, with the same number of students attending them. These two schools view these students, their work and external constraints they face in very different ways. A school's culture is shaped by its history, context and the people in it . Like the larger social culture, a school culture results from both conscious and unconscious perspectives, values, interactions, and practices, and it is heavily shaped by a school's particular institutional history . Students, parents, teachers, administrators, and other staff members all contribute to their school's culture, just like other influences such as the community in which the school is located, the policies that govern how it operates, or the principles upon which the school was founded . Schein provides a more specific definition of organizational culture by emphasizing its elements as "pattern of shared values, norms, beliefs, artifacts, underlying assumptions and value systems among a group of people. According to Schein, the values reflect the philosophy, ideology, moral and ethical codes, goals, ideas and standards of the organization, including basic essentials that provide judgments regarding what is wrong and what is right. Organizations possess a cultural perspective that is centered on a system of values that define patterns and functions of the organization . The key functions identified by Parsons as essential to the organization's cultural perspective and driven by its value system comprised of; goal achievement, change management, teamwork and agreement on values. Goal achievement is the manner in which the organization establishes goals and mobilizes resources to attain these goals. Change management relates to the way an organization can adapt to changes in goals and procedures while teamwork pertains the way members relate to each other and their commitment to the organization. In this study, school culture meant the extent to which teachers manage changes in schools, extent to which goals are achieved, extent of coordinated teamwork, extent to which students' needs are satisfied and the extent of agreement on values like respect for one another, in different school settings. This is in conformity with Kaggwa who defines school culture as a set of cultural unique behavioural conduct that differentiates a school from other schools in the same locality. Teachers' job performance mainly depends on the teacher characteristics such as knowledge base, sense of responsibility, and inquisitiveness; the student characteristics such as opportunity to learn and academic work; the teaching factors such as lesson structure, and communication; the learning aspects such as involvement and success; and the classroom phenomena such as environment and climate, and organization and management. If the teachers take care of these factors, their performance can be enhanced to the optimum level . Proxies implemented by states and districts of Nigeria to determine teacher quality have been woefully inadequate. Teacher entrance and exit, examination scores, years of experience, advanced degrees, and teaching credentials are either not related to student achievement or ratings of teacher effectiveness . Leigh and Mead clearly bring out the fact that the quality of teaching has come down gradually world over, the skills of teachers have come down due to outdated preparation on the part of the teacher and stagnant compensation schemes by the management of the educational institution. This condition in the recent years for the teacher has led to very few growth opportunities and inadequate compensation structures. The condition is worse with disadvantaged students who require excellent teachers but have the least capability. Leigh and Mead in their suggestion for lifting performance of teachers, have emphasized the need for periodical performance appraisal just as it is in the corporate or business organization. Teachers need to be periodically evaluated and the compensation structure will have to be based on performance. An effective policy needs to be developed in order to modernize and enrich teacher quality for hiring, evaluating and compensating . Aacha describes performance of teachers as the teacher's ability to integrate experience, teaching methods, instructional materials, knowledge and skills in subject matter delivery to students both inside and outside the classroom. On the other hand, Obilade states that teachers' job performance can be described as "the duties performed by a teacher at a particular period in the school system in achieving organizational goals. In this study, teachers' job performance meant the ability of teachers to perform their prescribed duties. Teachers' job performance was operationalized as scheming of work, lesson planning, involvement in extracurricular activities, involvement in discipline management, involvement in guidance and counseling, participation in staff meetings, lesson delivery/actual class teaching, maintenance of records of work and teachers' physical presence in school. Bukhuni and Iravo carried out a study to establish the effect of school culture on employee performance in public secondary schools in Bungoma North sub-county, Kenya. The study included 36 head teachers and 140 teachers from 36 public secondary schools. The study findings revealed that employee performance was more likely to be enhanced in schools with established cultures. The study further revealed a significant direct relationship between school values like, employee collaborations and employee performance. Discrete indicators like beliefs, convictions, values, norms, philosophy, mission, vision, goals, assumptions and moral values also contribute to higher performance of a school A positive and strong school culture motivates teachers, increases academic achievement by students, increases job satisfaction, commitment and cooperation among teachers, increases teachers' dedication to work . The six basic components of culture in schools with high achievement among teachers are; shared vision, traditions, collaboration, shared decision making, innovativeness and communication . In a related study, Gruernet reveals that collaborative school culture is effective in building cooperation, confidence, purpose and team spirit, leading to creativity and productivity in a school. Higher performance of teachers and students in schools is basically attributed to effective and strong cultures whereas negative cultures lead to poor performance . The existing culture in a school may either serve as a hindrance or boost towards achieving school results. According to Lomax , there is little chance for school improvement unless the issue of school culture is directly addressed. The development of a strong school culture leads to higher performance of the institution through increased participation of teachers and other stakeholders in the school affairs . --- Literature Review According to Edelstein , factors that reflect the school's culture like goal achievement, shared ideology or mission, cohesion and collaboration among teachers, tend to increase job performance and effectiveness of teachers. Related studies by Clark and Mills reveal nine features of usually effective schools, in which school culture is mentioned first. According to Clark and Mills, an effective school is characterized by an orderly environment with interpersonal relationships, discipline, collaboration, consensus, participative approach to decision making and all these cultures enhance commitment of staff members so that they perform their duties with due diligence. School culture formation makes all members of the school to get involved, making them able to identify and agree on the values, beliefs, ambitions, intentions and purposes of the school and then individually or in small groups develop key words to encapsulate these values that constitute the school culture . According to Hopking et al , effective schools exist within a climate of supportive culture. In a related study, Glasser asserts that cultural conditions like change management and goal achievement are necessary for the growth and expression of employee potential. According to Deal and Kennedy , school cultures can improve educational productivity. Culture provides an internal cohesion that makes it easy for teachers to teach, students to learn and for parents, administrators and the rest of the school community to contribute to instructional processes. School culture determines how the school needs to function to become effective. School culture impacts teachers' engagement by influencing the kind of attitude students demonstrate towards schooling and participation in school activities . Kruse and Louis further add that the culture of a school plays a significant role in fostering its outcomes, including teacher effectiveness. Some studies have revealed that the relationship between many cultural attributes and high job performance has not been consistent over time . According to Burke , changes in school culture are either positive or negative and these changes in culture affect every aspect of the school. Further studies concur with Burke since norms guide attitudes and behavior of members of a school and they act as strategies to bring about change and improve productivity of the school . Deal asserts that school reform policies are only successful when tied to school culture. Sarason further noted that if change is to improve the climate and outcomes of schooling for both students and teachers, then some features of the school culture need to be changed or else, the well-intentioned efforts get futile. Ghanney, Antwi and Humu conducted a study in Ga south municipality, Ghana to investigate the effect of school culture on teachers' job performance in private and public basic schools. Ghanney, Antwi and Humu collected quantitative data from 46 teachers using census sampling. The study findings revealed that school culture was a good predictor of teacher job performance. The study further revealed that adaptability culture significantly contributed to teacher job performance whereas involvement culture that involves team orientation did not significantly contribute to teacher job performance. Some scholars have revealed that establishment of a culture of professional learning communities is a good strategy for improving schools and helping students to learn at high levels. The professional learning communities help teachers collaborate and build social relations to discuss their profession to enable them to perform their fundamental duty of getting committed to the learning of every student . DuFour et al further explains that professional learning communities have a collaborative culture made of collaborative teams working interdependently to achieve common goals of impacting their classroom practices, hence better outcomes for learners, staff and the school as a whole. Some studies on effective schools reveal that clarifying and reaffirming the goal of a school makes the school effective . The clarity of the goals among members makes them to be more effective by establishing priorities and guiding decisions . The school goal helps members to create a compelling, attractive and realistic future of what they wish the school to become, hence the staff is able to drive the school to where it hopes to be when the goal is shared . School values as a constituent of school culture, are collective commitments that answer the question "how must we behave to create a school that will achieve our purpose?" . DuFour argues that when members understand the purpose of their school, know where it is moving to and pledge to act in certain ways to move it in the right path, they do not need prescriptive rules and regulations to guide their daily work. According to Fullan , creation of a school community with shared commitment increases its effectiveness. MacNeil, Prater and Busch support the notion that schools with good cultures have highly motivated teachers who have greater success in terms of student performance and outcomes. In a related study, Omusonga, Kazadi and Indoshi revealed that there was a strong correlation between school culture and students' performance in French language as a result of teacher participation in French co-curricular activities such as observance of French days, music and drama festivals. Many teachers in unfavorable school cultures lack commitment to their social systems and this inhibits teacher collaboration, motivation, problem solving, community building and achievement. According to Kamaroellah and Mubarak , school culture has a significant influence on work satisfaction and motivation which then influence job performance. The best employee is really satisfied with his job and this employee will show his best performance . Nakanwagi conducted a study in Wakiso district to examine the influence of school culture on students' academic performance in private secondary schools. The study findings revealed that school culture in form of strategic direction sets standards and helps to galvanize the efforts of all staff members towards a common goal. Owens The school culture plays a significant role in defining for teachers, their commitment to task; it evokes the energy of teachers to perform their tasks; loyalty and commitment to the organization and its ideals. Sergiovanni revealed that in successful schools, culture serves as a compass setting to steer people in a common direction, provides a set of norms that define what people should accomplish using suitable techniques, provides source of meaning and significance for teachers, school heads, learners and others as they work. Schools with strong cultures have a "vision" of excellence while those with weak cultures lack understanding, determination and drive to accomplish assignments . Cheng carried out a study in Hong Kong to observe the cross-sectional relationship between school culture and organizational characteristics. The study findings revealed that schools with strong cultures are effective, characterized by higher teacher job commitment, higher teacher morale, all which culminate into higher teacher job performance. Schools with weak and ineffective cultures on the other hand, had low teacher job commitment, low teacher morale and low job performance. --- Methodology --- Research Design The study used a cross-sectional survey design. This study design was used because it typically involves collection of data at one point over a short period of time to provide a "snapshot" of the outcome and characteristics associated with the population . --- --- Data Collection Tools Self -administered questionnaires were used to collect data from study participants. With perceived school culture, a thirty-item Organizational Culture Assessment Questionnaire with a validity of 0.96 and reliability of 0.89 adapted from Sashkin and Rosenbach was employed. This was based on a five-point Likert scale. With teachers' job performance, a sixteenitem Teachers' Job Performance Scale with a validity of 0.849 and reliability of 0.87 adapted from Kigenyi & Kakuru was used. This was also based on a five-point Likert scale. --- Reliability and Validity Content validity was established through inter-judgement. Four judges rated the items on a two-point scale of Relevant and Not Relevant . The content validity index was calculated by summing up the items considered relevant divided by the total number of items in the questionnaire. Organizational Culture Assessment Questionnaire had a validity of 0.96 and Teachers' Job Performance Scale had a validity of 0.849 were used, which were all above 0.70 making the scale valid. The reliabilities of the instruments were determined using the Cronbach's Alpha provided by SPSS. Internal consistency of the instruments was determined by piloting the questionnaire to 10% of the study population for those who will not be part of the actual study. The questionnaire was administered twice to the same population and the two scores were correlated using Cronbach Alpha Coefficient. A Cronbach Alpha Coefficient of 0.89 was obtained for Organizational Culture Assessment Questionnaire and 0.87 was obtained for Teachers' Job Performance Scale . --- Data Collection Procedure An approval was sought from Mbarara University of Science and Technology Research Ethics Committee . Permission was sought from Municipal Inspector of Schools and the head teachers of the respective government aided secondary schools. Consent was then sought from the respondents and thereafter administered questionnaires. Data from the structured questionnaire was coded by assigning values to responses and entered into SPSS version 20.0. Percentages were employed to describe the socio-demographic data of teachers. Descriptive analysis was used to generate means for objective one and two. Pearson Correlation Coefficient was used to determine the association between perceived school culture and teachers' job performance. --- Results and Discussion Research question 1: What is the perceived school culture in government aided secondary schools in Sheema Municipality? Mean scores in reported school culture ranged between 2.70 and 4.80 with a mean of 3.98 . The largest percentage of teachers reported strong school culture and 22.6% reported moderate levels. No significant mean differences in school culture were found among teachers of different; age groups , education level , teaching experience , and gender . --- Research question 2: What is the level of teachers' job performance in government aided secondary schools in Sheema Municipality? Mean scores in reported teachers' job performance ranged between 2.50 and 5.00 with a mean of 4.17 . Majority of teachers reported high level of job performance and 8.8% reported moderate levels of job performance. In terms of demographic variations , teachers' job performance was independent of age , gender , and time spent in service . However, statistically significant differences were found between teachers of different academic qualifications . Teachers with post graduate diplomas and master's degrees reported higher performance levels than all other groups. H0 Hypothesis: There is no statistically significant relationship between perceived school culture and teachers' job performance in government aided secondary schools in Sheema Municipality Pearson Correlation Coefficient between perceived school culture and teachers job performance was significant . There was a weak positive correlation between perceived school culture and teachers' job performance . School culture explained only 8.4% of the variance in teachers' job performance . The null hypothesis which states that "There is no statistically significant relationship between perceived school culture and teachers' job performance in government aided secondary schools in Sheema Municipality" was therefore rejected. --- Pearson Correlation Coefficient between perceived school culture and teachers' job performance --- Perceived --- Discussions The study findings showed that strong cultures exist among government aided secondary schools in Sheema municipality especially coordinated teamwork and satisfaction of students' needs. The findings further reveal that there are high levels of job performance among teachers in government aided secondary schools in Sheema municipality. The study findings revealed that a significant relationship exists between school culture and teachers' job performance. The findings agree with a study conducted by Bukhuni and Iravo who conducted a study in Bungoma North sub-county, Kenya to establish the effect of school culture on employee performance in public secondary schools. Their findings revealed high levels of job performance in schools with strong and established cultures. It further revealed that there was a direct relationship between school culture attributes like teacher collaboration and job performance. Cheng conducted a study in Hong Kong to examine the cross-sectional relationship between school culture and organizational characteristics, including job performance. The findings revealed that schools with strong cultures were characterized by high levels of job commitment, teacher morale and job performance. It was further revealed that schools with weak cultures had low levels of job commitment, morale and job performance. Cheng's findings especially on the relationship between school culture and job performance are in consonance with the findings of the present study that revealed a positive significant relationship between perceived school culture and teachers' job performance. Ghanney, et al conducted a study in Ga south municipality, Ghana, to investigate the effect of school culture on teachers' job performance in private and public secondary schools. Ghanney, et al collected quantitative data from 46 teachers using census. The findings revealed that some dimensions of school culture like change management contributed to teacher job performance. This agrees with findings of the present study it comprised of dimensions of school culture like change management and also revealed a significant relationship between school culture and teachers' job performance. However, in the same study, Ghanney, et al revealed that teamwork did not significantly contribute to teachers' job performance thus disagreeing with the findings of this study. The findings further agree with Edelstein who asserted that factors that reflect the school's culture like goal achievement, shared mission cohesion and collaboration among teachers, tend to increase job performance. The findings also agree with Kamaroellah and Mubarak whose study revealed that school culture has a significant influence on work satisfaction and motivation which then influence job performance. The major study limitation was that the study only focused on a quantitative research approach and this narrowed the researcher's understanding of the relationship between school culture and teachers' job performance. The study dwelt on only teachers and therefore, generalizing the results to other staff should be done with caution. The study was carried out in only government aided secondary schools and the results may not easily be generalized for private secondary schools. The research findings will assist school head teachers to uphold strong and good cultures, abandon weak and poor cultures and strengthen weak cultures where necessary and possible to boost job performance among teachers for increased teacher efficiency and effectiveness. The study findings will help teachers to understand the vitality of a school culture and work towards maintaining and strengthening meaningful school cultures. For policy makers, the understanding of school culture may be of paramount importance in the education management process and the findings may help them to come up with better strategic policies that can promote positive school culture that may improve teachers' job performance. The aim of the study was to explore the relationship between perceived school culture and teachers' job performance in government aided secondary schools in Sheema municipality. The study findings revealed that school culture has a positive significant relationship with teachers' job performance, implying that strong cultures promote better performance of jobs by teachers and weak cultures negatively affect teachers' job performance levels. --- Conclusion and Recommendations --- Conclusion The study findings showed that strong cultures exist among government aided secondary schools in Sheema municipality especially coordinated teamwork and satisfaction of students' needs. The findings further reveal that there are high levels of job performance among teachers in government aided secondary schools in Sheema municipality and this differs from Nsubuga and the Sheema Municipality Statistical Abstract . School culture significantly relates to teachers' job performance, implying that strong cultures promote better performance of jobs by teachers and weak cultures negatively affect teachers' job performance levels. --- Recommendations 1. School administrators, Board of Governors and other stake holders should put in place strategies to maintain high job performance among teachers through practices like proper definition of goals and school mission and vision, delegation, use of good human resource management practices and motivation among others. 2. The Ministry of Education and Sports should come up with clear training programs for teachers so that they are taught about the vitality of upholding strong cultures in their schools in a bid to support higher job performance.

The purpose of the study was to explore the influence of perceived school culture on teachers' job performance in government aided secondary schools in Sheema municipality, Uganda. The study was quantitative and adopted a crosssectional survey research design. A self-administered questionnaire was used to collect data from 252 respondents using census strategy. Descriptive statistics were used to analyze objectives 1 and 2. Pearson Correlation Coefficient was used to test the relationship between perceived school culture and teachers' job performance. The study findings revealed that 77.4% of the teachers reported strong school culture and 22.6% reported moderate levels of school culture. It was revealed that 91.2% of teachers reported high levels of job performance while 8% reported moderate levels of job performance. The study established that positive significant relationship exists between perceived school culture and teachers' job performance (r = .29, p ˂ .01). The findings agree with studies conducted by Bukhuni and Iravo (2015), Edelstein (2012) and Kamaroellah and Mubarak (2019). The study findings, however, disagree with Ghanney et al (2017). The study recommended that the ministry of education should come up with clear training programs for teachers to teach them about the vitality of upholding strong school cultures in their schools to boost job performance.

Introduction Unmarried adolescent pregnancy is a major health and social problem in many developed as well as developing countries with unique medical and psychosocial consequences for the patient and society. Despite a high female literacy rate in Kerala, observational studies suggest that many unmarried adolescents seek abortion; majority of them report in the second trimester and few go for unsafe abortions leading to complications like septic abortion, future infertility, and even maternal death. As pregnancy among unmarried is a highly sensitive issue, a large proportion of these abortions go unreported resulting in paucity of data on this. Results of few studies available cannot be implied to Indian scenario due to cultural difference as premarital pregnancy is considered a taboo here. In Kerala, no analytic study is available exclusively on unmarried adolescent pregnancy and their risk factors. It is in this background, this study was undertaken to identify the factors leading to pregnancy among unmarried adolescents and young adults of Kerala. The study was designed as a case-control study. Sample size: With the smallest acceptable Odds ratio as 2, alpha error 5 %, beta error 10 %, with single control per case, 181 cases and 181 controls were recruited for the study. Consecutive sampling technique was adopted in the selection of cases and controls. Cases: Unmarried adolescents and young adults , attending family welfare clinic of the three major government postgraduate teaching hospitals of Kerala, i.e., Medical College Thiruvananthapuram, Kottayam and Kozhikode, seeking abortion services were the cases. Controls: Unmarried non-pregnant adolescents within the same age range, receiving services at the same study hospitals were the controls. Those who failed to give consent and who did not co-operate were excluded from the study. Data collection: Data were collected from October 2004 to July 2006 using a pre-piloted structured interviewer administered, closed ended questionnaire, and information was elicited on sociodemographic characteristics, familyrelated factors, and individual factors. Owing to the sensitive nature of the research, for deriving community relevant variables to be included in the study, focus group discussions were also conducted among doctors, nurses, teachers, and mothers of unmarried girls B24 years. Ethical issues: The study was started after obtaining Ethical clearance from the Institutional Review Board. Informed consent was sought from respondents; for minors, parent's consent was taken. Data analysis: The data were analyzed by means of SPSS software and Epi info 3.3 versions. Bivariate and multivariate analyses were performed and the risk factors for unmarried adolescent pregnancy determined. Subgroup analysis: Stratified by age were also done. Test of association was by Chi-square. Odds ratio and 95 % confidence interval computed as the estimate of relative risk. --- Results of Bivariate Analysis Distribution of abortion seekers according to age is shown in the Pie Chart . Unmarried girls ranging from age 13 to 24 reported for pregnancy termination. Out of the total 181 abortion seekers, 20 cases were below the age of 16; 70 cases were between 17 and 19 years; and rest of them, i.e. 91 cases were [19 years. --- Sociodemographic Characteristics As shown in Table 1, majority of cases and controls were Hindus, hailing from nuclear families and also from rural areas. No significant difference was observed between the study and control group with respect to religion, family type, and residence. However, the bivariate analysis showed significant association between unmarried pregnancy and low socio-economic status. Those with low socio-economic status were found to be having 4 times higher risk for unmarried pregnancy . --- Family-Related Factors The three main variables studied under family-related factors were family problem, parental control, and intrafamily relationship. --- Family Problem It is evident from Table 1 that 64.6 % of the cases hailed from problem families when compared to 14.4 % of the controls. In the present study, family problem was due to any of the following reasons: broken family /children from polygamous families /parents with marital disharmony/problematic stepfather relations/ extreme poverty/mentally challenged parents/very old and debilitated parents/single parent without family support, etc. The study pointed out that the chances for unmarried pregnancy was nearly 11 times more among adolescents from problem families . --- Parental Control Majority of cases had lack of appropriate parental supervision and control; it is found that those with no control or strict control by parents had 14 times higher risk for unwanted pregnancy and this was found to be statistically significant . --- Intra-family Relationship Association between intra-family relationship and adolescent pregnancy indicated that 74 % of the cases had poor intra-family relationship compared to 15.5 % of controls and showed nearly 15 times higher risk for unwed pregnancy . --- Individual Factors The important variables studied were education, occupation, and knowledge about sexual and reproductive health. --- Educational Level More than three-fourth of the cases had studied only up to 10th standard or less , and those girls with less number of school years were 4 times more likely to have an unwanted pregnancy . --- Productive Engagement Adolescents and young adults not engaging in any productive activity were more in the cases than in the control groups . They showed 2 times higher risk for unwanted pregnancy . --- Sexual and Reproductive Health Knowledge This variable included knowledge about menstruation, conception, sexually transmitted infections, contraception, and safe sex. It was found that lack of knowledge on sexual and reproductive health had a significant association with unmarried pregnancy. The poor knowledge about sexual and reproductive health was higher than that of controls . This difference was found to be statistically significant . --- Multivariate Analysis Logistic regression was used to find out the independent association between unmarried teenage pregnancy and the study variables. The model was constructed based on P \ 0.25 in the bivariate analysis and clinically important variables. --- Results of Logistic Regression Analysis In the final model, children hailing from family problems, lack of appropriate parental supervision and control, poor intra-family relationship, lack of engaging in any productive activity, and lack of knowledge about sexual and reproductive health were found to be significant independent risk factors for unmarried adolescent pregnancy . Subgroup analysis showed strength of association of all the above risk factors to unmarried pregnancy more among the B19-year age group, compared to[19. --- Discussion Comparison of observations with other studies: Even though very few studies are available in the literature to know about the risk factors for pregnancy among unmarried adolescents and young adults, these studies and their findings are not really comparable as they were poorly designed and no acceptable sampling techniques were employed. Association of lack of participation in any productive activity and teenage pregnancy had been observed by different authors. In the present study, 44.5 % of the cases were not engaged in any productive or academic activities and they engage in risky behavior at a very young age. The observations made in this study is consistent with the reports observed by Kasen et al. [1] and Kirby [2], who have conducted studies on the influence of school dropout and school disengagement on the risk of adolescent pregnancy. Family problem as an immediate determinant of adolescent sexual activity has been pointed out by various studies [3][4][5][6]. The present study also supports this and revealed that more pregnant adolescents hailed from family problems when compared with non-pregnant controls . It is also found that these children having family problems lacked love and encouragement by the family and they felt insecure at home. They received less support for their problems in or outside the family and had lesser life satisfaction and happiness in general. Parents play a significant role in the sexual development and behaviors of their children. Parent-child closeness or connectedness, parental control, and parent-child communication have all been implicated in adolescent sexual behavior. Parental monitoring and supervision are important ways for keeping adolescents from risky situations and activities while the teen develops responsible decisionmaking skills. Association of lack of appropriate parental supervision and control and adolescent sexual activity has been shown in different studies [7][8][9]. The present study is in agreement with this and confirmed that either lack of parental control or strict control by parents is a highly significant risk factor for adolescent pregnancy . Another important observation made out from the study is that a few percentages of the cases were not staying with their biological parents. They were residing in the hostels , staying with relatives, or working as housemaids, all of which resulted in lack of appropriate parental supervision and control. The current study showed that poor intra-family relationship is associated with unwed adolescent pregnancy. Several mechanisms may underlie the associations between family relationships and adolescent sexual activity. A supportive relationship between the parent and adolescent is important for enhancing communication and supervision [10]. A study by Guijarro et al. [11] on family risk factors associated with adolescent pregnancy has shown that mother-child communication about sex contributed to decreased likelihood of sexual risk. Family connectedness may be a protective factor related to sexual risk-taking, even among high-risk youth [12]. The present study also backs up the above observations and showed that almost three-fourth of the pregnant adolescents reported poor intra-family relationship compared to controls and this finding was found to be statistically significant . These pregnant adolescents with poor intra-family relationships were not in the habit of openly communicating or discussing their problems with parents or other members of the family, and their parents in turn never discussed family life issues with them. Non-pregnant adolescents showed better parent-daughter communication, higher levels of cohesion, connectedness, and higher future expectations. The study also showed a close link between lack of knowledge on sexual and reproductive health and adolescent pregnancy. Overall knowledge about human sexuality, reproduction, and contraception was poor among the abortion seekers compared to controls . Majority of the girls did not have clear information on the consequence of sexual relationship . The observations made in this study supports the findings of Dr Anuragini Sharma and Joseph GA who have conducted cross sectional studies on Sexual Knowledge and Practices of school and college students in India [13,14]. The present study also revealed that ignorance, myths, and misconceptions concerning sexual matters were prevailing more among the pregnant adolescents. Around One-third of the abortion seekers were unaware that it is possible for a girl to get pregnant the first time she has sex and about half of them believed that sexual activity after marriage only would result in pregnancy. Another observation was that 54 % of the cases did not suspect pregnancy for them as they had irregular menstrual cycles. Hence, there was delay in seeking abortion services as evidenced by the fact that out of the total 181 pregnant adolescents, only 36.5 % reported in the first trimester, more than half reported in the second trimester, and the rest progressed to third trimester. Delay in seeking abortion services was largely the result of unawareness of their pregnancy, fear of revealing the matter to family members, and due to social stigma. Contraceptive knowledge was found to be low. The reasons for non-use of contraceptives were the unplanned nature of the sexual act and no prior discussion between the partners regarding contraception. Even though ''low level of education of subjects'' and ''poor socio-economic status of parents'' were found as significantly associated with unmarried adolescent pregnancy in bivariate analysis, they were not found to be statistically significant in multivariate analysis. The major policy implication of this study is the finding that girls with teenage pregnancy had 5 times less knowledge about sexual and reproductive health, highlighting the importance of introducing Adolescent health/Family life education in schools. --- Limitations of the Study 1. As the study was conducted in Government hospitals, there may be a possibility of exclusion of women from high socio-economic status. 2. ''Peer group influence'' was not studied as a risk factor as both cases and controls viewed their friends as being ''good'' and there was no independent way to find out if their influence was actually compounding the problem. 3. Even though ''childhood sexual abuse'' was included as a study variable, only very few subjects gave a positive history, hence was not considered for analysis. --- Conclusion This study suggests that family-related matters, namely family problem, poor intra-family relationship, and lack of appropriate parental supervision and control have independent association with unmarried adolescent pregnancy. The study also points out that lack of engaging in any productive activity and lack of knowledge about sexual and reproductive health have significant roles leading to unmarried pregnancy. The number of adolescent pregnancies can be reduced by introducing Adolescent Health Education programs through educational institutions and for out of school adolescents through Anganwadis under ICDS scheme and Non-Governmental Organizations. Family counseling centers should also be organized at Taluk and District hospitals under National Rural Health Mission to guide parents in improving intra-family relationship, appropriate parental supervision in managing adolescents, and to give correct information on sexual and reproductive health issues.

To determine the risk factors for pregnancy among unmarried adolescents and young adults. Method(s) Case-control study was done over a period of 2 years. Data collected from 181 unmarried abortion seekers and 181 unmarried non-pregnant controls (B24 years) attending three Medical Colleges of Kerala. Results Logistic regression analysis showed a strong association between unmarried adolescent pregnancy and lack of parental supervision and control (OR 8.74, P = 0.000), poor intra-family relationship (OR 7.01, P = 0.000), family problem (OR 4.41, P = 0.000), lack of knowledge on sexual and reproductive health (OR 4.95, P = 0.0003), and nonengagement of adolescent in any productive activity (OR 4.41, P = 0.0373).Lack of parental control, family problem, poor intra-family relationship, lack of knowledge on sexual and reproductive health, and lack of engaging in any productive activity were found to be significant predictors for unmarried adolescent pregnancy.

InTRoduCTIon The UK is in a period of superdiversity that is characterised by 'an increased number of new, small and scattered, multiple-origin, transnationally connected, socio-economically differentiated and legally stratified immigrants'. 1 This presents challenges for the delivery and configuration of maternity services in achieving equality of provision which forms a key aim of the National Health Service in the UK. 2 One in four births in the UK is to foreign-born women. 3 Indeed some immigrant women appear disproportionately in confidential inquiries into maternal and perinatal mortality, 4 perhaps indicating possible deficits in the delivery of care, access and utilisation. Our review contributes to amelioration of this situation by synthesising knowledge related to maternity care access and interventions so as to configure appropriate interventions as identified per the NHS Midwifery 2020 vision to guide professional development of healthcare professionals . 5 Reshaping care to ensure culturally safe and congruent maternity care that will not only benefit immigrant women but also improve the health of future generations in the UK. 3 4 6 Without the delivery of culturally appropriate and culturally safe maternal care, negative event trajectories may occur that range from simple miscommunications to life-threatening incidents, [7][8][9] risking increased maternal and --- ConCepTuAl defInITIonS There is no consensus definition in the UK regarding the definition of the term immigrant 20 with the terms immigrant and migrant which are frequently used interchangeably across different data sources and datasets whilst conveying the same meaning. Country of birth is used by The Annual Population Survey of workers and Labour Force Survey as a precursor for defining a 'migrant'. This survey therefore declares a person born outside the UK is classified as a 'migrant'. Noteworthy is the fact that workers born outside the UK may become British citizens with increasing residence in the UK. A second source of data on migrants is applications made to obtain a National Insurance Number. This differs from the former in that the term migrant is conferred on the basis of nationality. All applicants who hold nationality other than the UK are therefore considered migrants. However, the situation is dynamic in that the nationality of a person may also to change over time and in some cases individuals may acquire dual citizenship involving several nation states. A third and significant source of data on migrants is the Office for National Statistics . ONS utilises a different strategy classification which focuses on the notion of short-term international migrant and long-term international migrant. In this definition, the term 'long-term' refers to holding the intention of residing longer than a year, whereas short-term is intention of residing less than a year. The implication of this is that the ONS considers length of stay of a person in the UK as critical in determining migrant status which reflects the United Nations recommended classification of migrant into short and long term. Additionally, ONS utilises the UN definition of long-term international migrant. Accordingly, 'a migrant is someone who changes his or her country of usual residence for a period of at least a year, so that the country of destination effectively becomes the country of usual residence'. 20 In long-term international migration data, students and asylum-seekers are also included which differs for example from the situation in the USA. --- Immigrants and the uK nHS In respect of service provision, the NHS adheres to the mandates set by central government that determines immigrant's entitlement to free NHS care. These mandates are concerned with the immigrant status and the type of service provision. 21 Within these mandates, an asylum-seeker woman may not be entitled to full maternity care because of immigration status. 22 Moreover, data collection by the NHS on this topic is not well established or comprehensive. Currently, the NHS usually collects data on ethnicity and nationality and not on migrationrelated variables such as length of stay, country of origin and so on. The National Institute for Health and Care Excellence which provides clinical guidelines for healthcare practice in the UK see NICE 23 identified recent migrant women as having complex social needs in its guidelines on Pregnancy and complex social factors: a model for service provision for pregnant women with complex social factors identified recent migrant women having complex social needs. Within the NICE definition, a recent migrant woman is a woman has who moved to the UK within the previous 12 months. This generic definition of the term migrant conflates migrant women of all classifications . This suggests that there is implicit acceptance of the term migrant women in healthcare in respect of being born outside the UK, and being subject to immigration regulations, together with possible challenges in English language proficiency. --- The operational definition of an immigrant women used in this review The preceding paragraphs suggests that the term 'immigrant' is defined in various ways in different countries and by different authors. However, two features are frequently referred to in these definitions, namely 'country of birth' and 'length of stay'. These factors are noted by the NICE guidelines 23 on the provision of maternity care as important in entitlement, access and ability to use healthcare in the UK. For example, if you are born outside the UK, it is unlikely that you are knowledgeable about the UK healthcare provision. We adopted the following definition of an immigrant woman for the purposes of our review, and most importantly, to inform our inclusion and exclusion criteria. We defined a woman as an immigrant if she was: ► Born outside the UK. ► Living in the UK for more than 12 months or had the intention to live in the UK for 12 months when first entered. We therefore included studies on immigrant women where the population studied fulfils these two characteristics and included population groups of foreign students, asylum-seekers, recent legal refugees and immigrants, and illegal immigrants. In cases where the study populations/sample was not accurately or fully described, we employed the criteria of linguistic ability, as demonstrated by the need for an interpreter as a proxy for immigrant status. Notwithstanding all of these perspectives, we acknowledge that the term 'immigrant women' is generic and refers to a highly heterogeneous group of individuals with a complex and vast array of ethnocultural groups. --- Aim and rationale We consider in this paper how accessibility and acceptability manifest, as important dimensions of access to maternity care services in terms of women's perception about availability of services and their experiences of accessing these services. We also consider whether evaluated interventions exist that challenge inequalities in maternity healthcare provision. Our review employed two theoretical frameworks. These are Gulliford and colleagues' theory of access and second the concept of cultural safety. A theory of access to services developed by Gulliford et al 12 map out four dimensions : 1. Service availability. 2. Utilisation of services and barriers to access . 3. Relevance, effectiveness and access. 4. Equity and access. We used this theoretical model in our systematic review which was based on a synthesis project funded by the National Institute for Health Research. Unlike most access models in the USA, this framework reflects the philosophy of the NHS in that its key principles are to provide horizontal access in terms of ensuring equality of access in the population and to achieve vertical access in terms of meeting the needs of particular groups in the population, such as minority ethnic groups. The application of these principles is influenced by availability, accessibility and acceptability. The Gulliford model 12 has been widely used in empirical research, with the main paper cited over 730 times. This model with its emphasis on accessibility, acceptability, relevance and effectiveness, is entirely appropriate for assessing the provision of maternity services to minority ethnic groups and was employed in this review to assist in initial theme development and to examine how this access model intersected with our evidence. Second, concepts of cultural safety provided a theoretical lens for the production of recommendations. Cultural safety is a theory that aims to assist the understanding of deficits in care by considering the historical and social processes that impact power relationships within and beyond healthcare. 24 Cultural safety is achieved when programmes, instruments, procedures, methods and actions are implemented in ways that do not harm any members of the culture or ethnocultural group who are the recipients of care. Those within the culture are best placed to know what is or is not safe for their culture which suggests the need for increased dialogue about immigrant and partner approaches. [25][26][27][28][29] --- MeTHodS We employed Popay's approach to Narrative Synthesis 30 which consists of four elements (for a comprehensive explanation please see our published protocol. 31 The unique feature of this approach is that it provides highly specified steps. Team members have successfully employed NS previously and have vast expertise in its usage. 7 ► Element 1: Developing a theory of why and for whom. ► Element 2: Developing a preliminary synthesis of the findings of the included studies, following implementation of the search strategy. ► Element 3: Exploring relationships in the data. ► Element 4: Assessing the robustness of the synthesis. TheNS approach relies primarily on text to summarise the findings and produce asynthesis of the narrative findings of included papers. NS may be used with all paradigms of research quantitative, qualitative studies and mixed methods research studies, as the emphasis is on an interpretive synthesis of the narrative findings of research rather than on a metadata analysis. 30 --- Search strategy refinement and implementation The search strategy employed key terms used in consistently formulated text-based queries and search statements. These terms were based on subject headings, thesaurus terms or related indexing and categorisation terms appropriate for each literature database. An example of a detailed final search strategy is given in online supplementary file 1. First, we searched 10 electronic databases using the aforementioned strategies . Following this, we searched for appropriate grey literature in SI Web of Knowledge Conference Proceedings Citation Index , Dissertations and Theses, and the Cochrane Methodology Register. We also searched using Google and Google Scholar and consulted with the study expert advisory group. In conclusion, we hand searched the reference list of all included studies and relevant systematic reviews. Citations were downloaded into an ENDNOTE library, and following this all duplicates removed. The bibliographic databases that we searched are listed in box 1. We adopted the PICO approach to implement the search strategy as follows: P=immigrant women I=maternity care C=non-immigrant women-implicit comparator emerging in the results O=experience of care Our search strategy development was therefore based on: Search concept 1=pregnancy, childbirth , explicit terms covering women/females requiring all types of maternity care . Search concept 2=immigrant populations . Search concept 3=terms used to identify access to, use of, deficiencies in and so on, service provision This comprehensive search strategy generated high rates of retrieval of records, however many were not pertinent. --- Screening for relevance In many cases, the study populations/sample was not fully described. In this situation, we contacted the authors for further clarification and in some cases used linguistic ability, for example, the need for an interpreter as a proxy for immigrant status. Our focus was on first-generation immigrant women regardless of their phenotype which led to inclusion of women of white ethnicities, although we encountered few studies that focusing on these groups. Study screening was undertaken independently by two team members who employed our screening tool to assess the relevance of titles and abstracts in respect of our screening tool. The entire team reviewed papers classified as ambiguous papers in order to achieve a consensus agreement and where necessary full text papers of potentially included studies were retrieved and appraised. The exclusion and inclusion criteria can be found in online supplementary file 3. When we retrieved full-text papers which were later rejected, we have documented these excluded papers and presented a rationale for exclusion. These can be found in online supplementary file 4. --- ReSulTS Studies included in the review, findings and evidence Our systematic review identified 40 empirical research studies in the scientific and grey literature. The included studies embraced a broad range of ethnocultural groups and methodological genres . The search outcomes are comprehensively detailed in figure 2, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart. 32 The distribution of the studies across the themes are shown in figure 3 and publication dates in figure 4. --- data extraction and assessment of relevance We conducted the following foundational activities in order to extract data . 1) Textual description. A systematic textual narrative was written for each study. We used headings adapted from Popay et al Setting, Participants, Aim, Sampling and Recruitment, Method, Analysis, Results. 30 Tabulation and summarisation of all studies to be included. These tables described the attributes of the studies and the results. Information was extracted from the textual description using the same headings as above and additional headings as necessary. Papers in the PDF format were imported into ATLAS. ti qualitative data analysis software using the 'Attributes' option to allow the tabulation of relevant data. --- Quality assessment In element 4, we conducted the quality appraisal . 33 All included studies were critically To determine the nature of the barriers confronting women when they used antenatal and postnatal services. Pollokshields, Glasgow Qualitative: semistructured questionnaire. Not specified. Not specified. Qualitative: thematic analysis. Twenty women were interviewed in depth by a Centre's Health Development Worker. Of these, 17 were born outside the UK. --- 47 To determine the current clinical practice of maternity care in England, including the service provision and organisations that underpin care, from the perspective of women needing the care; to identify the key areas of concern for women receiving maternity care in England; and to determine whether and in what ways women's experiences and perceptions of care have changed over the last 10 years. --- England: not specified Quantitative: survey. Not specified. Survey: not specified. Quantitative: crosssectional design. Random samples of women selected for the pilot and main studies were identified by staff at the ONS using live birth registrations for births within two specific weeks: 2-8 January and 4-10 March 2006 . The same method of sampling was used as had been employed in 1995 to enable direct comparisons. Random samples of 400 women for the pilot survey and 4800 women for the main survey who were aged 16 years and over and who had delivered their baby in a 1-week period in England were selected. The sampling was stratified on the basis of births in different geographical areas . No subgroups were oversampled. The usable response rate was 60% for the pilot survey and 63% for the main survey. The samples included 229 women of BME born outside the UK. --- 48 To explore the healthcare experience of vulnerable pregnant migrant women. --- London Mixed methods: participants were contacted by phone and interviewed using a pro forma questionnaire designed to determine their access to antenatal care; barriers to that access; and their experiences during pregnancy, labour and the immediate postnatal period. Further data was extracted from their records at the Doctors of the World clinic to see how they had accessed the clinic. Not specified. Phone survey. Qualitative: thematic analysis. Quantitative: not clear. Pregnant women who presented to the drop-in clinic of the DOTW in London were approached between January 2013 and June 2014. To study the maternity care experiences of Somali refugee women in an area of West London. This article focused particularly on findings relating to the language barrier, which to a large degree underpinned or at least aggravated other problems the women experienced. West London Qualitative: case study. Six semistructured interviews and two focus groups . Not specified. Not specified. Qualitative: thematic analysis. Snowball sampling: 12 Somali women were selected from a larger survey involving 1400 women. Theory of interactions and transformational educational theory. Hospital settings or women's homes. Qualitative: thematic analysis. Women were approached if the status of 'asylumseeker' or 'refugee' was written in the hospital notes taken at their booking appointment. Fourteen women were approached, but nine women declined to participate. Five women consented, but one woman was dispersed before 20 weeks gestation and therefore was not included in the study. Of the remaining four participating women, three were asylum-seekers and one was a refugee. To address the research question that postulates that immigrant women experience sensitive care through the use of an ethnically congruent interpreter and that such women prefer to meet health providers of the same ethnic and gender profile when in a multiethnic obstetrics care setting. Greater London Qualitative: in-depth individual and focus group interviews. Open-ended questions were presented by an obstetrician and an anthropologist. Framework of naturalistic enquiry. Not specified. Qualitative: naturalistic inquiry. Participants were recruited throughout Greater London between 2005 and 2006. Snowball sampling was used to recruit 36 immigrant Somali women, and another three were selected by a by purposive technique for a total of 39. A purposive technique was used to select further 11 Ghanaian women who had delivered at least one child within the British healthcare system and who were living within the study area at the time of data collection. --- 61 To study the relationships between Somali women and their Western obstetric care providers. The women interviewed came from 14 different countries and had been dispersed or relocated to or within six regions of the UK. At the time of dispersal, 14 had been awaiting a decision on their asylum claim and six had been refused asylum. PROTOS is a maternity care recording system. HCP, healthcare professional. --- Table 1 Continued appraised by two reviewers using tools from the Center for Evidence-Based Management . 34 We used Good Reporting of A Mixed Method study 35 for the mixed-methods studies. Differences were resolved in our reflective team meetings. We also used high, medium and low as appraisal categories This is approach is congruent with recent publications from the Cochrane Qualitative Research Group's Confidence in the Evidence from Reviews of Qualitative research publications and was previously used by in published studies by Higginbottom and colleagues. 7 9 Studies were classified in three into domains, high, medium and low, to enable a 'macro' evaluation. ► High was assigned to studies that used a rigorous and robust scientific approach that largely met all CEBMa benchmarks, perhaps equal to or exceeding 7 out of 10 for qualitative studies, 9 out of 12 for cross-sectional surveys or 5 out of 6 for mixed-methods research. ► Medium was assigned to studies that had some flaws but that did not seriously undermine the quality and scientific value of the research conducted, perhaps scoring 5 or 6 out of 10 for qualitative studies, 6 to 8 out of 12 for cross-sectional surveys or 4 out of 6 for mixed-methods research. ► Low was assigned to studies that had serious or fatal flaws and poor scientific value and scored below the numbers of benchmarks listed above for mediumlevel appraisals in each type of research. The past decade has witnessed a growth in approaches to assessing quality and Popay et al 30 recommends evaluating not only the scientific quality of studies but also the 'richness' of studies, defined as 'the extent to which study findings provide in-depth explanatory insights that are transferable to other settings' 30 'Thick' papers create or draw on theory to provide in-depth explanatory insights that can potentially be transferable to other contexts. By contrast, 'thin' papers provide a limited or superficial description and offer little opportunity for generalising. Each paper was assessed against the criteria as set out in Higginbottom et al 28 33 and categorised as either 'thick' or 'thin' . --- Analysis and synthesis Following construction of the preliminary themes, we produced code/narrative theme tables to demonstrate how the basic meaning units related to the theme. This involved utilising the codes produced in ATLAS. ti and aligning these to the manually extracted key findings . We reviewed all these processes in our reflective team meetings to ensure the rigour and robustness of our analytical steps. This iterative process is similar to the process of qualitative research and involved grouping the narrative findings into meaning units and social processes as they manifested in the maternity care experiences of immigrant women. Individual team members engaged in independent theming of tabular and coded data. We subsequently merged these individual perspectives to form the final harmonised themes representing a 'meta-inference' which is a term used in mixed methods research to describe merging of findings from the positivistic and the interpretative paradigms. Tashakorri and Teddlie 36 describe meta-inference as 'an overall conclusion, explanation of understanding developed from the integration of inferences obtained from the qualitative and quantitative strands'. Following construction of the preliminary themes, we produced code/narrative theme tables to demonstrate how the basic meaning units related to the theme. Utilising the codes produced in ATLAS. ti and aligning these to the manually extracted key findings . During the analytical processes we interrogated the data identifying using the concept suggested by Roper and Shapira. 37 We have constructed the themes in a policy directive fashion in terms of containing implicit indications in order to provide tangible guidance for policy and practice that might be developed into relevant strategies that benefit immigrant women and the NHS. Rigour, reflexivity and the quality of the synthesis Reflexivity in the review process requires a self-conscious and explicit acknowledgement of the impact of the researcher on the research processes, interpretations and research products. Reflexivity therefore demands acknowledgement of inherent power dimensions, hierarchies and prevailing ideologies that might shape and determine interpretations and the consequent knowledge production and research products. Gender, sexuality, professional socialisation, ethnocultural orientation and political lenses as these impact on social identities further coalescing to provide a specific perspective on any given phenomena. The review team members are imbued with a strong personal and professional commitment to the eradication of inequalities and allegiance to contemporary equality and diversity agendas. From a reflexive perspective, this is important given that immigration is global phenomena and the inherent vulnerability of some immigrant women. --- Open access Figure 3 The total numbers of studies involved in each theme. Figure 4 The range of publication dates for the included studies . Reflexive analysis alerts us as researchers to emergent themes and informs the formal and systematic process of analysis, with reflexivity defined as: sensitivity to the ways in which the researcher's presence in the research setting has contributed to the data collected and their own a priori assumptions have shaped the data analysis 38 Our collaborative decisions required constant review and reading and, in some cases, reviewing the theme allocation and evidence to reach consensus. Therefore, we believe we achieved a nuanced and comprehensive approach. Higginbottom et al have successfully employed this review genre previously and have vast expertise in its usage. 39 Within the published NS reviews, we have not given great attention to the issue of publication bias. However, we strove to eradicate any potential bias by undertaking a comprehensive and exhaustive literature review that included grey literature and follow-up emails with authors seeking greater clarity and explanation of opaque issues. A number of the included research studies were identified via ProQuest and E-theses and do not appear as publications in peer reviewed scientific journals. We also held a national stakeholder event during which we presented our preliminary findings to a wide range of health professions , academics, voluntary and community workers. Possibly this approach may be considered contentious in the respect of systematic review, as attendees had no previous knowledge of the original included papers although they held deep topic knowledge. Notwithstanding this, we found broad support for our findings and facilitated groups work activities in order to --- Open access Table 2 Thick and thin criteria Higginbottom et al 33 --- Richness Operational definition --- Thick papers ► Offer greater explanatory insights into the outcome of interest. ► Provide a clear account of the process by which the findings were produced-including the sample, its selection and its size, with any limitations or bias noted-along with clear methods of analysis. ► Present a developed and plausible interpretation of the analysis based on the data presented. Thin papers ► Offer only limited insights. ► Lack a clear account of the process by which the findings were produced. ► Present an underdeveloped and weak interpretation of the analysis based on the data presented. challenge our initial interpretations. These challenges resulted in the construction of Theme 5: Discrimination, racism, stereotyping, cultural sensitivity, inaction and cultural clash in maternity care for immigrant women. These focused activities collectively contribute to the confidence in the review findings, providing verification and validation of the themes. We identified 40 research studies that met our inclusion criteria, and we extracted and synthesised key findings into five themes for the publications informing each theme. --- Methodological genres --- Quantitative studies We identified eight quantitative studies that all used a questionnaire for data collection. [40][41][42][43][44][45][46][47] These populationbased studies and cohort surveys were all cross-sectional: none were longitudinal. --- Mixed-methods studies We identified 10 mixed-methods studies that employed both qualitative and quantitative dimensions. 2 48-56 For example, Duff et al 49 reported a two-stage psychometric study in which focus groups and interviews were used in the first stage to develop a questionnaire for an ethnocultural group In the second stage, quantitative methods were used to test and evaluate the acceptability, reliability and validity of the questionnaire. Other mixedmethods designs included interviewing a small sample of the participants after collecting data from a large-scale survey; conducting semistructured interviews with a small sample of participants based on quantitative data routinely collected from a large group of participants; and using face-to-face, postal and online questionnaires to collect data. One of the studies used Q methodology which uses questionnaires with structured and unstructured questions. --- Qualitative studies Of the 40 studies included in this review, we identified 22 as qualitative research studies employing a range of qualitative methodologies and approaches. 17 57-77 However, many of these studies did not specify a qualitative methodological genre but instead employed a more generic qualitative approach and described only the data collection tools used. For example, some presented multiple longitudinal case studies of participants about their maternity care experiences that included photographs taken by the participants, field notes and observations in addition to researcher interviews. Another example was a case study of an ethnocultural group, immigrant women of Somali origin, that used semistructured interviews and focus groups. Some studies used focus groups and interviews conducted in the language of the population group; for example, Bengali, Sylheti, Urdu and Arabic. Others used in-depth interviews, open-ended questions, group storysharing sessions and individual biographical life-narrative interviews. In contrast, a few studies specified a qualitative interpretive approach that used hermeneutic phenomenology and focused ethnography. --- Studies focusing on specific ethnocultural groups The chosen studies included participants from a wide range of ethnocultural groups that originated in diverse countries in different continents, including Asia , Africa and Europe . In some cases, the sample was drawn from a single ethnocultural group, such as Pakistani. 72 However, most of the studies were undertaken with mixed samples of immigrant women originating from different countries . --- Studies focusing on immigrant women without a clearly specified ethnocultural group We identified 16 studies that used the term immigrant women generically and not clearly specify an ethnocultural group. In deciding to include these studies, we believed that legitimate proxies for immigrant status could be the specified use of an interpreter or the participants having countries of origin or birth outside the UK. Some studies reported immigrant women arriving from 14 different countries but did not specify the country of Theme 1: access and utilisation of maternity care services by immigrant women Late booking emerged as an important dimension in this theme with immigrant women study participants often booking and accessing antenatal care later than the recommended timeframe of during the first 10 weeks of pregnancy. This delayed utilisation was found to be multifactorial in nature with influencing factors including the effects of limited English language proficiency, immigration status, lack of awareness of the services, lack of understanding of the purpose of the services, income barriers, the presence of female genital mutilation , factors associated with differences between the maternity care systems of their countries of origin and the UK, arrival in the UK late in the pregnancy, frequent relocations after arrival, the poor reputations of antenatal services in specific communities and perceptions of regarding antenatal care as a facet of medicalisation of childbirth. The range of factors affecting the access and utilisation of postnatal services were similar to those reported for antenatal services. Theme 2: maternity care relationships between immigrant women and HCPs Our included studies identified the perception of service users in this group and their interactions and therapeutic encounters with HCPs as significant in understanding access, utilisation, outcomes and the quality of their maternity care experience. Included studies identified both positive and negative perceptions of study participants regarding the ways HCPs delivered maternity care services were both positive and negative. A number of studies illustrated positive relationships between HCPs and immigrant women with the HCPs described as caring, respecting confidentiality and communicating openly in meeting their medical as well as emotional, psychological and social needs. Conversely, some studies provided evidence of negative relationships between participants and HCPs, with HCPs described from the perspective of immigrant women as being rude, discriminatory or insensitive to the cultural and social needs of the women. The end result of these negative encounters was that these women tended to avoid accessing utilising maternity care services consistently. --- Open access Theme 3: communication challenges experienced by immigrant women in maternity care It is axiomatic that limited English language fluency presents verbal communication challenges between HCP and their patients, families and carers. Moreover, this is compounded when HCPs use complex medical or professional language that is difficult to comprehend. Nonverbal communication is culturally defined and challenges can occur through misunderstandings of facial expressions, gestures or pictorial representations. Poor communications result as illustrated in our included studies in limited awareness of available services in addition to miscommunication with HCPs. Study participants often expressed challenges in accessing services, failed to understand procedures and their outcomes and were constrained in their ability to articulate their health or maternity needs to healthcare providers and disempowered in respect of their involvement in decision-making. They therefore sometimes gave consent for clinical procedures without fully comprehending the risks and benefits, and did not always understand advice on baby care. Studies also identified communication as not reciprocal with HCPs often misunderstanding participants. These issues of communication were described as leading to feelings of isolation, fear and a perception of being ignored. Theme 4: organisation and legal entitlements and their impacts on the maternity care experiences of immigrant women The study participants in our included studies had mixed experiences with the maternity care services in the UK. Positive and commendable experiences included feeling safe in giving birth at hospital rather than at home, being able to register a complaint if poor healthcare was received, being close to a hospital facility, not being denied access to a maternity service, and having good experiences with postnatal care. Conversely, negative experiences included lack of continuity and being unaware of the configuration of maternity services work that limited appropriate use. Participants in our included studies found services bureaucratic and perceived within the UK maternity care model as having a propensity towards medical/obstetric intervention and lower segment caesarean section births. The legal status of an immigrant women in the UK has a profound influence on their on their access to maternity care. Women without entitlement to free maternity care services in the UK were deterred from accessing timely antenatal care by the costs and by the confidentiality of their legal status. Moreover, some women arrived in the UK during the final phase of their pregnancies that resulted in interruptions in the care process, loss of their social networks, reduced control over their lives, increased mental stress and increased vulnerability to domestic violence. Positive experiences included receiving information from their midwives on the benefits of breastfeeding together with demonstrations on how to position the baby. Negative experiences included poor support from hospital staff on how to breastfeed their babies consequently these reported experiences are mixed. Theme 5: cultural sensitivity, inaction and cultural clash in maternity care for immigrant women Inequalities in access, navigation, utilisation and the subsequent maternity care outcomes are influenced by discrimination and cultural insensitivity in maternity care services according to the perspectives of women in several included studies. Although discrimination is often subtle and difficult to identify, direct and overt discrimination was reported in some studies. Specifically, study participants of Muslim faiths challenged assumptions held by HCPs, including those held regarding Muslim food practices and that their partners or husbands should help the women during labour. Moreover, HCPs were reported in some studies to lack cultural sensitivity and cultural understanding. For example, these women did not optimally benefit from antenatal classes facilitated by a non-Muslim educator who had no knowledge of the relationships of Muslim culture to maternity. Furthermore,Muslim participants often expressed dissatisfaction with antenatal classeshaving a gender mix, which contravened religious edicts. Studies illustrated that some women of Muslim faith also regarded their cultural and religious needs were not met, and they felt that the staff lacked insight, knowledge and understanding of FGM. Evidence from our included studies suggests some immigrant women perceived that the staff did not treat them with respect or attended fully to their healthcare needs, and they felt devalued, unsupported and fearful while receiving maternity care. Our findings also identified instances of cultural clash and conflicting advice during pregnancy and maternity care, mostly resulting from differences between the cultural practices and medical systems of the home countries of the immigrant women and those in the UK. In a few cases, however, midwives were happy to meet the cultural and religious needs of the study participants in our included studies in both antenatal and postnatal settings which is a positive finding. We conceptualise the findings graphically in figure 5. --- patient and public involvement The systematic review questions were developed in consultation with our project advisory group including service users' priorities experience and preferences. This systematic review did not include empirical research therefore there were no human participants. --- dISCuSSIon And ConCluSIonS The UK is in a period of superdiversity, defined as being 'distinguished by a dynamic interplay of variables among an increased number of new, small and scattered, multiple-origin, transnationally connected, Open access socio-economically differentiated and legally stratified immigrants'. . 1 Responding to this level of diversity is challenging for UK maternity care health services and may require the development of new and innovative strategies. The experiences of immigrant women in accessing, navigating and utilising maternity care services in the UK are both positive and negative. In order to enhance services, it is essential that strategies are developed to overcome the negative experiences reported. The experience of maternity care services is multifactorial in nature with a number of issues appearing to coalesce to determine the poorer experience reported by some immigrant women. Important factors identified by the review included a lack of language support, cultural insensitivity, discrimination, poor relationships between immigrant women and HCPs, and a lack of legal entitlements and guidelines on the provision of welfare support and maternity care to immigrants. Implications of findings and recommendations for maternity care policy, practice and service delivery Inequitable access appeared to be a consequence of the immigration and legal status of asylum-seeking women which has a profound impact on healthcare experiences and consequently health, and was also influenced by language fluency. We concluded that addressing language barriers and ensuring culturally sensitive care are essential elements of providing optimal maternal care for immigrant women. The issue of confidentiality may be compromised by having known interpreters in small communities. One solution may be the setting up of a national-level website offering standard information on maternity care and the option of translation in a wide range of languages. Additionally, the identification of best language practices should be identified in order to improve the current language service model. The knowledge, understandings and attitudes of maternity care healthcare providers is a critical determinant of care. Ethno-culturally based stereotypes, racism, judgemental views and direct and indirect discrimination require eradication requires challenging discrimination and racism at all levels: individual, institutional, clinical and societal Interventions to improve maternity care for immigrant women are scant, and formal evaluations of these interventions were largely absent. Increasing the social capital available to immigrant, health literacy and advocacy resources may empower women to access and use maternity care services appropriately. Maternity care staff require a greater level of mandated education to have better cultural awareness of needs of diverse client groups including newcomers to the UK. Our findings highlight the importance of demonstrating compassion, empathy and warmth in their relationships with these women to reinforce positive attitudes among immigrant women. It is contingent on maternity care providers to value diversity among service users and to offer individualised and culturally congruent care. One way to achieve this goal would be through birth plans that can be jointly agreed and discussed in advance by the maternity care staff and recently arrived newcomers and immigrant women. Maternity care staff should seek to empower immigrant women by providing comprehensible information and better education concerning the configuration of the maternity system in the UK, conveying accurate information about care delivery. Central to these suggestions may be to enable volunteer and third-sector organisations to work as links between the statutory maternity services and immigrant women. We found evidence of such links in our national networking event. Representatives of immigration control agencies may feel obligated to adheres to immigrant rules and consider the maternity care needs of immigrant women's and baby's health as a secondary issue. The policy context regarding data protection and sharing information with the Home Office about the immigrant status of women was at issue as well, especially since variabilities have been seen in the policies for sharing this information. The results suggest that the legal and policy context is important in addressing the maternity care needs of immigrant women. It would seem imperative, as reflected in current policy directives, to adopt a universal of aim of achieving optimal maternity care for all and not just for immigrant women. However, maternity care services should strive to give more information to immigrant women about their rights to care, the availability and configuration of maternity services, and how to navigate maternity care systems. The child in utero of an immigrant is a future UK citizen and optimising maternity care is a dimension of securing the future health of the nation. In a period of super diversity is incumbent on health professional to have an awareness of immigrant women's legal rights and perhaps education on this topic should be mandated for maternity HCPs. Continuity in maternity caregivers and compulsory provision of interpreters would also help to improve the experiences of these women. Decision-makers and healthcare leaders should address the findings at a strategic level. A focus on diversity, equality and the needs of immigrant women could reasonably be embedded in the role and responsibility of 'Board level Maternity Champion' and of 'Maternity Clinical Networks'. Maternity service providers could consider the appointment of one obstetrician and one midwife jointly responsible for championing maternity care provision to immigrant women in their organisation. As these dimensions feature within the 'Bespoke Maternity Safety Improvement Plan', 78 key areas of action include: ► Focus on learning and best practice: issues of equality and diversity should be featured in the Saving Babies' Lives care bundle for use by maternity commissioners and providers. ► Focus on multiprofessional team working: continuous personal and professional training. Open access ► Focus on data: greater focus on ethnicity and immigration within the Maternity Services Dataset and other key data sets. ► Focus on innovation: create space for accelerated improvement and innovation at local level. Gaps in the evidence Some locally developed and locally based interventions to address inequalities in access and quality in maternity care for immigrant women were described during the final feedback meeting. However, there are very few interventions to address these issues in the published literature and their effectiveness has not been evaluated robustly. None of the interventions had also included economic evaluation of the intervention. Studies of the usual 6 weeks postnatal checks by a general practitioner were not identified nor studies that focused on the intrapartum period. As mentioned earlier, we found few studies that focused on immigrant women with 'white ethnicity' in our review time period, for example, women of Eastern European origin. --- Strengths and limitations ► We were challenged and constrained by the lack of consistency in describing immigrant population sin the published literature. There exists a great deal of variation and no unified approach within the UK literature. ► Immigration is an international phenomenon, and this review increases understanding of how immigrant women navigate maternity services in the UK, ► The review systematically maps our positive and negative aspects of maternity care provision as experienced by immigrant. ► The review provides strategic policy-level direction for enhancement of maternity care services. ► The review does not address the experiences of maternity care for second-generation women nor does it consider refugee and asylum seeking women as a separate group. --- Implications for future research More research is required into how the term 'immigrant' is used, and the changes in its use over time that may affect immigrant women's care. At present, the term is used very broadly and simplistically which masks its inherent heterogeneity. Furthermore, more research is required to understand how the intersections of particular characteristics-such as gender, education status, time in the UK, immigration status, wealth and country of originmay influence or alter the experiences of these women in their maternity. Research is also required that focuses on developing and evaluating specific interventions to improve maternity care for immigrant women. Twitter Gina Marie Awoko Higginbottom @Awoko1 and Catrin Evans @Catrin_notts --- Open access

► Immigration is an international phenomenon, and this review increases understanding of how immigrant women navigate maternity services in the UK. ► The review systematically maps the positive and negative aspects of maternity care provision as experienced by immigrant women. ► The review provides strategic direction for enhancement of maternity care services. ► The review does not address the experiences of maternity care for second-generation women (eg, women of black and minority origin born in the UK).

INTRODUCTION Tobacco smoking is associated with higher risks of non-communicable diseases 1,2 and is responsible annually for >6 million deaths globally 3 . Over the last forty years, smoking prevalence shifted, becoming concentrated in low-and middle-income countries 4 as the tobacco industry employed a variety of tactics to exacerbate tobacco smoking in LMICS, such as targeted marketing and combatting effective tobacco control measures 5 . Smoking is a leading NCD risk factor in the Pacific [6][7][8] , with particularly high levels among men, ranging from 30-50% in this region. In other LMICs, tobacco use is higher among men than women, most often through smoking manufactured cigarettes, and higher in those with lower socioeconomic status and educational attainment, and among manual laborers [9][10][11][12][13] . Samoa is one LMIC where smoking is identified as an essential area for focused health promotion activities 14 . Smoking prevalence in Samoa among older and younger adults declined significantly from the late 1970s to 2013 15 . From 1978-2013, the percentage of men who smoked cigarettes daily decreased from 76% to 36%, and among women from 27% to 15%. From 1991-2013, current tobacco smoking decreased from 64% to 40% in men, and 21% to 17% in women 15 . The Samoa 2014 Demographic and Health Survey reported that 47.2% of men and 17.0% of women 25-54 years of age used tobacco 16 . In 2015, as part of a nationally representative screening survey of 2234 adults aged ≥18 years, conducted before WHO-sponsored village-level intervention programs, 45% of men and 15% of women reported smoking tobacco in the last 12 months 17 . To date, most tobacco research among Samoan adults focused very little on societal and individual correlates of smoking 18,19 , with a few exceptions. In a 2005 study among Samoans in American Samoa, Hawaii and California, smoking was associated with being male, younger, married, and less educated 20 . In the temporal trends study in Samoa among men, non-smokers had higher education than smokers, 0.5-0.8 years, but among women there were few differences by education, except in 2013 when smokers had 0.6 years more education than nonsmokers 15 . Temporal trends of lower cigarette smoking were accelerated with more education in men, but impeded with more education among women 15 . Among men aged 15-49 years, the 2014 Samoa DHS described lower tobacco use in the highest wealth quintile, those with more than secondary education, and residents of Northwest 'Upolu; in women in the same age group, there were few sociodemographic differences except for lower use among rural residents, especially those from Savai'i 16 . A 2009 study of 1056 adults aged ≥40 years residing in 11 villages in rural Savai'i found a smoking prevalence of 52.6% for men and 16.2% for women, with an apparent decrease with older crosssectional age only in women 21 . A recent qualitative study identified specific cultural patterns to smoking in Samoa, including insights into the early initiation of smoking at young ages and the deeply rooted social contexts of smoking 22 . Given these prior findings, current smoking among adults warrants aggressive measures to combat its financial and health burdens. Identifying specific sociodemographic correlates of smoking is necessary to target tobacco control efforts to subgroups, and even to tailor interventions based on specific individual characteristics 12 . Designing and adapting tobacco interventions for specific groups of Samoan adults may reduce NCDs and be cost effective for under-resourced health systems 23 . Younger adults aged <40 years, in particular, would benefit from targeted interventions as smoking cessation by the age of 40 years reduces all-cause mortality by 90% compared to those who continue to smoke 24 . Reducing adult smoking may also help reduce adolescent smoking, considering that 44.7% of young Samoan adolescents use tobacco and 32.2% smoked cigarettes, both the highest levels among 68 LMICs 25 . Given these gaps in the existing literature, the purpose of this study is to present prevalence of smoking and number of cigarettes/day and their sociodemographic correlates from a large nationwide survey of adults aged 25-65 years, in 2010 in Samoa. We focus on sex and age stratified multivariable analyses to provide detailed information on correlates of tobacco smoking necessary for designing more targeted interventions. --- METHODS --- Research design and participants This is a cross-sectional study of health survey data from a genome-wide association study conducted from February to July 2010 in Samoa. Details on the design and methods for the GWAS have already been reported 26 . There were 3745 participants who: a) selfreported having four Samoan grandparents; b) were aged 25-65 years; c) were non-pregnant; d) had no severe physical or cognitive impairments; and e) completed the interview in the Samoan language. Participants resided in 33 villages from all four census regions ranging from the most rural to the one urban or town region: Savai'i , Rest of 'Upolu , Northwest 'Upolu, and Apia urban area. The sample was representative of the wider Samoan population in terms of marital status, educational attainment, and access to basic household amenities, but had a greater proportion of women, residents from the more rural Rest of 'Upolu and Savai'i regions, and older participants compared to 2011 census data 26,27 . The study was approved by the Health Research Committee of the Samoa Ministry of Health and the Brown University Institutional Review Board. Written informed consent was given before participation in the study. --- Assessing outcome of interest Current smoking was assessed using the question: 'Do you currently smoke cigarettes, cigars or pipes?'. We considered stratification of outcome into current smokers, those who quit smoking, and those who had never smoked, but we chose a binary outcome of current smokers versus those not currently smoking, due to a negligible proportion of those who quit smoking . As such, we classified former smokers as 'not currently smoking' along with those who had never smoked. The reported total number of manufactured cigarettes, hand-rolled cigarettes, and cigars, smoked per day among smokers was measured continuously and also classified as: ≤5; 6-10; 11-19; or ≥20 cigarettes or cigars per day. Information on number of tobacco products per day was provided by 93% of men and 92.3% of women , who were current smokers. --- Correlates Several sociodemographic characteristics were explored as potential correlates of tobacco use. In addition to age and sex, education was classified into three categories: less than secondary education completed ; secondary education completed ; and university completed or greater. The three categories used to classify occupation were: 1) unemployed, retired, or student; 2) subsistence farmers and fisherman, and semi-skilled and skilled wage laborers, and clerks; and 3) administrative managers, professionals, and political leaders. Among men in the first occupation group, for example, 93% were unemployed, and in the second group, approximately 72% were subsistence farmers and fishermen. Among women in the first group, about 96% were unemployed. Census region was included as described above given substantial rural to urban differences across Samoa. A household assets inventory of 10 items as a sensitive measure of family economic resources in modernizing societies was used, previously used to study the adult Samoan population 28,29 . Higher scores represent higher income and/or wealth levels. Following exploratory data analyses, we dichotomized this index into the lowest tertile versus the two top tertiles in order to assess the link between low income and wealth position and smoking. Current alcohol use was identified if participants indicated they had consumed alcohol at least once in the past twelve months. --- Statistical analyses Sex-stratified analyses were used throughout. Bivariate analyses provided descriptions of current cigarette smoking in association with the categorical covariates. ANOVA and Spearman's correlations were estimated to determine associations of number of cigarettes/day and sociodemographic characteristics. Logistic regressions identified significant associations of covariates with current smoking and produced final models with variables that were significantly associated . Covariates included in the final model were those significantly associated with smoking in initial analyses and initial multivariate models. Lastly, because age in a transitioning society may influence exposure to tobacco, levels of education, and health literacy, we performed age-and sex-stratified logistic regressions to identify significant associations with current smoking. We classified age as ≤40 years or >40 years old, in the age-stratified analyses. Analyses were performed using the statistical software packages Stata 15 and SAS 9.4. --- RESULTS More than half of men and more than onefifth of women reported current tobacco smoking; approximately one-third of adults smoked Tob. Prev. Cessation 2019;5:50 https://doi.org/10.18332/tpc/114093 . The vast majority of current smokers used manufactured cigarettes ; only 6% used handrolled cigarettes, 3% used cigars, and <1% used pipes. Bivariate results showed that among men, cigarette smoking was significantly associated with being single, unemployed, or a subsistence-farmer/ laborer, having a low level of household assets, and current alcohol use . Among women smoking was associated with younger age, being single, residing in the less rural census regions of 'Upolu Island, and consuming alcohol. Adjusted odds ratios among men showed that subsistence workers, laborers or clerks had significant 44% higher odds of smoking compared to men working as professionals or administrators . Men who consumed alcohol had more than three times the odds of being a current smoker. Compared to men with middle or high household assets, those with low assets had significant 36% higher odds of current smoking. Adjusted odds ratios among women indicated that those not completing secondary education had twice the odds of smoking compared to those who had a tertiary education. Compared to women aged 25-34 years, women aged 55-64 years had nearly half the odds of being current smokers. Compared to women in the Apia urban area, residents in the Northwest 'Upolu region had lower odds of smoking, while residents of Savai'i had less than half the odds of being a current smoker. Women who were current alcohol users had more than ten times the odds of being a current smoker compared to non-drinkers, though this adjusted association had a wide confidence interval driven by the low percentage of women reporting alcohol consumption in the past 12 months. Adjusted odds ratios from age-stratified logistic regression models for men are provided in Table 5. Among men aged ≤40 years, being married was associated with lower odds of smoking while those who had consumed alcohol in the past 12 months had more than three times the odds of smoking. Residents of Northwest 'Upolu had higher odds of smoking than those in the Apia urban area. Among men aged >40 years, only alcohol consumption was associated with higher odds of smoking. Among women aged ≤40 years, those not completing secondary education had nearly 2.7 times the odds of smoking compared to those with a tertiary education . Younger women who reported alcohol consumption had more than ten times the odds of smoking, though the confidence interval was very wide due to the low prevalence of women who were current smokers and alcohol consumers. This low prevalence also influenced our results for women aged >40 years, as current alcohol users were far more likely to smoke than those who did not consume alcohol, but with a drastically wider confidence interval. Among men who smoked, the mean number of cigarettes/cigars per day was 10.9 with a median of 10 and a range 0.5-60 per day. Among women who smoked, the mean daily consumption was 8.7 with a median of 6 and a range 0.5-40 per day . Men smoked significantly more cigarettes daily than women . There was a significant gender difference in the percentages of total cigarettes/ day in the categories: ≤5, 6-10, 11-19, or ≥20/day; for men they were 27.7%, 39.9%, 12.4% and 20.0%, compared to women with 41.8%, 35.0%, 10.3%, and 13.0%, respectively. One-fifth of men who smoked consumed ≥20 cigarettes/day. In multivariable models among men who smoked, age and household assets were positively associated with number of daily cigarettes consumed, as well as occupation and census region . Subsistence workers and laborers consumed more cigarettes daily, mean 11.8, than professionals or administrators, mean 10.7, and the unemployed or retired, mean 9.4. Cigarettes/day increased across the four census regions from rural Savai'i to urban Apia . In multivariable models among men aged ≤40 years, those in the upper two tertiles of household assets smoked 2.8 cigarettes/day more than the lowest tertile and the unemployed smoked fewer per day than managers/ professionals or subsistence workers and laborers . Among older men, only household assets were significantly and positively associated with daily cigarette consumption . Education was marginally associated with daily consumption in older men; those completing university smoked per day more than those completing secondary or primary school . Among women, multivariable models of daily cigarette smoking showed only marginal associations with increasing age and census region , with lower consumption among women from rural Savai'i. Multivariable models stratified by age among women found no significant associations. --- DISCUSSION The prevalence of smoking in 2010 in Samoa among men was 51.3% , and for women it was 21.8% , for the age group 25-65 years; these were higher than those reported for 2013 from the STEPS survey 19 and used in the publication on temporal trends in tobacco use in Samoa 15 . The 2010 prevalences of current tobacco smoking for men and women are also somewhat higher than levels reported in the Samoa 2014 Demographic and Health Survey report 16 , 47.2% among men aged 25-54 years and 17.0% among women aged 25-49 years; while among adults, screened in 2015, these were 45% for men and 15% for women 17 . The slightly lower levels we found in 2010, compared to those from several studies from 2013-15, may be part of the temporal decrease in current smoking levels in Samoa 15 , as well as due to methodological differences. New national level surveys are needed to look at the trajectory of these positive trends into the third decade of the 21st century, considering that, regardless of several recent studies in the last decade, 45-50% of men and 15-20% of women smoke tobacco. We found among tobacco smokers that 20% of men and 13% of women consume ≥20 cigarettes or cigars daily. Estimating daily tobacco use among all participants, including non-smokers, 9.5% of men and 2.6% of women smoked ≥20 cigarettes or cigars daily in 2010, which compares well with 2013 estimates in the temporal trends report 15 . The 2014 DHS reported that approximately 67% of all men and 40% of women smoked ≥10 cigarettes daily 16 . The similarities of the results from 2010 shown here and for later years from various sources [15][16][17][18] provide confidence in the findings. Prevalence of adult tobacco smoking and daily smoking in Samoa appears to have decreased, notably in men, within the second decade of the 21st century, suggesting both the need for more research on factors associated with this decline to sustain these likely temporal changes into the third decade of this century, and on developing targeted tobacco interventions at the structural and individual levels. Having shown that tobacco smoking levels in 2010 are similar, but somewhat higher, than those of later studies, we focused on the age and sex specific socioeconomic and behavioral correlates of smoking in Samoa in order to provide evidence for intervention development. Our multivariable results stratified by age and sex identified risk factors in different sub-groups not seen in aggregate calculations of smoking prevalence for participants of all ages in these regions, such as the 2014 DHS report 16 . Although the 2014 Samoa DHS reports plausible patterns such as lower tobacco use in men among the wealthiest, those with the highest education, and residents of Northwest 'Upolu, and lower use in rural women, it is impossible to disentangle associations with educational attainment, wealth and age on tobacco use. Our findings, especially among men of all ages, follow similar trends across LMICs that link lower economic position, socioeconomic status and agricultural and physical labor to increased prevalence of cigarette smoking [9][10][11][12][13] . Among women, smoking was higher in those aged <55 years, residents of the capital Apia census region, and those with lower education levels. Further research on the context of initiation, continuation and cessation of tobacco use in younger urban Samoan women is urgently needed to follow up on recent qualitative insights from that sub-group 22 . Education was associated with current smoking only among younger women and in the inverse direction. These findings in women differ from the temporal trend analysis, which found education was not associated with current smoking until 2013 when there was a positive association 15 . Our 2010 results in younger women did show independent associations of smoking with both urban residence and alcohol use. Further research on the expanding economic roles and potential NCD risks of women is needed to clarify the nature of these associations for intervention development. Global patterns of targeted tobacco advertising to lower social status communities and young urban people have been shown elsewhere 5 , and, despite Samoa's tobacco 2013 law regulating tobacco advertising, more research is needed on its community and individual impacts. Findings from age-and sex-stratified models demonstrate some key differences in correlates across age and gender that may be important for the design of tobacco interventions. While no factor was associated with smoking among older adults besides current alcohol consumption, the variety of correlates identified among younger men and women provides some interesting considerations. Men aged ≤40 years who were subsistence farmers, fisherman, laborers or clerks, had higher odds of smoking, consistent with results in other LMICs [10][11][12][13] , while married men in this age group had lower odds of smoking. This indicates the need to target younger single men in manual labor and clerical positions in future smoking-based programs in Samoa, including assessment, prevention, and interventions focused on cessation. This age and sex group generally uses less primary care than other adults, so implementation research is needed on community level tobacco use interventions 16,23 . Indicators of social position and socioeconomic status, i.e. household assets and occupation, were associated with daily consumption of tobacco products in men. This could be attributable to higher incomes associated with work and household income and wealth, some of which is used for cigarette purchases. The Samoa household income and expenditures report, from 2013-14, found that, nationally, 0.3% of all household expenses were for tobacco, ranging from 0.2% in Apia to 0.4% in rural Savai'i 30 . These expenditure estimates for the purchase of cigarettes in Samoa are of considerable interest given evidence across many LMICs that household expenditures for tobacco were associated with lower education and health care expenditures 31 . These opportunity costs of tobacco in combination with the likely substantive financial burden of health care costs for individuals and for households with cigarette smokers, are concerning. The findings that almost one-fourth of women of reproductive age smoked cigarettes, and those with low education levels had more than twice the adjusted odds of smoking, is alarming given the well-documented associations with later adult health outcomes as well as maternal smoking and child health. In the context of the existing high levels of adult adiposity and cardiometabolic risks among women 26 , there is an urgent need to develop societal and public health programs to prevent cigarette smoking, by avoiding initiation and helping cessation among women. Pregnancy was one reason for smoking cessation cited by women in Apia 21 . For example, research is needed to determine the efficacy of screening for tobacco use and behavioral interventions within standard prenatal care. The recent report on individual-level MHealth intervention for smoking cessation being developed and tested in Samoa is noteworthy 32 . There may be opportunities for information, education and communication campaigns sponsored by employers, religious organizations and local villages to encourage smoking cessation and prevent initiation among young people. More broadly, research on smoking cessation and primary prevention of tobacco use is urgently needed in Samoa, as there is very little systematic evidence. After adjusting for all other variables, regional differences in smoking correlates among younger men and women are notable. Men aged ≤40 years in the Savai'i and Northwest 'Upolu regions had higher smoking percentages compared to those in the Apia urban area. Women aged ≤40 years in the Savai'i region had lower smoking levels compared to those in the Apia urban area, and women in the Northwest of 'Upolu area had an odds ratio that was marginally insignificant. These regional findings for women are consistent with previous research, but the findings for men are in conflict with previous reports that smoking prevalence was also lower in Savai'i compared to other regions 16,18,21 . Among men who currently used tobacco, we found that those in rural Savai'i smoked fewer, and those in urban Apia smoked more cigarettes/cigars per day, in multivariable models. But results from agestratified models of daily tobacco consumption in men found no census region or rural-to-urban effect; rather, in young men, indicators of social position and income such as household assets and occupation were associated with tobacco consumption. This suggests that the census region socioecological influences on smoking prevalence and daily consumption need more detailed examination for the purpose of developing public health interventions. Across all models, consuming alcohol in the prior 12 months was associated with current smoking regardless of age, in both men and women. As LMICs like Samoa continue to develop, patterns of co-occurrence of smoking and drinking may begin to match those in developed countries, where the association of smoking and drinking is well-established 33,34 . The co-occurrence of these risk behaviors represents opportunities for structural interventions targeted at both behaviors simultaneously. Structural level tobacco interventions currently in operation in Samoa including smoking bans in offices, advertising bans in the media, and raising the excise taxes on retail prices, need to be continued and expanded 35 . Given the high proportion of our sample and the Samoan adult population who were unemployed, e.g. among women or subsistence workers among men, workplace smoking prohibitions need to be complemented by broader community interventions. Such interventions should be informed by the sociocultural context of smoking, as has been recently described 22 . The government of Samoa has been enacting laws since the early 21st century to address tobacco use, including the Tobacco Control Acts of 2008, 2013 and 2019 36,37 , after having joined and ratified the Framework Convention on Tobacco Control a few years earlier. Finally, it is worth noting that Samoa has increased the excise tax on retail prices of all tobacco products. As of 2015, total taxes account for 51.6% of the cost of tobacco purchases, and an estimated WST 28.6 million or US$ 10.9 million in tobacco taxation may be available for increasing public health and clinical efforts to reduce the health consequences of cigarette smoking 38 . --- Limitations This study has a few limitations. First, we used crosssectional data, limiting interpretations about causality. Second, social desirability bias may have caused under-reporting of smoking and drinking behaviors, though our findings related to prevalence are roughly in line with prior research [14][15][16][17] . Third, unmeasured confounders like stress and social interaction patterns of smoking were unmeasured making it impossible to interpret more proximate determinants, which are important for tailoring intervention designs in tandem with qualitative work 17,32 . Fourth, our research data were collected in 2010, which creates the potential that correlates of smoking may have shifted over the last decade. Although our estimates of current smoking prevalence are somewhat higher than those of studies 3-4 years later, the correlates identified are plausible and similar to other work, and will be useful in developing interventions. We are following up a sub-sample of 500 adults from 2010 and will report on changes in prevalence. --- CONCLUSIONS Our findings point to the need for continued structural and sub-group interventions targeted at reducing smoking prevalence and intensity among Samoans, especially those aged ≤40 years, which may reduce their likelihood of developing serious cardiometabolic and lung disease later in life. Interventions designed for different age, sex and socioeconomic status groups within the Samoan context are important 22,32 . Implementation research will be needed to determine 10 Tob. Prev. Cessation 2019;5:50 https://doi.org/10.18332/tpc/114093 intervention content, customization and evaluation across sub-groups. In addition, work on biological susceptibility to tobacco may be needed. Recently, the number of cigarettes/day was associated in Native Hawaiians with a single-nucleotide polymorphism variant that is very rare globally but of >17% occurrence in Native Hawaiians 39 . As Samoa continues to grapple with unacceptable high smoking levels, particularly among men, and women of reproductive age, efforts to reduce the prevalence of smoking and understand what interventions work is necessary to protect the health of Samoans moving into the future. --- CONFLICTS OF INTEREST The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none was reported. --- AUTHORS' CONTRIBUTIONS STM conceived and led the study. ACA performed the majority of the statistical analysis and with STM wrote the initial draft of the manuscript. With guidance from STM, NLH led the fieldwork data collection in 2010 and contributed to writing the manuscript. MSR facilitated the 2010 fieldwork in Samoa, and with TN, contributed to the discussion of the interpretation and public health implications of the findings. All authors read and approved the final manuscript. --- PROVENANCE AND PEER REVIEW Not commissioned; externally peer reviewed.

INTRODUCTION Tobacco use in Samoa has been described over time by age, sex and education, but little work exists on other sociodemographic factors associated with tobacco use. We describe current smoking and daily tobacco use in adults from Samoa, with a focus on sex and age stratified analyses of the influence of occupation, education, census region, household asset ownership and alcohol use in order to help develop potential targeted interventions. METHODS In 2010, a nationwide survey of 3745 adults aged 25-65 years from 33 villages was completed in Samoa. Current smoking status, daily tobacco use, as well as current alcohol use, and a variety of sociodemographic factors were assessed by interview. Bivariate and multivariable models, and sex and age group stratified analyses, were performed to determine different patterns of correlates. RESULTS More than half of all men (51.3%) and 21.8% of women were current tobacco smokers. Men and women smoked on average 10.9 and 8.7 cigarettes/ day, respectively. Twenty per cent of men consumed ≥20 cigarettes/day. In men, being married, a subsistence-farmer/laborer, an alcohol user, and having low household assets, were independently associated with being a tobacco smoker (all p<0.01). Among women, not completing secondary education, being 25-34 years, residing in urban Apia, and being an alcohol user, were independently associated with being a tobacco smoker (all p<0.01). CONCLUSIONS Tobacco use in Samoa remains high and correlates of smoking suggest that interventions for cessation need to be developed within the contexts of sex, age, education, and household socioeconomic status.

INTRODUCTION Children are the fundamental guarantee of human sustainable development 1 . As an important resource for the future construction of a country, their physical health is a strategic issue directly related to the future and destiny of the country . Children's physical fitness has always been a health issue of global concern. A survey of children and adolescents aged 2-19, including toddlers, showed that obesity rates rose by 6.9% for males and 6.4% for females in developed countries; obesity rates in developing countries rose by 4.8% for males and 5.0% for a female between 1980 and 2013 . According to the 2014 National Physical Fitness Monitoring Survey, the level of morphological indicators such as height has improved in China, while the level of physical fitness has decreased, ranging from 0.1 to 2.3%. The Report on Childhood Obesity in China points out that the obesity rate of children in large-and medium-sized cities alone is 4.3%. Without effective intervention measures, the obesity rate of children will rise to 6% by 2030, which seriously threatens children's physical health. Therefore, it is of great significance to pay attention to children's physical quality for the sustainable development of children's physical health . Since 2000, by the national physical fitness monitoring indicators issued by the state, China has conducted regular and unified tests on the monitored objects nationwide in the form of sampling surveys; children aged 3-6 are the important monitoring objects. The age of 3-6 is a critical period for children's growth and development. A follow-up investigation found that obesity in early childhood can increase the probability of obesity in adulthood . And obesity can cause chronic diseases, such as hypertension, diabetes, and cardiovascular disease . Therefore, it is of great significance to investigate and monitor children's family environment, sports behavior and shape, quality, function, and other content indicators at this stage . Among physical measurement indexes, Body Mass Index , which is calculated as BMI = weight /height 2 , is an objective index commonly used in the world to measure the degree of body fat and thinness . BMI, as one of the important indicators for child nutrition monitoring and obesity screening, has reached broad consensus on theory and practice. Many organizations and countries around the world have established the standard curve of BMI percentile and the threshold of overweight and obesity screening for preschool children, such as the International Obesity Organization , WHO . Relevant research found that BMI was not only significantly correlated with the occurrence of hyperglycemia, hypertension, hyperlipidemia, and other diseases but also can directly reflect the body fat level of the human body. This index was also significantly correlated with the health of children and adolescents after adulthood ). Due to their cognitive characteristics, children's cognitive ability is lower than that of adults. Generally, children cannot rely on direct experience when acquiring information, and they need to rely more on people around them to provide information. Studies have found that parents' upbringing, static screen time was too long, and inadequate activity would lead to children's physical deterioration, and parents' upbringing was the integrated embodiment of its education behavior and idea, which also was a kind of performance in the process of child-rearing behaviors and tendency; beyond that, family structure and the cultural level can significantly affect their parents' upbringing . Although there are many studies on the influence of the family environment on the next generation's sports healthy behavior, previous studies mainly focused on the influence of the family capital on its healthy behavior. Moreover, very few studies focused on the influence of family sports attitude on their sports behavior, screen time, and BMI. In addition, previous studies on children's BMI mainly focused on descriptive analysis without effectively exploring its path relationship. Therefore, the path relationship between the influence of family sports attitude on children's BMI, especially in different family structures, remains to be further studied. Therefore, based on the analysis of previous studies and by the characteristics of children's growth and development, this study constructed a theoretical framework model of family sports attitude, affecting children's BMI. This study explored the influence mechanism of family sports attitude on children's BMI.It is found that family sports attitudes can affect children's physical health by influencing children's screen time and sports participation. The results of this study will provide a useful reference for the rational design and planning of children's physical and mental health courses, and it will provide a useful reference for improving children's physical activity level and preventing children's obesity. Therefore, based on the analysis of previous studies and by the characteristics of children's growth and development, this study constructed a theoretical framework model of family sports attitude affecting children's BMI, thus providing a useful reference for the future design and planning of children's physical and mental health courses, promoting their physical activity levels and the prevention of children's obesity in the future. --- LITERATURE REVIEW Family Sports Attitudes With Children's Physical Activity and Screen Time Social psychological studies have proved that values affect attitudes and attitudes affect behaviors. Sports attitude is an outer variable, which can only affect children through sports behavior. The characteristics of children's physiological cognition lead to the need for nurturing children to provide them with sports experience and information, and sports are highly infectious. Researchers have found that children from families with supportive attitudes toward physical activities were significantly better than those who did not . Therefrom, family sports attitude would have a certain influence on children's physical activity participation behavior and psychology . It was found that children were more affected by the feedback of parents' sports evaluation, and they were more susceptible to the influence of peers in adolescence. From that, parents play a key role in the influence of children's physical activity behavior in early childhood . Compared with children without parental support, children with parental support were more likely to participate in sports activities and believed that they have the better athletic ability . A study that followed daily activities of 155 children aged 4-7 for 3 years found a high correlation between low levels of gaming activity and high BMI, while the high level of physical activity was positively correlated with the BMI-Z score in preschoolers . The BMI of children with moderate to highintensity physical activity was significantly lower than that of children with low-intensity physical activity . However, there are also inconsistent results with the above studies. Researchers conducted a randomized controlled intervention experiment on children in 36 nurseries in Douglas, Scotland, three times a week, and found that the BMI of the children in the low-intensity physical activity experimental group was not lower than that of the control group. However, there was a significant difference in BMI between the moderate-to-high-intensity group and the control group, so the amount of moderate-to-high-intensity physical activity would significantly affect the BMI of children . Fitzgibbon et al. found that moderate to high-intensity hip-hop was effective in reducing subsequent increases in BMI in preschoolers in Chicago, Illinois, but had no significant effect in Latin America. In addition, sedentary behavior has also attracted a lot of attention from researchers, especially since the screen time of sedentary behavior has almost become the focus . The survey found that the average screen time of primary school students in Hong Kong was 4 h per day on weekdays and 6 h on weekends there was a positive correlation between screen time and BMI among children . Three-year-old children with more than 8 h of screen time per week had an increased risk of obesity at the age of 7, a finding found in another study, which also reduced BMI, triceps sebum thickness, and abdominal circumference when screen time was reduced . --- Parents' Education Level and Children's BMI As the core of the family and the first teacher of children, the influence of parents on their children's behavior and attitude is the focus of scholars' research. Parents' education levels are an important background characteristic that affects children's development. A large number of studies have reported that parents' socioeconomic status is closely related to their children's cognitive ability, social emotion, and physical health. The indicators of social and economic status include parents' education levels, income, occupation, etc., among which parents' education levels are a stable and important indicator . Cheng et al. found that a mother's occupation and length of education had a significant impact on the BMI of her children. The higher her occupational class and status, the longer her education, the lower the BMI of her children. The more time children spent in physical activity with their parents, the lower their BMI was . In the study of family social economy and BMI of children and adolescents, Shi et al. found that the education level of parents can significantly affect the BMI of children and adolescents, and the higher the education level of parents, the lower the probability of the children's BMI being obese. In consequence, the mechanism of the parental educational level in the influence of family sports attitude on children's BMI needs to be further studied. --- Family Structure and Children's BMI Early childhood is a critical period for the development of physical health. When children choose physical activities at a young age, they can only complete them under the guidance of caregivers, so caregivers play a pivotal role in this process, and the family environment in which children grow up and its composition are of great importance. The researchers found that the physical activity style of the caregivers, the degree of support and supervision for their children's physical activity, and the parenting methods would all affect children's sports participation . Therefore, the composition of family members and the time and the way of accompanying children in their growth process have a significant impact on children's sports participation and eating habits, and these factors are related to children's BMI. There are mainly nuclear families living with their parents and non-nuclear families, such as singleparent families, intergenerational families, and three-generation families. Children with different family structures have different primary caregivers, and children's sports participation and other factors are different due to different parenting styles and health concepts. In recent years, the influence of family structure on children's health behaviors has gradually attracted the attention of scholars. However, the focus is mainly on the mental health problems caused by intergenerational rearing and single-parent families, while the influence of family structure on children's physical health and sports behaviors is rarely studied . Then, the specific mechanism of children's health differences caused by family sports attitude is still unclear, and the influence of different family structures on children's physical fitness needs to be further studied. In conclusion, first, previous studies have confirmed that family sports attitude has a certain influence on children's physical activity and screen time. Second, exercise participation and screen time were important factors affecting children's BMI. However, whether exercise participation and screen time have a mediating effect on the influence of sports attitude on children's BMI has not been verified. Third, since the family structure can influence children's sports behavior, it has not been confirmed whether the influence path of sports attitude on children's BMI will change in different family structures, that is, nuclear family and non-nuclear family. Because of this, this study hypothesized that H1: family sports attitude was negatively correlated with infant BMI; Hypothesis H2: children's activity participation and screen time played a mediating role in the relationship between family sports attitude and children's BMI. --- MATERIALS AND METHODS --- --- Instrument The instrument of this research was a questionnaire, which was designed and based on a large number of research literature; the following three parts were sorted out and formed: --- Personal Background The content includes children's gender, age, height, weight, and other actual basic information, and the corresponding BMI is calculated according to their height and weight. --- Children's Sports Participation In this part, the exercise participation calculation formula used by Fox was adopted: exercise participation = exercise frequency × . The higher the value, the higher the exercise participation degree. This part includes the number of days for children to have screen activities in a week, the number of times of exercise in a week, activity intensity, and activity time, etc. Specifically, the classification of exercise time is as follows: 1 means 30 min or less, 2 means 31-60 min, and 3 means 61-90 min, 4 means 91-120 min of exercise, 5 means 121 min or more. Exercise intensity is also divided into five levels: 1 means not tired at all, 2 means not tired, 3 means a little tired, 4 means very tired, 5 means very tired. The frequency of exercise is the number of exercises per week: 1 means less than one time per week, 2 means one to two times per week, 3 means three to four times per week, 4 means five to six times per week, and 5 means more than six times per week. --- Home Environment This part includes the basic information of children's main caregivers, occupation, education level, physical exercise frequency, and attitude toward sports. The physical exercise attitude scale contains nine items, such as Do you think your child's physical fitness is enhanced by playing sports games. The scale was quantified by a Likert five-point scale, according to the options "disapprove, not very agree, general, somewhat agree, and strongly agree"; the scale was rated as 1-5 points, respectively. The higher the total score, the more support the parents have for children's sports. The pretest of the scale showed that the retest reliability was high, and the overall Cronbach α coefficient was 0.87. The overall verification results of the measurement model were as follows: χ2/df = 1.84, RMSEA = 0.03, AGFI = 0.98, CFI = 0.99, TLI = 0.99, IFI = 0.99, GFI = 0.99, which shows that the questionnaire has good measurement validity and reliability. --- Statistical Analysis SSPS Statistics 21.0 was used for the statistical analysis. After the Shapiro-Wilk normality test, the independent sample t-test and one-factor ANOVA were used to analyze the differences in demographic variables. Pearson's cross-production correlation was used in parents' education levels, family sports attitude, sports participation, screen time, and children's BMI. AMOS 21.0 was used to construct the structural equation model and conduct path analysis. The mediation test process proposed by Gerbing and Anderson was used to explore the mediation effect. First, the independent variable had an impact on the dependent variable, and the regression coefficient was significant. Second, the independent variable affected the intermediary variable, and the regression coefficient was significant. Third, the joint influence of the independent variable and intermediary variable on the dependent variable was significant. The significance level of all indexes was set to α = 0.05. --- RESULTS --- Demographical Variance Analysis of Sports Attitude and Sports Behavior The results of the independent sample t-test show that, in terms of gender, there was no difference in boys' and girls' family attitudes toward physical activity . There was no difference in the physical activity behavior between boys and girls . There was no gender difference in screen time . There was no gender difference in infant BMI . The results of analysis of variance showed that: In terms of a mother's education level, there were differences in the family sports attitude score, children's sports participation, screen time, and children's BMI among different mothers' education levels. Specifically, the scores of the mother with primary school education were lower than those with high school education, while those with high school education were lower than those with a college education or above . There were also significant differences in children's sports participation between fathers with different education levels . Children with fathers' education levels had the highest sports participation in college or above, followed by children with fathers' education levels in middle to high school. There was no significant difference in children's sports participation between fathers' education levels in middle and high school and primary school children. The screen time of children was significantly different only between primary school and college or above groups. The screen time of mothers with primary school education was significantly higher than that of college or above groups . In terms of children's BMI, there were differences between mothers with different education levels. The BMI of children whose mothers had primary school education was significantly higher than that of those from middle school to high school, and the BMI of children whose mothers had middle school to high school education was higher than that of children whose mothers had a college education or above. In terms of fathers' education levels, there was a significant difference in sports attitude among families with different fathers' education levels . The score of a father's education from middle school to high school was significantly higher than that of primary school, and there was no significant difference in the score of family sports attitude when a father's education was higher than middle school . There were also differences in children's sports participation among fathers with different education levels. Fathers' education levels in college and above were significantly higher than those in primary school , while there was no difference between fathers' education levels in primary school and middle school to high school . In terms of screen time, there was no difference between fathers with college or above education background and those with middle school to a high school education level , while there was no difference between fathers with middle school to high school and primary school education levels . In terms of BMI of children, the father's education level from middle school to high school was significantly lower than that of primary school, and the father's education level from university or above was also significantly lower than that of primary school . There was no difference in BMI of children between a father's education level from middle school to high and those education levels from a college or above . Family structure: there were differences in family sports attitude, sports participation, and screen time among children with different family structures, but there was no significant difference in BMI. Multiple comparisons showed that the scores of family sports attitude in families with parents living together, intergenerational parenting, and three generations living in the same house were significantly higher than those in single-parent families . The results showed that the children's sports participation of three generations in the same family, intergenerational parenting, and single-parent family was significantly higher than that of a small family . Screen static time per week in families with three generations and parents living together was higher than that in single-parent families . --- Correlation Analysis of Family Sports Attitude, Parents' Education Levels, Sports Participation, Screen Time, and Children's BMI Correlation analysis showed that there were significant correlations among the variables. There was a significant negative correlation between family sports attitude, sports participation, parents' education levels, and children's BMI, and there was a significant positive correlation between screen time and children's BMI. Hypothesis H1 of this study was confirmed. There was a significant positive correlation between sports participation and family sports attitude . There was a significant negative correlation between family sports attitude and children's screen time . There was a significant negative correlation between family sports attitude and children's BMI . --- Linear Regression of Family Sports Attitude, Sports Participation, Screen Static Time, and Children's BMI The results of linear regression showed that family sports attitude was negatively correlated with children's BMI. Sports participation was negatively correlated with young children's BMI. One week of static screen time was positively correlated with children's BMI. And they all reached a significant level . Path Relationship Analysis of Family Sports Attitude, Sports Participation, Screen Static Time, and Children's BMI The relationship of family sports attitude, sports participation, screen time, and children's BMI, the mediating effect between screen time, and exercise participation were examined by Wen and Ye mediating effect test procedures . Figure 1 shows a structural equation model testing the mediating effect of exercise participation and screen time. The structural equation model contains four potential variables, and the results show that the fitting indexes of the structural equation model are as follows: x2/df = 1.84, df = 1, RMSEA = 0.03, AGFI = 0.98, CFI = 0.99, TLI = 0.99, IFI = 0.99, which indicated that a model can be established well. Figure 1 shows the standardized coefficient in the mediation model diagram, and Table 5 shows the results of indirect effects. As shown in Table 5, the confidence interval of each mediation path does not contain 0, so the mediation effect is significant. Hypothesis H2 in this study is valid. As shown in Figure 1, the direct effect of family sports attitude and children's BMI in the model was not significant , but the negative prediction effect of family sports attitude and screen time was significant . Family sports attitude and children's sports participation had a significant positive predictive effect . The BMI of children was predicted negatively by sports participation , while the screen time could predict BMI positively . From the significant levels and the standardized path coefficient b values in the structural model in Figure 1, it is not difficult to find that screen time and sports participation in the model play a mediating role between family sports attitude and children's BMI, and the mediating role includes three paths: The single mediating effect of screen time and sports participation, and the chain mediating effect of screen time → exercise participation. To verify the mediating effect of sports participation and screen time between family sports attitude and children's BMI, nonparametric percentile Bootstrap was used to test the significance of mediating effect. The original data were sampled 2,000 times, and a 95% CI was estimated. If the confidence interval did not include 0, the mediating effect was significant. The results in Table 5 show that screen time and sports participation played a complete mediating role between family sports attitude and children's BMI, while screen time → sports participation played a chain-mediating role. None of the three intervals contain 0, indicating a significant mediating effect. Therefore, hypotheses H1 and H2 of this study have been confirmed. --- Difference Test of a Mediation Model in Different Family Structures Relevant studies found that different family structures in children's lives can lead to significant differences in children's sports participation, screen time, and BMI. This study verified whether sports participation and screen time have a stable mediating role in different family structures. Firstly, the mediating effects of children's sports participation and screen time in nuclear and non-nuclear families were examined respectively. The results showed that the model fitting indexes of nuclear families were: χ2/df = 2.51, AGFI = 0.96, CFI = 0.99, TLI = 0.96, IFI = 0.99, RMSEA = 0.067. The model fitting indexes of non-nuclear families were: χ2/df = 1.67, AGFI = 0.96, CFI = 0.99, TLI = 0.98, IFI = 0.99, and RMSEA = 0.052. The results showed that the fitting indexes of each group were good, which could be compared with different groups of models. Then, the set equivalent model was used to compare the different groups of models, and the results showed that the model had good fitting indexes, and the fitting index difference CFI and TLI values were less than 0.05 . The results show that the mediation model is stable in different family structures. --- DISCUSSION Analysis of Influencing Factors of Infant BMI, Family Sports Attitude, Sports Participation, and Screen Time This study found no gender differences in children's BMI, family sports attitude, weekly sports participation, and static screen time, which is consistent with the results of the Xinran Shi's study . At the same time, this study found that children with different family structures have significant differences in sports attitudes, screen time, and sports participation . Further analysis found that children with single-parent family structure scored the lowest in family sports attitude, and there were significant differences in living with their parents, grandparenting, and three generations in the same family. Previous studies showed that children living in the single-parent family structure were more likely to have physical and mental health problems than those who live in a two-parent family structure . Quarmby et al. confirmed that the lower socioeconomic status of single-parent families compared with two-parent families partly explains the poorer physical health of their children. Compared with European and American countries, the major feature of Chinese family structure is grandparenting. There are different conclusions about the advantages and disadvantages of grandparenting. Opponents believe that grandparents are prone to develop bad health habits due to poor health habits and lack of nutrition knowledge. Supporters believe that grandparents' love for their grandchildren, abundant time and energy investment in a company, education, and nurturing experience are all beneficial to the healthy development of grandchildren . In this study, we found that the sports participation of three-generation families was higher than that of families living with parents only, which may be related to the time advantage of grandparents to pay attention to and accompany their children's activities. Other researchers have suggested that the three-generation family structure is more conducive to the formation of healthy concepts in young children, as the role of family bonding is played by grandparents living in the same family. In addition, this study found that the higher the mother's education level, the lower the children's BMI. Qian et al. conducted a study on Chinese students aged 8-10 and found that the group whose parents have a low educational level has a low score in dietary habits and nutritional knowledge, while dietary habits are significantly correlated with BMI . Han Hui et al. found a significant correlation between children's food preference risk and their mothers' education levels, and the higher the mother's education level, the lower the risk of children's food preference risk . The study also found that parents' education levels were negatively correlated with children's screen static time. Researchers also found that the daily screen time of children whose parents have a college education level or above was significantly lower than that of children whose parents have a primary school education level or below. The study also found that children whose parents had college education levels spend more time on moderateintensity activity than children whose parents had middle school education levels or below . The results of this study are also consistent with those of other studies . So, the results of this study are not only limited to Chinese children but also consistent with those of Spanish and Caucasian children; the results showed that the educational level of the parents was related to the health habits of children . Wang et al. found in their research that the concept of exercise will gradually internalize into selfidentity with the improvement of the education level, and the influence of the education level on physical exercise behavior will be increasingly enhanced. Therefore, parents' education levels have a profound influence on family physical exercise. --- Family Sports Attitude and Children's BMI There is a significant negative correlation between family sports attitude and children's BMI, which is consistent with the previous research results of Gao Yan et al. A large number of studies have shown that the decisions of children's sports participation at the beginning, middle, and end are related to the environment. Human behavior is formed by the interaction between individuals and the environment. The environmental models that affect human development include micro, medium, and macro-environmental models. Family is an important part of the microenvironment, which plays a key role in the process of children's growth . Family is not only the "gatekeeper" of children's sports activities but also the key factor of children's screen time duration . Effective intervention measures aimed at family factors can affect children's physical condition. Therefore, a better understanding of the influence of family sports attitude on children's physical health is helpful to reduce the incidence of childhood obesity and malnutrition caused by family factors. Previous studies have shown that the influence of family, parents, and elders on children's sports behavior can be divided into two aspects: the support of elders for children's sports and the shaping of the role model of elders . All the above results indicate the importance of family sports attitude in children's physical health among the influencing factors of the family on children's physical health, and family should be regarded as the implementation focus of children's physical health intervention. --- Chain-Mediated Effects of Children's Screen Time and Exercise Participation In this study, screen time and sports participation were used as mediating variables to construct a mediating model of family sports attitudes affecting children's BMI. The mediating effect of sports participation and screen time is significant, which indicates that they played an important role in family sports attitudes in interpreting children's BMI. Fan Huiying et al. found that family support has a significant impact on children's physical activities, and family as the starting point of children's sports participation there was the strongest correlation between their support for children's physical activity attitude and children's sports . So, family sports attitudes play an important role in young children's sports participation and static screen time limitation. Parents, as the first teachers of children's physical education class, will have a positive impact on children's sports by setting an example in sports participation and supporting the concept of physical education, whereas the negative performance of family physical education attitude and behavior will have a negative impact on children. Reilly et al. found that Scottish children's physical fitness declined due to insufficient sports participation in recent years. One important and direct reason is that parents have a negative attitude toward sports participation . Researchers found that the longer children's moderateintensity activity time was, the shorter their screen time was, which was related to children's family sports cognition . This view was consistent with the results of this study. Parents' support for children's sports can improve their children's sports participation, while children's sports exercise is negatively correlated with children's BMI . Sports attitude can influence children's BMI through sports participation. Another important factor affecting BMI is sedentary behavior, which includes screen time, homework time, and so on. Related studies have found that sedentary behavior mainly based on screen time has a negative impact on health problems such as overweight and myopia in adolescents and children. Moreover, the study found that family and peer support had a significant impact on sedentary screen time, and parents' support for children's physical activity was negatively correlated with screen time . It can be seen that the family environment constitutes an external driving force for children's screen time, which is mainly manifested in exercise support and participation. When family members form a state of "high screen time-low exercise time, " children's screen time can be invisibly increased, thus leading to the increase of their BMI. This indicates that the mediating effect of screen time and exercise participation in this research model is feasible, and it can be seen from Figure 1 that the mediating effect is completely mediating. To some extent, this result explained the influence mechanism of family sports attitude on children's BMI and provided evidence for the intervention of family factors in children's physical health development. --- Differences in Mediating Effects of Screen Time and Exercise Participation Among Different Family Structures Screen time and exercise participation were fully mediating effects in different family structures, and the chain-mediated effects held in both nuclear and non-nuclear families. In the family structure, the nuclear family is the family composed of husband and wife and unmarried children, while the nonnuclear family includes three generations living under the same family, single parent, and grandparenting families. This study found that the total mediating effect of screen time and exercise participation in non-nuclear families was higher than those in nuclear families, which may be related to the higher proportion of three generations living in the same family in non-nuclear families. The family structure of three generations under one roof is the most abundant and complete. Children grow up in this environment with both parents and grandparents. The study found that grandparents had more time and energy to spend with and participate in children's activities than busy parents. However, the lack of parents' company and upbringing may result in children's bad health behavior habits due to the lack of scientific sports knowledge, nutrition knowledge, and poor health concept of their grandparents . Therefore, both parents and grandparents play important roles in the development of healthy behaviors in young children. In this study, there were significant differences in children's family sports attitude, sports participation, and screen time among different family structures, but there was no significant difference in the mediating role model of sports participation and screen time. The results indicated that the three factors of family sports attitude, sports participation, and screen time were different due to the influence of family structure variables, but the chain-mediating effect of sports participation and screen time in the influence mechanism of family sports attitude on children's BMI had internal similarities among different children's family structures. --- CONCLUSION AND PROSPECT Conclusion There are significant differences in children's family sports attitude, sports participation, screen time, and BMI with different parents' education levels. Specifically, the higher the parents' education levels, the higher the family sports attitude score, the better the sports participation, and the lower the screen time. Family sports attitude is positively correlated with sports participation and a mother's education level and negatively correlated with children's screen static time and BMI. Sports participation and screen time play a completely mediating role in the influence path of family sports attitude on children's BMI. The mediating mechanisms of sports participation and screen time were not affected by family structure , but the total mediating effect of sports participation and screen time was higher in non-nuclear families than in nuclear families. The results of this study will provide a useful reference for teachers and parents to control children's physical health and the influencing factors of children's physical health research in the future. --- Limits and Prospects This study identified the mediating role of sports participation and screen time. The structural equation model of the relationship between the attitude of plus family and children's BMI was constructed and verified. And the study still has the following limitations and suggestions for future research: Among the family influencing factors of children's BMI, this study focused on examining the mediating effects of children's sports participation and static screen time. This study was a cross-sectional study. Future research can further support the conclusion of this study by using the exercise tracking method and providing direct evidence for variable causality. --- 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. --- ETHICS STATEMENT The studies involving human participants were reviewed and approved by the Ethics Committee of Sports School of Southwest University in China. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin. --- --- 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.

Background: Children's physical health is an important resource for a country's future construction. However, researchers found that the physical fitness of young children around the world has declined during the two decades, from 1992 to 2012. The decline in the physique of young children has caused widespread concern around the world. Children's main living places are families and kindergartens, so this research explores the impact of family factors on children's body mass index (BMI) from the perspective of family attitudes, children's sports participation, and screen time. Methods: A cross-sectional study was used to conduct a questionnaire survey of children in China. A total of 600 children were investigated, and 589 valid questionnaires were obtained. SPSS21.0 statistical analysis software was used for descriptive analysis, mean comparison, and correlation analysis of the data. AMOS 21.0 was used to construct a structural equation model and carry out path analysis.(1) There are significant differences in children's family sports attitude, sports participation, screen time, and BMI with different family structures, and parents' education levels. (2) Family sports attitude is significantly positively correlated with parents' education levels and children's sports participation, and negatively correlated with children's screen time and BMI. (3) Children's sports participation and screen time play a chain-mediating role between family sports attitude and children's BMI, and the role is a complete mediating role. Therefore, family sports attitudes can affect children's physical health by affecting children's sports participation and screen time. To promote children's physical health, we should pay attention to the intervention of family sports attitude. (4) The mediating effects of exercise participation and screen time are similar in different family structures, so the structure of this study can be applied to different family structures.

Introduction The region of Southeast Asia exemplifies both the benefits and drawbacks of globalisation. Due to its wealth in resources and its advantageous location, this region has, for a very long time, been subject to the influence of outside forces. The Association of Southeast Asian Nations , which has introduced a new degree of regional cooperation to the area, has the purpose of integrating SEA countries in the region. The region of SEA is split into two sub-regions: the mainland and the islands . There are 11 different countries that make up this region. Membership in ASEAN which was founded on August 8, 1967 in Bangkok, Thailand, and this has served to bring together the states that are located in this region. Indonesia, Malaysia, the Philippines, Singapore, and Thailand were the founding member states of ASEAN. Brunei Darussalam became a member of ASEAN on January 7, 1984. This was followed by Vietnam on July 28, 1995, Lao People's Democratic Republic and Myanmar on July 23, 1997, andCambodia on April 30, 1999, bringing the current total number of ASEAN Member states to 10 . SEA comprises 10.5% of the overall land area of Asia and 3% of the total land area of the Earth. The population of this region is over 675 million people, which accounts for around 8.5% of the total population of the world . The mainland subregion of SEA is referred to by names such as the Indochinese Peninsula and Indochina. It is located to the east of the Indian subcontinent and to the south of the Chinese mainland. To the west, it is bounded by the Indian Ocean, while to the east, it is bounded by the Pacific Ocean. Cambodia, Laos, Myanmar, Thailand, and Vietnam are all members of this group of countries . Buddhism predominates across the board in the cultural landscape of this sub-region . It is also one of the regions in the world with the lowest rate of urbanisation and features a multicultural and multiethnic population. Human rights are routinely violated in this region as a direct result of the political instability that exists here. For instance, in the span of just more than a century, Thailand has seen 13 successful and 9 unsuccessful military coups. The most recent military coup occurrence in this sub-region took place in February of 2021 in Myanmar . Malaysia, Singapore, Brunei, East Timor, Indonesia, and the Philippines are the countries that make up the insular SEA sub-region known as the Islands of SEA. This area is a subrealm that is made up of a number of peninsulas and islands. However, despite its status as a developing region, it is frequently the site of environmental and regional strife. All of the states' histories in this sub-region include periods of colonial rule. In fact, Thailand was the only country in SEA that was never colonised. Religions were introduced into this region through Western colonisation and traders. Colonial powers such as the Portuguese introduced Roman Catholicism to Malacca, Malaysia, and areas of eastern Indonesia during the 15th century. Buddhism is currently widely practised in four states such as Myanmar, Thailand, Laos, and Cambodia. Muslims traders were the first to bring Islam to SEA. By 1650, Islam had surpassed other religions and became the dominant religion in Malaysia and Indonesia . As a direct consequence of this, SEA is one of the regions in the world with the lowest religious diversity. For example, with approximately 87% of its 244 million people identifying as Muslims, Indonesia is now the world's largest state with a Muslim population . So, when Thailand became the first country in Asia to decriminalize the consumption of cannabis in 2022 several states were not comfortable with the new law. Thailand's new law allows marijuana foods, beverages, and medical use in Thailand. Such behaviours are part of Thailand's subculture but remain counterculture in other SEA states. The impact on SEA subregions was investigated in this study. This study seeks to answer the question: Could Thailand's recent cannabis culture have an impact on other SEA countries? Because this is a new study, qualitative and digital ethnography were employed. The secondary government data were analysed via thematic analysis. It is anticipated that due to tourism, people in this SEA region will be able to learn, respect, and, indeed, practise Thailand's new subcultures themselves. Thailand's tourism industry recovered well in 2022, but international arrivals will not return to pre-pandemic levels until 2024. The gradual border reopening has resulted in an increase in international visitors in 2022, with 4.4 million tourists by the end of August 2022, up from slightly more than 380,000 in 2021. The increase in visitor arrivals is expected to continue in the coming years, supporting tourism jobs and revenues in Thailand. Tourism shall be a primary sector of economic growth and it is expected to increase Thailand's GDP in 2023 . The new cannabis law is expected to attract more international tourists and promote 'weed' tourism in Thailand. Thailand is currently the home of thousands of cannabis dispensaries, and is unofficially known as the "Amsterdam of the East" . --- Literature Review --- Social Conservatism The historical and religious influences in the region contribute to the social conservatism that is an essential characteristic of SEA. Social conservatism is a political philosophy whose primary concern is the maintenance of long-established norms, values, and beliefs . It places emphasis on a concern with moral and social values, which many adherents of the ideology believe have deteriorated in today's modern society as a result of liberalism. Social conservatives in SEA organise in support of duty, and move to uphold traditional values and social institutions such as traditional family structures, gender roles, sexual relations, and national patriotism . Social conservatives in SEA also uphold traditional religious practices. The anti-democratic and anti-liberal social forces in SEA have successfully outweighed democratic and liberal elements via laws and policies in civil society . As a result, political and ideological divides exist in this region and they have become more pronounced over the course of the most recent years. As a case in point, on December 6, 2022, the Indonesian legislature unanimously enacted a new, comprehensive criminal code that criminalises extramarital sexual interactions. The law is applicable to everyone in Indonesia including international tourists . Recent modifications to the Indonesian penal code are a clear reflection of the trend of rising social conservatism that is prevalent throughout SEA. There are trends that are suggestive of tension in this regard, however. Only a few days before the Indonesian law came into force, for instance, on November 29, 2022 Singapore decriminalised male sexual activity . But, concurrently, the constitution in Singapore was altered to preclude legal challenges to the conventional standards of marriage. Several public issues, particularly those pertaining to public housing, education, and family planning, are affected by the way marriage is interpreted in Singapore. As a result, same-sex couples cannot enjoy the rights and advantages of marriage, including social security, inheritance, adoption, common ownership of property and goods, tax and benefits. Meanwhile, the Vietnamese government announced on November 11, 2013, that it will no longer impose fines on anyone who conduct public same-sex wedding ceremonies . However, such ceremonies are neither recognised by the law nor offered legal protection. In spite of its rapid economic progress, Singapore is still often regarded as a conservative state that is especially stringent with some regulations. Violators of these laws face the possibility of being arrested or imprisoned. The authorities even continue to use a centuries-old procedure called corporal punishment, which consists of hitting a person with a rattan cane. As a legacy of British colonisation, caning is still employed as a form of disciplinary action in Malaysia and Brunei. The province of Aceh in Indonesia, which has a mainly Muslim population, is the only one in the country to observe Islamic law and to publicly cane people for offences such as stealing, gambling, and adultery. In Aceh, Sharia law is strictly implemented, which includes prescribing the outward appearance of society members via a dress code . Thus, being modestly dressed in SEA, such as being covered from the waist to the knees for men, and having the arms, legs, back and chest covered for women are typically regarded as essential. For instance, despite being a multicultural society, Malaysia has many local dress code traditions and expectations. In Malaysia, it is strongly recommended that members of the public, particularly tourists, stay away from certain types of apparel, such as singlets and short skirts. Meanwhile, in Singapore, Section 27A of the Miscellaneous Offences makes it illegal to be naked in public, which includes being naked in private locations that are visible to the general public . However, despite the fact that going topless or nudist while sunbathing is considered impolite in Thailand, Thailand's legal system does permit clothing-optional resorts. In a move consistent with this trend toward social liberalisation, meanwhile, on June 9, 2022, cannabis in Thailand was removed from the list of illegal substances . --- Weed Tourism: Subculture vs Counterculture This legal framework for cannabis dates back to the 1970s. The Narcotics Act was enacted by Thailand in 1979. The classification of cannabis as a Class 5 controlled substance was determined. However, in 2019, Thailand passed the Narcotics Act , which legalises the use of cannabis for scientific and medical study. Therefore, initiatives to use cannabis in research and development have been given the go-ahead to instigate the cannabis subculture . In the year 2020, the Thai Food and Drug Administration announced that cannabis leaves and roots generated from recognised sources will not be matched the criteria for classification as type 5 narcotics. These actions made it possible to legalise cannabis while setting some restrictions. During the following year, the FDA registered approximately one hundred locations such as government entities and agricultural communities with state registrations to cultivate marijuana . On June 9, 2022, Thailand made history by permanently removing cannabis and hemp from the list of Category 5 narcotics . This decision was highly significant enormous with regard to the tourism industry in Thailand. Initially, all aspects of the cannabis plant, including possession, manufacture, distribution, consumption, and sales, were legalised in Thailand for medical use only. However, since the beginning of June 2022, numerous food and beverage companies around the nation have capitalised on the recent liberalisation of marijuana by stocking the shelves of convenience stores with cannabis-infused foods and beverages containing marijuana-derived ingredients. It should be noted that while in order for food and drink products to be legally sold in Thailand they must contain less than 0.2 percent tetrahydrocannabinol , cannabis-infused food and drinks can be served including to tourists in cafes and restaurants provided the THC content of these products remains below the 0.2 percent threshold. Notwithstanding this point, a recent study conducted at a renowned institution in Thailand revealed that thirty percent of these items exceeded the 0.2% THC limit set by the government . THC is the principal psychoactive component of marijuana. As a result, without a doubt, the cannabis counterculture has begun to grow in Thailand following the drug's legalisation. It is plausible to argue that in global terms the cannabis subculture functions as a cultural subgroup inside the larger culture. Members of this group could hold beliefs and passions that are not shared by the larger culture. Around the world there are now neon-lit signs featuring cannabis plants, and cannabis is making its way into the food, drinks, and cosmetics industries. This tendency is bolstered by the prospect of profit from weed tourism, which is anticipated to be worth more than one billion dollars for Thailand by 2025 . The revenue from cannabis has been a lifeline for cities devastated by the restrictions related to Covid, which have wrecked Thailand's tourism-based economy . Cannabis in Thailand has now been advertised as a vacation experience by the Tourism Authority of Thailand , which has published a series of guidebooks promoting places where tourists may visit organic cannabis fields and receive cannabis oil massages . Since June 2022, more than one million people have been granted permission to cultivate cannabis through the government's mobile phone application. According to the Thailand Ministry of Public Health, this number is anticipated to continue rising . However, throughout the rest of the SEA, drug prohibitions are stringent. In Singapore, in 2022 itself, five men were executed for drug trafficking offences . Human Rights Watch reports that since assuming office on June 30, 2016, President Rodrigo Duterte of the Philippines has waged a war on drugs that has resulted in the deaths of over 12,000 Filipinos, largely the urban poor. The Philippine National Police has been implicated in at least 2,555 killings. In a campaign that may constitute crimes against humanity it has been reposted that Duterte and other senior officials have approved and incited those killings . Thus, with the prominent exception of Thailand, the cannabis culture in SEA remains a prominent counterculture. The ideals and conventions of cannabis consumption behaviour diverge significantly between Thai and other SEA mainstream societies. The Thai society's tolerance towards cannabis appears to be strongly contradictory to SEA mainstream cultural values. This is not the first time Thailand has legalised a subculture that contrasts with that of other SEA nations. In 2007, NAT, which stands for Naturist Association Thailand Co., Ltd., was founded. In 2011, Chan Resort became the first resort in SEA to legally permit visitors to disrobe on the property . In 2015, when the Barefeet Naturist Resort in Bangkok initially opened its doors, visitors were not permitted to choose whether or not to wear clothing. Guests are required to be naturists or to be accompanied by a naturist at all times . The NAT has three thousand members as of 2016, including individuals from India, Australia, Hong Kong, the Philippines, Myanmar, Thailand, Singapore, and Taiwan. In 2022, Thailand was home to eight resorts that did not restrict tourists to wearing clothing. The resorts are well established around Thailand's major cities such as in Bangkok, Pattaya, Phuket, and Chiang Mai, despite the pandemic . Thailand is a popular destination for naturists from other nations since the climate is good throughout the year. The majority of non-Western clothing optional tourists are not just from SEA but also from Australia, India, the Middle East, and China . Meanwhile, in 2015, the Thai Parliament passed the Gender Equality Act. This law is intended to protect lesbian, gay, bisexual, and transgender community members from discrimination based on gender identity and sexual orientation . Although Thailand has a long history of tolerant attitudes towards LGBT individuals, this is the first law that clearly guaranteed their rights . As a result, in 2017, Bangkok was voted the second most LGBTfriendly city in Asia due to its thriving LGBT community. Bangkok was ranked as the 61st greatest LGBT destination in the world, despite the fact that same-sex marriage remains outlawed. The score was determined based on surveys conducted in 80 nations, in which respondents ranked the LGBT tourism such as dating places, nightlife, openness, safety, and legal rights in each location . As expected, no other ASEAN cities were ranked and listed on this survey. Meanwhile, travelling to Bangkok or other cities in Thailand is not a challenge for ASEAN community members. The ASEAN Framework Agreement on Visa Exemption, which was launched during the 39th ASEAN Ministerial Meeting on July 25, 2006 in Kuala Lumpur, was a significant initiative designed to boost and speed up intra-ASEAN and labour liberalisation . This programme permits ASEAN individuals including tourists travelling throughout the SEA to travel visa-free for 14 working days from the date of admission . As a result of geographical distance and visa exemption, tourists from other SEA have been visiting Thailand in great numbers . Thus, SEA tourists passing through Thailand would have easy access to behaviours or substances, including cannabis, forbidden in their own countries. --- Research Methodology The research objectives were to understand the effect of Thailand's Subcultures on other SEA States' Countercultures. The primary research question was to investigate the effect of Thailand's Subcultures on other SEA states in the context of weed tourism particularly, that is, Thailand's recent cannabis culture. As a result, a qualitative research approach drawing on data and analysis were utilized. Qualitative data that was non-numerical in nature such as text documents was compiled and analysed in order to gain a better understanding of ideas, perspectives, or experiences. The method of digital ethnography was selected for the study of individuals in a real-world setting. Digital ethnography, also known as mobile ethnography, is a modern, digital-first approach to conducting ethnographic research remotely . In terms of data gathering, the social and behavioural sciences rely heavily on ethnography, which is also a common qualitative research technique. Secondary data collection was employed to collect data, which was then transformed into knowledge that may be utilized to develop conclusions on the functions of societies and individuals. This methodology also allowed researchers to observe responses from the SEA states in their natural surroundings. To participate in this study, SEA states were required to meet two inclusion criteria. The consequences of the recent cannabis counterculture and cannabis law in Thailand were the sole factor taken into account. Firstly, only the views and opinions after June 9, 2022, from the SEA government authorities, specifically government ministries and leaders were considered. Second, only viewpoints from ASEAN member states were examined. Consequently, only the opinions of the governments of the 10 member states were considered. The East Timor government's views on the cannabis subculture in Thailand were not included because East Timor is not an official ASEAN member state. Thematic analysis was conducted to investigate the recurring themes within the secondary data . As it is a method rather than a methodology, this approach is not beholden to a particular epistemological or theoretical stance, unlike many other qualitative research studies. Thus, this analysis makes it a very versatile strategy added a substantial advantage given the diversity of this research topic . This qualitative data analysis was used to analyse primarily non-numerical data such as audios, videos, and texts from government and official press websites, as well as from the official ASEAN website. The researchers classified all the data systematically and then organised the codes, based on similarities, into larger and larger categories, resulting in a hierarchical structure represented by the likes of code, subtheme, and theme. The researchers followed Braun and Clarke's six-phase framework, which included familiarising themselves with the obtained secondary data, generating initial codes, searching for common themes, evaluating the themes, identifying themes, and writing up the results . Using Voxco software, the researchers analysed the selected reports and searched for salient patterns in the data's repeated themes. --- Findings Since the 1970s, ASEAN governments have primarily employed strong punitive measures in an attempt to kerb drug consumption and distribution. Nonetheless, on June 9, 2022, Thailand legalised the cultivation of cannabis and its consumption in food and beverages, and removed cannabis from its list of prohibited drugs. Thailand's new law aim to enhance its agriculture and tourism industries and allow the use of plant for medical purposes. Immediately following its introduction, this new Thai law was reported in the media of all ten ASEAN member states. Hence, the researchers reviewed and analysed the responses of the ASEAN member states' leaders and government representatives to the reports of the change in Thai cannabis law with the objective being to address the primary research question. The findings of this research indicate that weed tourism activities in Thailand will help the inhabitants of this region to comprehend, respect, or perhaps even themselves adopt Thailand's subculture. Geopolitical factors have an intriguing effect on the cannabis culture in SEA. In terms of Thailand's subcultures, mainland and insular states tend to hold opposing viewpoints. --- Mainland States The Kingdom of Cambodia -Khan Samban the director of the industrial crops department at the Ministry of Agriculture said "while cannabis cultivation will be still illegal here, the government could consider lifting the ban. It is an easy crop for planting and can grow in many areas, the government would consider amending the law for special cases." . Republic of the Union of Myanmar -There has been no official remark from the military government of Myanmar regarding the legalisation and usage of cannabis in Thailand. The Narcotic Drugs and Psychotropic Substances Law in Myanmar, which was passed in 1993, regulates the use and possession of cannabis in Myanmar. Cannabis is still prohibited in Myanmar. However, this law was changed in 2018 to state that drug users are no longer considered criminals. Drug use has been decriminalised, although possession of minor quantities has not. Some argue that this weakens the entire regulation because it is difficult to use cannabis without also possessing it . Lao People's Democratic Republic -Similarly to Myanmar, Laos government officials made no official statements regarding cannabis legislation in Thailand. However, the Lao Ministry of Health informed the Thai Ambassador to Laos in a letter dated July 2022 that the kratom plant shall remained illegal in Laos. This includes cultivating, possessing, purchasing, selling, processing, storing, and importing or providing the plant to Laos . Kratom is a herbal substance that has effects similar to those of opioids and may cause psychotropic effects. It can also be addictive. Socialist Republic of Viet Nam -The Viet Nam government also issued no formal pronouncements regarding this topic. Even for medical purposes, cannabis is outlawed in Vietnam, however the nonpsychoactive chemical contained in marijuana is permitted and readily distributed. Vietnam classifies marijuana as a controlled substance, like heroin and cocaine. As such, it bears the same severe punishments as imprisonment or the death penalty. Recent reports indicate that police in Vietnam do make arrests on recreational cannabis consumption-related offences. However, offenders, including tourists, only face a maximum of a USD100 fine and a harsh warning . --- Insular States Malaysia -The Home Minister Datuk Seri Hamzah Zainuddin in a speech read by Home Ministry deputy secretary-general Datuk Abdul Halim Abdul Rahman stated "Malaysians traveling to Thailand would be easily exposed to cannabis-based products in the form of food, drinks and supplements sold there. Furthermore, cannabis-based products in the form of food, drinks and supplements may be smuggled into Malaysia. Malaysia is concerned that some Malaysians may become addicted to cannabis because they had unwittingly consumed such products." . Republic of Singapore -The Minister for Home Affairs and Law, Kasiviswanathan Shanmugam shared the point that "freer availability of cannabis in Thailand, to which a lot of Singaporeans go to and from, where a lot of tourists come to Singapore, is going to present more challenges." Republic of Indonesia -"The Royal Thai Embassy in Jakarta warn Thai people not to carry cannabis, hemp, or products with cannabis or hemp ingredients into Indonesia. Violators of the law could face a fine of at least 5 years to life imprisonment, capital punishment or a fine of roughly 2.4 million baht ". Negara Brunei Darussalam -On 13 July 2022, the Narcotics Control Bureau made a formal statement stating that people convicted of ingesting Class 'A' drugs face up to ten years in jail, a fine of up to USD 15,000, or both. Therefore, NCB reminded Brunei citizens and permanent residents who visit foreign countries that they must be aware of this penalty. Under Brunei's law, Bruneian and its permanent residents caught ingesting drugs outside of the country are also subject to prosecution, regardless of where the drugs were ingested. Chapter 27 of the Misuse of Drugs Act of Brunei criminalises the possession, consumption, and cultivation of cannabis . Distribution or possession of more than 500 grams of cannabis will result in the obligatory death penalty . Republic of the Philippines -The Thai cannabis law has not as yet impacted on the legal position regarding drugs in the Philippines. There was no formal statement but the new President of the Philippines, Ferdinand Marcos Jr., has pledged to take a softer stance on illegal narcotics than his predecessor. However, the government made it clear that cannabis possession will remain illegal in the Philippines. In 2002, the Comprehensive Dangerous Drug Act was passed, which classified cannabis as a dangerous substance. Possession and use of cannabis are therefore regarded as illegal in the Philippines, with possible punishments including life imprisonment or the death penalty . According to the findings, the mainland and insular states of SEA are currently engaging with prepared to face the new cannabis law in Thailand. Cannabis culture will continue to exist as a counterculture in all SEA states for the foreseeable future. However, mainland SEA states such as Myanmar, Laos, Cambodia and Vietnam may have the propensity to tolerate and observe Thai cultural characteristics within their own territorial boundaries. The degree of these states' acceptability of cannabis use can be deduced from their official proclamations that have been disseminated through various media outlets. However, it does not appear that the insular SEA states are prepared for the establishment of such subcultures, and it is apparent that they will continue to categorise cannabis-related activities as those belonging to countercultures. There is a chance that the role played by politics and religions is the essential factor that requires further examination in future studies of this matter. For example, the Philippines and Vatican are the only two independent jurisdictions in the world where divorce remains illegal under their respective legal systems. According to the Philippines Commission on Women , the Family Code enacted in 1987 authorises relative divorce or legal separation but still imposes no regulation on divorce per se. According to Human Rights Watch , Brunei's Sharia law requires the death penalty for offences such as having sex with a nonspouse or having anal intercourse . Therefore, with these strong markers of social conservatism in mind, it would be difficult for these insular states to ease their cannabis regulations. Consequently, in the future, there may be significant geopolitical issues in SEA as there could be differences between the two sub-regions, as governments on the mainland may adopt a more relaxed and liberal attitude on their cannabis regulations, while those on the insular may take the opposite posture. Therefore, the ASEAN Framework Agreement on Visa Exemption could be stalled or revised as SEA states could face severely divided views on cannabis subculture and weed tourism in Thailand, which could also jeopardise ASEAN and the region's stability and prosperity. --- Conclusion The SEA region is divided into mainland and insular sub-regions. The ten member states of the ASEAN region have always been known for opposing social liberalisation and upholding traditional values. Thailand decriminalised cannabis in June 2022, permitting cannabis-infused foods, beverages, and medical use of the drug widely throughout Thailand. In contrast, other SEA nations, prohibit the usage of cannabis in their states and consider cannabis use to belong to a counterculture in Thailand. This study assessed the impact of this law change on SEA subregions. This study addressed the question of whether Thailand's recent cannabis law could influence other SEA nations. Due to the novelty of this research, qualitative and digital ethnography were employed. The approach included the study of content extracted topics from secondary government data. The findings of this research suggested that mainland states may accept and apply their existing cannabis counterculture as a subculture. However, Thailand's cannabis subcultures and weed tourism in Thailand shall remain countercultures in the insular states' region. Given this point regarding the relevant geopolitical differences, such as the acceptance and tolerance of cannabis culture between insular and mainland states, it is stated that the issue could pose a future challenge to the prosperity of this SEA region.

Southeast Asia (SEA) is geographically divided into two subregions, mainland and insular. This region consists of 11 states that consistently oppose social liberal changes and uphold traditional values. Conservatism in politics, economy and society, including the legal realm, is always favoured in this region. However, recent developments in Thailand could test the conservative beliefs and practices in this region. In June 2022, Thailand decriminalised the cultivation and consumption of cannabis. This new law allows people in Thailand to consume marijuana edibles, marijuanainfused drinks and use marijuana for medical purposes. This is not the first time Thailand has liberalised its laws. For example, unlike other SEA countries, Thailand does not regulate the use clothing in private spaces; therefore, there has been a significant increase in the number of clothing-optional resorts in Thailand over recent years. Meanwhile, in 2015, Thailand became the first SEA country to recognise and protect the rights of Lesbian, Gay, Bisexual, and Transgender (LGBT) individuals. Such laws and tolerance have now become part of Thailand's practices and subculture but remain taboo in other SEA states. Hence, this research has investigated the impact of developments in Thailand's subculture on the SEA subregions, with a particular focus on cannabis law reform. Therefore, this research addresses the primary research question: in the context of 'weed tourism', could Thailand's recent cannabis culture impact other Southeast Asian countries? Due to the infancy of this research, a qualitative and digital ethnographical research method was chosen. The thematic analysis was adopted from the secondary government-published data by eliciting key themes. The findings revealed that weed tourism activities would encourage the people in this region to understand, respect, or even practice Thailand's subculture themselves. The mainland states may tend to accept and observe Thailand's cannabis subculture in their states. However, the insular states will not accept such subcultures and will always consider such activities as countercultural.

INTRODUCTION Globally, 83% of births in 2020 occurred with skilled birth attendance , but coverage continues to be uneven around the world with significant discrepancies between regions with only 64% of births in sub-Saharan Africa being attended to by SBA. 1 About 303 000 maternal deaths are registered annually with 99% being recorded in low-income and middle-income countries. 2 3 SBA has been documented as an effective intervention for reducing maternal and neonatal deaths. 4 5 Skilled attendance at birth can reduce intrapartum-related complications by up to 20%. 6 Therefore, ensuring increased utilisation of SBA can substantially contribute towards achievement of the "Sustainable Development Goal 3 that aims at reducing the global maternal mortality ratio to less than 70 per 100, 000 and neonatal mortality ratio of ≤12 per 1000 live births by 2030''. [6][7][8] A skilled birth attendant is 'an accredited health professional such as a midwife, doctor, or nurse who have been trained with adequate skills needed to handle uncomplicated pregnancies, childbirth and the immediate postnatal period, and in the identification, management, and referral of complications in women and newborns'. 6 Besides the women losing their lives, effects of maternal mortality and morbidity are also experienced at the household and community level. 9 10 Children left behind after maternal deaths have increased odds of mortality or other health challenges including undernutrition and the society loses resources when women die in their most productive years. 9 In Sierra Leone, pregnancy is associated with a 1 --- Strengths and limitations of this study ► This is the first nationally representative analysis that explores the rural-urban correlates of skilled birth attendance in Sierra Leone. ► We used the latest nationally representative sample from the 2019 Sierra Leone Demographic and Health Survey, hence findings are generalisable to women in Sierra Leone. ► Given the cross-sectional nature of the data, we could not establish the temporal relationship between the outcome variable and the independent variables. ► Since the data were collected from women who had childbirths within 5 years prior to data collection, we anticipate recall bias in the process of collecting this data among the respondents. Open access in 17 lifetime risk of maternal death making it among the highest globally. 6 Despite several measures being implemented in the country, utilisation of maternal health services such as utilisation of at least four or more antenatal care contacts marginally increased by three percent points between 2013 and 2019 while initiation in the first trimester decreased by 1% point . 11 In 2017, the Ministry of Health adopted the latest 2016 WHO guidelines for ANC, recommending eight or more ANC contacts during pregnancy. 12 To date, there are no data available about the progress made regarding the utilisation of eight or more ANC contacts. The latest Sierra Leone Demographic and Health Survey only reported on the utilisation of at least four ANC contacts. 11 Postcivil war and Ebola epidemic Sierra Leone era has witnessed left a fragile health system having poor infrastructure and inadequate skilled health personnel who are irregularly paid low salaries. 13 Despite the government's efforts to improve maternal health with approaches such as exemption of user fees for maternal healthcare services, 14 the country ranks among the top three countries with the highest MMR, globally. 3 6 15 Furthermore, the exemption of user fees is challenged by inadequate skilled health personnel, increasing workload and inadequate supplies and equipment. 16 17 Secondary and tertiary care in Sierra Leone is provided by 14 district and regional governmental hospitals. 17 At national level, there are four tertiary referral hospitals which are all located in the Western Area Urban District. 18 The country has one of the lowest nurse densities in the world, at approximately 0.2 nurses and midwives per 1000 people. 13 Although differences in the levels of utilisation of SBA between Sierra Leone's rural and urban women have been documented, 6 11 there is a paucity of information on this topic as it is not adequately explored. Therefore, it is important to further understand these factors when stratifying by rural-urban place of residence among women because this may be key to designing effective contextspecific strategies and interventions targeting rural and urban areas. We aimed to determine the correlates of SBA in Sierra Leone, stratified by rural-urban place of residence. --- METHODS --- Data source Secondary data from the 2019 SLDHS was analysed for this study. SLDHS data collection occurred between May and August 2019 by Statistics Sierra Leone with technical assistance from Inner City Fund international through the DHS programme. --- Study sampling and participants A stratified, two-stage cluster sampling design was used for the survey leading to 13 872 households. 11 The 2019 SLDHS final report contains a detailed description of the sampling procedures. 11 19 Women of reproductive age who had a live birth within 5 years preceding the SLDHS were included in this secondary analysis. Originally, a weighted sample of 15 574 women was included in the individual women's data set of which 7326 had given birth within 5 years prior the survey , 3 as shown in online supplemental file 1. --- Variables --- Dependent variables SBA was defined as delivery conducted by a doctor, nurse or midwife 11 and was coded as 1 while unskilled birth attendance was coded as 0. --- Independent variables The analysis included independent variables based on evidence from available literature and data. 6 9 20 Sixteen explanatory variables were included and categorised as shown in table 1. --- Statistical analysis Due to the multistage cluster study design used by SLDH, complex sample package of SPSS statistical software was used with the analysis plan designed to include sample : individual weight, strata for sampling errors/ design and cluster number. [21][22][23] Associations between independent variables and SBA were assessed by cross tabulation and p values presented. Before the final adjusted model, each independent variable was assessed individually for its association with SBA using bivariable logistic regression and the crude OR, 95% CI and p values are presented and independent variables with a p≤0.25, and not strongly collinear with other independent variables were included in the final multivariable logistic regression model. 24 In the final adjusted model, adjusted ORs , 95% CI and p values were calculated at significance level set at p<0.05. Online supplemental file 2 shows the Strengthening the Reporting of Observational Studies in Epidemiology checklist. Sensitivity analysis was done with unskilled birth attendance as the outcome and the results are shown in online supplemental file 3. --- Patient and public involvement Patients were not involved. However, local authorities in the different regions were contacted before data collection. A comprehensive report on the survey results was released and openly available on the DHS website. 11 --- RESULTS Table 2 shows a comparison of background characteristics of study participants. Rural areas had more participants compared with urban areas . Remarkable differences were observed in region with 1.1% of rural women residing in Western region compared with 51.1% in urban areas. Furthermore, 63.2% of rural women had no education compared with 35.5% in urban areas, 34.8% in rural areas belonged to the poorest quintile compared with 0.4% in urban areas and 36.2% had exposure to mass media in rural areas compared with 69.7% Open access in urban areas. Over 60.3% of rural women had big problems with distance to the nearest health facility compared with 25.8% in urban areas. Overall, 88.3% of the women had SBA. SBA was higher in urban areas at 94.9% compared with 84.2% in rural areas. Factors associated with SBA Tables 3 and4 presents the predictors of rural and urban SBA. Our analysis revealed that region of residence, exposure to mass media and distance to the nearest health facility have significant positive association with SBA among women from both regions of residence. In the rural areas, the likelihood of being delivered by a skilled birth attendant was three times higher in the Southern , Northern and six times higher in the Eastern regions , one and a half times higher among women who had been visited a field worker , two times higher among women with postprimary education , one and a half times higher among women with exposure to mass media , twice higher among women not having big problems with distance to the nearest health facility while the likelihood was 0.8 times lower among women who initiated ANC after the first trimester . In the urban areas, the likelihood of being delivered by a skilled birth attendant was five times higher in the The SLDHS collected data on household asset ownership and calculated wealth index using Principal Component Analysis. 68 Among rural women, only 0.9% and 5.7% belonged to the richest and richer quintiles, hence these were combined into one to have rich, middle, poorer and poorest quintiles in logistic regression. Among urban women, only 0.3% and 3.0% belonged to the poorest and poorer quintiles, hence these were combined into one to have poor, middle, richer and richest quintiles in logistic regression. --- Region Northern, Eastern, Southern, Western and Northwestern Among rural women, only 1.1% belonged to the Western region hence in logistic regression, Western and Northwestern regions were combined. --- Education No education, primary education, secondary and tertiary education Among rural women, only 0.5% of the women had tertiary education and only 7.1% in urban hence secondary and tertiary were combined to have postprimary in the logistic regression analysis. --- Household size Less than seven members and seven and above members Based on the dataset average of seven members per household --- Sex of household head Male or female Marital status Married and not married Marriage included those in formal and informal unions while not married included the never married, divorced, separated and widowed. --- Religion Muslims and Christians and others Problem seeking permission to access healthcare --- Big problem and no big problem In the original SLDHS questionnaire, three responses had been suggested: no problem, no big problem and big problem. However, the no problem response was not reported by anyone . --- Difficulties accessing nearest health facility --- Big problem and no big problem In the original SLDHS questionnaire, three responses had been suggested: no problem, no big problem and big problem. However, the no problem response was not reported by anyone --- Open access Southern , 12 times higher in the Eastern region , one and a half times higher among women from households with less than seven members , twice among women who had exposure to mass media and one and a half times among women who had no big problems with distance to the nearest health facility compared with those from the western and northwestern regions, households with seven and above household members, with no mass media exposure and those with big problems with distance, respectively. Wealth index was imprecisely significant with urban women belonging to the richest quintile being more likely to have SBA compared with those in the poor quintile. --- DISCUSSION In this study, we looked at factors associated with SBA utilisation in Sierra Leone stratified by rural-urban place of residence. Overall, 88.3% of the women had SBA. The overall, urban, rural and SBA prevalence in our study shows 28, 15 and 31 percentage point increases respectively compared with that of 2013. 6 25 This shows a tremendous improvement in the uptake of the SBA between 2013 and 2019 in Sierra Leone which could Open access Open access be attributed to the changes in health-seeking behaviour and transformation of the health systems witnessed after the Ebola epidemic. 26 27 The introduction of free maternal healthcare services in 2010 could also partly have contributed to the observed increase in SBA utilisation. 28 29 SBA was higher in urban areas at 94.9% compared with 84.2% in rural areas. Higher SBA utilisation among urban women has also been shown by Ameyaw and Dickson 6 and this could be partly explained by factors such as the huge negative effects of the conflict on the rural healthcare system, high concentration of health centres and hospitals and healthcare workers in urban areas enabling easier access to maternal healthcare services. 6 30 31 Higher SBA utilisation among urban women compared with rural women has been shown in several other studies. [32][33][34] The mismatch between high coverage of SBA and the persistently high numbers of maternal and perinatal deaths is not only unique to Sierra Leone. This may be partly attributed to delayed seeking of childbirth care and inadequate quality of care provided by skilled birth attendants. [35][36][37] Available evidence from similar low resource settings in sub-Saharan points towards poor quality of services offered. 29 38 The inadequate quality of care may be attributed to factors such as; poor remuneration which demotivates health workers, increased workload on health workers, lack of essential drugs and low quality pre-service and refresher training. 36 37 In Sierra Leone, preservice training for SBAs produces three cadres of nursing staff, namely; maternal and child health assistants who train for 2 years, state enrolled community health nurses spend two and half years in training, and state registered nurses whose training lasts 3 years. These cadres then have the option to undertake further midwifery training that lasts between 18 and 24 months depending on the nursing qualification and experience. 39 40 However, the quality of training is affected by factors such as; poor student attendance, delayed and low tutor allowances and poor schools' infrastructure especially for rural training schools. 30 40 Region of residence, exposure to mass media, and distance to the nearest health facility had higher likelihood of SBA uptake in both rural and urban areas. Household size was only significantly associated with SBA in urban areas while being visited by a fieldworker, level of education and timing of initiation ANC were only significant in rural areas. Being a resident of the South, the Eastern and Northern regions was associated with more odds of SBA utilisation among rural areas compared with those in the Western and North-western regions which was a similar finding for urban women in the Eastern and Southern regions. This is an unexpected finding since the Western region has the highest concentration of skilled personnel and health facilities, the most developed and is the most economically vibrant region and therefore has better quality social amenities compared with other regions. 28 30 However, the Western areas have witnessed increasing numbers of urban poor who are experiencing high standards of living and inequitable distribution of social amenities hence negatively affecting their ability to access quality healthcare. 41 42 Furthermore, the documented staff challenges in urban areas such as poor delegation, favouritism and a lack of autonomy could partly affect quality of services in public health facilities which further limits utilisation of healthcare. 28 30 The government's efforts to ensure better service delivery in the less developed regions that are far away from the developed Western region could also have contributed to this observation. 12 Region has been documented to have an association with SBA in other studies. 43 Exposure to mass media was associated with more odds of SBA utilisation in both rural and urban areas. Mass media have been documented to improve health literacy Open access by sensitising communities on the positive outcomes of timely healthcare seeking and utilisation hence leading to positive attitudes, challenging negative social norms and improving health seeking behaviour. 44 45 Furthermore, women who are exposed to mass media are more likely to be educated, have discussions with their peers which interpersonal interactions contribute greatly in challenging negative norms that might affect health seeking and hence lead to positive health seeking behavioural change. 46 47 Hence, enhancing mass media exposure can be used to provide targeted maternal health messaging that can lead to increase in the utilisation of SBA. 48 Exposure to media has been shown in previous studies done in similar contexts to have a positive association with SBA. 6 49 50 Rural and urban women who reported that distance to health facilities was not a major challenge had higher odds of SBA utilisation. Our study observed that the mothers in rural areas and urban areas who had no big problem with distance to a health facility had 2.25 and 1.62 higher odds, respectively, of being attended to by a skilled birth attendant compared with their counterparts who had challenges of distance to the nearest health facility. The strong association between distance to health facility and SBA utilisation among the rural mothers compared with urban can be partly explained by the fact that rural areas of Sierra Leone have poor road networks compared with urban areas with most roads being only accessible by offroad vehicles or motorbikes. This is further compounded by the lack of access to affordable transport and health facilities that far apart from each other, which contributes to delays faced by women in rural areas. 35 51 Distance to health facilities has been shown to impede access to maternal child health services including SBA in several other studies. 9 52 53 Unlike in urban areas, being visited by a field health worker, such as a community health worker among rural women was significantly associated with SBA utilisation. The high demand of CHWs in rural areas due to limited accessibility of healthcare because of shortage of health facilities and large distances needed to be covered by rural women 30 51 compared with easier access of health facilities in urban areas could partly explain the observed difference in association. The increased SBA utilisation among rural women who were visited by field health workers could be partly explained by the fact these field health workers equip mothers with knowledge on the dangers of using unskilled birth attendants and complications of pregnancies in addition to encouraging them to seek care within health facilities. 54 Being visited by field health workers has been shown to be associated with SBA in several other studies. 55 56 Level of education was significantly associated with SBA in rural areas but not urban areas. Women with post-primary education had more odds of SBA utilisation compared with women with no education. Educated women are believed to easily understand counselling given from healthcare workers, more health literate hence informed on obstetric danger signs, which enables them to seek early maternal healthcare. 48 Educated women have also been shown to develop greater confidence, be more conscious of their health and better abilities to make wise decisions about their own health, hence better SBA utilisation. 6 9 Furthermore, higher levels of education have an influence on women's positive interpretation of mass media messages leading to positive healthcare seeking behaviour change. 48 In predominantly patriarchal African societies and mainly in rural areas, 57 men are the main providers with the highest decision making powers. 58 Women in rural areas are usually Open access less empowered due to the more conservative societies in rural areas hence factors such as education that might increase women's status and decision making are more likely to have an impact on healthcare seeking. [59][60][61][62] This might partly explain the significance of education in rural areas and the non-significance in urban areas. Our findings indicate the need for government to strengthen access to quality girl child education among rural areas to at least secondary school level. Level of education has been shown to be associated with SBA utilisation among several other studies. 48 63 Delayed initiation of ANC among rural women was associated with less odds of SBA utilisation. ANC utilisation has been shown to be associated with several other studies. 48 52 64 Delayed initiation could partly reflect poor health seeking behaviour which is further observed by reduced odds of SBA utilisation. However, there is need for further studies to explore the association of ANC utilisation and SBA given the fact that ANC frequency was not significantly associated with SBA but timing of ANC initiation was. Besides the three factors that were significant in both rural and urban areas, household size was the only factor that showed significance in urban areas. Women who belonged to households with less than seven members had more odds of SBA utilisation compared with their counterparts. This is in agreement with a study done in Nigeria and India. 65 66 Although wealth index was marginally significant in urban areas, women belonging to the richest wealth quintile had 2.5 odds of SBA utilisation compared with their counterparts in the poorest households. We hypothesise that families with smaller sizes tend to have less expenditure which enables savings that can be used for the direct and indirect costs involved in accessing healthcare. 66 Furthermore, smaller sizes could be attributed to better maternal healthcare seeking such as modern contraceptives utilisation which is further translated into SBA utilisation. 66 Lastly, having smaller family size might lead to less time spent by women while doing household chores and providing care to other family members and increase their time to seek healthcare. 67 However, given the dearth of information regarding household size and SBA utilisation, we recommend further studies to explore this. --- Strengths and limitations The study used a nationally representative sample for the analysis and thus the results can be generalised to all Sierra Leone women. Since the data was extracted from DHS surveys, we are confident that standardised procedures such as validated questionnaires were used in data collection to ensure the validity of the results. This being a cross-sectional study, this creates a limitation in establishing casual relationships from the established associations. In addition, since most of the data was for women who had childbirths within 5 years preceding the survey, we anticipate recall bias in the process of collecting this data among the respondents. --- CONCLUSION AND PUBLIC HEALTH IMPLICATIONS In Sierra Leon, SBA utilisation has greatly improved in the last decade. Utilisation is higher in the urban compared with the rural areas. Region of residence, exposure to mass media, and distance to the nearest health facility had a significant association with SBA uptake in both rural and urban areas. Household size was only significantly associated with SBA in urban areas while being visited by a fieldworker, level of education and timing of initiation ANC were only significant in rural areas. Hence ensuring context specific policies and strategies is crucial to ensure effective SBA utilisation. Generally, maternal stakeholders need to focus on Western region, use of mass media for awareness and sensitisation and ensuring increased availability of affordable and accessible health facilities in both rural and urban areas. In addition, urbanspecific programmes need to focus on women residing in larger households and rural specific programmes need to focus on use of field health workers, women educated to primary level and below and ensuring timely initiation of ANC services. Further research is need to explore reasons why maternal mortality is high despite the high SBA focusing on areas such as quality of care provided. --- Contributors QS conceived the idea, drafted the manuscript, performed analysis and interpreted the results. IM, KK and MWM reviewed and interpreted the results, reviewed the first draft and drafted the subsequent versions of the manuscript. All authors read and approved the final manuscript. QS is responsible for the overall content as the guarantor. Competing interests None declared. Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. --- Patient consent for publication Not applicable. Ethics approval This study involves human participants and was approved by High international ethical standards are ensured during MEASURE DHS surveys and the 2019 SLDHS protocol was reviewed and approved by the Sierra Leone Ethics and Scientific Review Committee and the ICF Institutional Review Board. Besides, the local authorities before implementing the survey and well-informed verbal consent are sought from the respondents prior to data collection. This data set was obtained from the MEASURE DHS website after getting their permission, and no formal ethical clearance was obtained since we conducted a secondary analysis of publicly available data. Note: The SLDHS report does not provide the IRB approval number 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 may be obtained from a third party and are not publicly available. All data are available from the Demographic and Health Surveys website upon registration. 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 Understanding the rural-urban contextspecific correlates of skilled birth attendance (SBA) is important to designing relevant strategies and programmes. This analysis aimed to assess for the ruralurban correlates of SBA in Sierra Leone. Setting The latest nationally representative Sierra Leone Demographic and Health Survey of 2019. Participants The study included a weighted sample of 7326 women aged 15-49 years. Each of them had a live birth within 5 years prior to the survey (4531 in rural areas and 2795 women in urban areas). Primary and secondary outcome measure SBA (primary) and predictors of SBA (secondary). Results SBA was higher in urban areas at 94.9% (95% CI 94.1% to 95.7%) compared with 84.2% (95% CI 83.8% to 85.9%) in rural areas. Rural women resident in the Southern, Northern and Eastern regions, with postprimary education (adjusted OR (aOR) 1.8; 95% CI 1.3 to 2.5), exposure to mass media (aOR 1.5; 95% CI 1.1 to 1.9), not having difficulties with distance to the nearest health facility (aOR 2.3; 95% CI 1.7 to 3.0) were associated with higher odds of SBA. Urban women resident in the Southern, Eastern region, with households having less than seven members (aOR 1.5; 95% CI 1.1 to 2.3), exposure to mass media (aOR 1.8; 95% CI 1.1 to 2.9) and not having difficulties with distance to the nearest health facility (aOR 1.6; 95% CI 1.1 to 2.5) were associated with higher odds of SBA. Conclusion Given the observed differences, improving SBA requires programmes and strategies that are contextspecific.

Significance Clinical domains other than obstetrics have recognized the importance of care management interventions to reduce hospital emergency department visits. While recent literature has identified clinical predictors of postpartum ED visits, interventions to reduce unnecessary ED visits and improve postpartum health for high-risk women are limited. This study reveals opportunities for early intervention to reduce the widening gap of maternal morbidity and mortality for --- Introduction Racial and ethnic disparities in maternal health are a public health crisis in the United States. Black women are two to three times more likely to experience severe maternal morbidity and three to four times more likely to suffer pregnancy-related mortality compared to non-Hispanic white women . These elevated health risks exist across the pregnancy continuum and extend into the postpartum period. Black women are more likely to have an emergency department visit after delivery , and Black and Latina women have a two-fold increased risk of postpartum hospital readmission relative to white women . Postpartum ED visits and hospital readmissions are markers of the excess morbidity Black and Latina women experience and studies indicate that some instances of emergency service utilization following delivery are preventable . The postpartum period presents a critical opportunity to improve outcomes , given that 51% of maternal deaths occur within 1 day to 1 year after delivery . Recent value-based payment models have focused on moving care upstream and centering intervention on providing incentives to increase adherence to the 6-week postpartum visit. However, these interventions may have missed the mark as many women experience health problems long before the 6-week time point, resulting in increased emergency department visits . In fact, studies indicate that 60% of all maternal ED visits occur prior to 6 weeks postpartum . Understanding trends in postpartum ED utilization is central to addressing disparities as a marker of poor maternal health, future morbidity, and pregnancy's long-term impact on the health and wellbeing of new mothers and their families. Clinical risk factors for postpartum complications and ED usage have been extensively described . Yet, the current body of literature lacks the patient perspective on what brings women to the ER, and whether opportunities for intervention existed along their journey. Over the last decade, research has focused on understanding the role of social determinants of health -the conditions in which people are born, grow, live, work, and age-in relationship to health outcomes . However, increasing evidence argues that addressing SDoH alone does not completely improve outcomes for women of color because it fails to recognize the larger systemic context that serves as the primary driver of social inequity . Newer approaches suggest investigating the structural determinants of health, defined as the cultural norms, policies, institutions, and practices that define the distribution of SDoH, as the root cause of health disparities and a primary driver of SDoH . The historical allocation of greater resources to white communities while disproportionately depriving communities of color is an important lens to consider in maternal morbidity and mortality outcomes because it allows us to understand how racism, classism, and sexism intersect to affect the lives of women of color. Identifying key factors that lead to increased rates of postpartum ED visits can inform quality improvement efforts. However, a more thorough investigation into the patient lived experiences driving ED visits, especially for disadvantaged women of color, is needed. The objective of this study is to explore the postpartum experiences of publicly-insured women of color to identify additional risk factors that may contribute to postpartum emergency room usage and identify potential opportunities for intervention. --- Methods We used purposive sampling to recruit postpartum women into four focus groups in the summer and fall of 2020. All women were eligible to participate if they spoke Spanish or English, ≥ 18 years old, delivered an infant between June 1, 2019 and May 1, 2020 at Mount Sinai Hospital, and had Medicaid coverage for their delivery. Eligible participants were identified using either an EPIC query or outreach to the Ambulatory Clinic by posting IRB approved recruitment fliers in the waiting room. All participants verbally gave their informed consent by phone with use of a witness prior to inclusion in the study and received a copy of the consent form via mail or email. Participants also completed a brief anonymous demographic survey via email and mail. Participants were offered a $100 gift card and round-trip Met-roCard for compensation of their time upon completion of the focus group. Race/ethnicity-concordant moderators led the focus groups. Three English focus groups were conducted in July and August 2020 and one Spanish focus group was conducted in October 2020. Due to COVID-19 health and safety concerns, the focus groups were held virtually using a HIPAA compliant platform. Approval of this study was obtained by the Program for the Protection of Human Subjects at the Icahn School of 1 3 Medicine at Mount Sinai School. This study was conducted in accordance with the COREQ criteria for reporting qualitative research . A moderator guide containing open-ended questions was used by the moderators to ensure consistency among focus groups and techniques such as reflection were used to clarify statements. The moderator guide was developed by the research team based on prior research and published literature. The same questions were asked in each focus group and were only slightly modified to follow the natural flow of the conversation. Each focus group lasted approximately an hour in length. The focus groups were both video and audio-recorded and transcribed in their entirety. The Spanish focus group transcript was professionally translated into English and bilingual members of the research team verified the accuracy of the translation. Following the focus groups, the research team debriefed about initial thoughts and impressions. The moderator then wrote detailed notes summarizing each focus group, including verbatim quotes from the transcripts. We used a combined inductive and deductive approach to analyze the data. First, we developed deductive codes based on the team's impressions from the focus groups and the moderator guide topics. Iterative review of the transcripts and coder triangulation were used to thematically analyze the data. Members of the research team individually reviewed the transcripts and then met to create an inclusive master themes list. Two independent readers then coded the transcripts and a third reader reviewed the coded transcripts for inter-rater agreement. Disagreements were then discussed and resolved. Dedoose software was also used to facilitate data management and retrieval . --- Results Eighteen women participated in one of four focus groups . Majority of participants self-identified as Black/African-American and were over the age of 30. Of eighteen participants, five women had an ER visit within 30 days of delivery. Reasons for ER visits included preeclampsia, postpartum cardiomyopathy, and infection. Four major themes and 13 sub-themes were identified that described women's postpartum experiences . The four major themes included: lack of access to and communication with a medical team, lack of preparation, importance of social support and participant recommendations for improving quality of care. Each theme is discussed below. Tables 4,5, 6, and 7 include the themes and representative quotes from participants. --- Lack of Access to & Communication with Medical Team Participants reported varying degrees of accessibility and communication with their medical team following delivery. Some women reported no problems with scheduling appointments. Other participants, particularly women who had postpartum complications, identified hardships 1 3 in accessing prompt care following delivery. Many women described feeling like a "bother" or "pain in the butt" to their doctor when calling to ask questions. As one woman explained: It's hard getting in contact with someone when you're going through something. Sometimes you call and they tell you, "we are going to send the message and the doctor's going to call you back within X amount of days. But sometimes they don't call you back and it's like I have this thing going on, I'd rather be seen. [FG2,participant with ED visit] Participants also spoke about their concern with not having a single team to watch over them prenatally, during delivery, and after birth. Due to being Medicaid-insured, most participants had rotating providers throughout their pregnancy course and postpartum. For one participant who experienced an adverse health event, inability to identify a single provider made the emergency room seem like the next-best option to receive the immediate care she needed. Women also described having rotating providers as a barrier to building trusting relationships with their physician. Participants described being "a part of [a] huge system" that prevented development of emotional connections to their doctors, and instead relying on visiting nurses and midwives for support. Additionally, some women reported racism and bias in their interactions with healthcare providers throughout prenatal care and during delivery, which dampened their communication and desire to reach out to them. Prompts about racism and bias sparked a robust conversation in the Spanish-speaking focus group, with one woman remarking: When asked directly about racism and bias, some participants described providers as "very kind" and the services they received as "excellent." Others emphasized the structural barriers they experienced due to being low-income and Medicaid-insured, such as lengthy clinic wait times and financial disadvantages, that made it difficult to access care. One participant compared her experience being Medicaidinsured with being privately insured, stating: The first time I gave birth, I had insurance through work…But I lost my insurance and this time I delivered with Medicaid…I never felt like I had a team of people watching. Even while I had the gestational diabetes, it was a total experience from the last time because I had a specific doctor that only looked at me. [FG2, participant with ED visit] --- Lack of Preparation for Postpartum Period A general response among all focus groups was feeling unprepared for what to expect following delivery. Some participants specifically emphasized being ill-prepared for the emotional aspect, while others spoke about underestimating the lifestyle changes of bringing home a new child. One participant stated, "nothing can really prepare you for what is about to go down" [FG3, participant without ED visit]. Participants who had experienced a postpartum complication tended to recall the fear they felt during the postpartum period and beyond by using words such as "terrifying" and "scary" when recounting the post-delivery experiences with which they felt unequipped to deal. Women stated that they did not receive enough health education about what to expect or look out for during the postpartum period, which contributed to feelings of Women, both who did and did not have an ED visit, expressed difficulties in retaining information immediately following delivery. Women repeatedly described it as a "blur" and therefore advised against verbal education during clinic or hospital visits as the main educational tool. It was very awkward and I didn't want to speak to that man after that ii. I feel like a lot of women don't get services because of financial reasons or don't have the insurance that covers those types of things. Services should be offered to women and let them decide if they would like to decline iii. I contacted the doctors so much in the hospital that someone spoke to the social worker, and said I should see someone else, like a psychiatrist or something Participants also spoke about competing demands at the time of delivery such as other children or wanting to return home from the hospital, that made it difficult to focus on the information that was being provided. --- Importance of Social Support to Postpartum Success The value of social support spanned all focus groups. Women either described feeling grateful for the support they received, or reported feeling overwhelmed because support was lacking. One participant spoke directly about the challenges of asking for help as a new mom, stating: "there is no longer a village" to help with your child [FG1, participant without ED visit]. Some participants stressed the importance of identifying a close support person during pregnancy to prepare for unexpected events in the postpartum period. Women discussed how their partners, mothers, mothers-in-law, sisters and older children helped them in the first few weeks or months after coming home from the hospital. Most participants spoke about the important role of partners in coping with new responsibilities during the postpartum period. For women who needed to go to the ER, partners played an essential role as family caretakers. However, not everyone had a husband or a partner to help them. For example, one woman mentioned that she split from her partner right after the baby was born and another had a husband who lived in another country. Other women expressed disappointment of unmet expectations when family members were not supportive or able to help. [on] how to be a parent. I feel like the support system is very very important in order to raise a child, have a child, and take care of the child. Even if you went to classes, it is very difficult. So have a good support system Subtheme 3.2: Role of partners i. [W]hen we came home, my husband was here… we were doing it together, like we were a team, he was helping a lot with the baby ii. My husband also took me to the Emergency Room. It was just him who was with me, and he was in charge of rushing home to see my eldest children, and rushing back to the hospital and to be with me, just him Subtheme 3.3: Unmet expectations i. I live with everyone at home, my mom, my dad, and my siblings. And at that time, when I gave birth, I'd have liked some of my family to help me the day my child went to sleep, maybe I could go to sleep too… but for some reasons, they had to work, each of them had to do their things, they… didn't do it ii. Certain expectations that I had when they didn't happen, that was really disappoint [ing] Participant Recommendations for Improving Quality of Care Women in each focus group were asked to provide direct feedback on how postpartum care could be improved. In this section, we provide a high-level summary of the identified opportunities for intervention that were consistent across focus groups. Several of our participants mentioned that enhancing health education by providing adapted material formats for both themselves and to share with caregivers could be helpful, particularly in printed form to take home or an easily accessible electronic backup in case the printout gets lost. One participant expressed a desire for more health education about possible warning signs. Women in our focus groups spoke about the need for health care providers to engage caregivers and family members from the start. Although partners and family members played an important support role for many of our participants, they were not always knowledgeable or helpful. According to one woman, "...maybe, even if it is a very severe case like the one I had, maybe if those things would have been listed, maybe my husband would have read it and kept it." [FG2, participant with ED visit] Aside from family members, one participant mentioned seeking support from peers who experienced a similar postpartum complication. Another woman agreed that creating informal, peer support groups of women who are going through the same experience can be helpful and indicated that participating in a focus group gave her a different perspective. One participant described a positive experience with being part of a group of women treated in the same clinic who were expecting around the same time. However, the continuation of such connections into the postpartum period remains unclear. In closing remarks, multiple participants added that providers should "just listen to women more." --- Discussion Our study identified the following factors that influenced women's postpartum experiences: access and communication with their medical team, perceived level of preparedness for the postpartum period, and social support network following delivery. While some women had good communication with their providers, many women reported barriers due to rotating providers and racism/bias. Generally, Table 7 Participant recommendations for improving quality of care Subtheme 4.1: Enhancing health education i. …maybe, even if it is a very severe case like the one I had, maybe if those things would have been listed, maybe my husband would have read it and kept it ii. Pay attention to moms postpartum. Providing resources or information right before you give birth or while they are in the hospital, 'this is what you need to be careful of.' A list of things that you could use or that may use could be life changing. Critical for mom iii. Come up with the right kind of resources, whether it's pamphlets, whether it is [an] online app Subtheme 4.2: Engage caregivers i. …in case you are ill, the family already knows what you have, and you feel that support because the doctor… already told them that she needed that support on this, this and this. Or for example, when you have an appointment, that your relative doesn't stay outside but that they may enter with you and listen to your opinion ii. it's a good initiative to do at least a kind of talk, in order to involve the family member at the beginning and near the end of the pregnancy, so that they can understand that there is a new member coming into the family, that is, that everyone must collaborate, that everyone must be supportive. Because, sometimes, as a new mother, as it is my case, I didn't know if I am doing well with the baby, or I am feeling very tired, or I need someone to tell me "I am going to stay with you tonight to watch your baby". I do think that family members should be more involved directly from the center itself, so that family members feel more engaged than if I tell them myself." Subtheme 4.3: Support groups i. I felt like I had enough support at home but to be connected to other women who are actually going through it. Because to be honest it's hard for me to find anyone in New York that has a group like that. And it would've been nice I guess to be connected. With someone who has experienced it or is going through it right now ii. So I did… my doctor visits with a whole bunch of other girls who were expecting around the same time. So, I was supposed to give birth in June. So, my doctor was really really available when it came to emailing and answering questions. If I didn't ask her, I'd ask my nurse. If it wasn't the nurse then I would look for the mommy, the mommy, the little groups that were of women pregnant at the same time as well to ask them questions Subtheme 4.4: Listen to women i. A doctor in the room would not listen to me. I kept telling him, I don't want forceps. He got frustrated ii. One thing doctors can do is listen more. A lot of simple thing[s] but a lot of doctors are not doing it all women felt a level of unpreparedness for the postpartum period due to ineffective education methods. This lack of knowledge often resulted in fear for women who experienced warning signs prompting ED use. Social support networks were essential to postpartum satisfaction, and greatly influenced women's ability to juggle the changing responsibilities of bringing home a new baby. Our results are consistent with previous studies that have found that primary care accessibility, health education, and social support are critical factors to decreasing emergency service utilization and preventing adverse health outcomes among vulnerable populations . However, our study adds that addressing these factors is particularly important in reducing postpartum morbidity and mortality among publicly-insured women of color due to the baseline systemic inequities that exist as a consequence of racism, classism, and sexism. To our knowledge, this study is the first to draw directly on the narratives of Black and Latina women to identify non-clinical risk factors that may be important for postpartum ED utilization. Although we expected to find notable psychosocial differences between women who visited the emergency department and women who did not, each group had comparable experiences and shared similar concerns. Instead, our results indicated that the upstream structural factors of their experience due to being low-income, publicly insured women of color disproportionately affected the women who developed health complications during the postpartum period. Many of the concerns including inaccessibility of care, low levels of health literacy, and insufficient support reflect the historical, systemic, structural, and political forces that have limited the ability of communities of color, and in this context specifically women of color, to lead healthy lives . As pointed out by Dr. Whitney Laster Pirtle in her paper on racial capitalism, the systems of racism and capitalism have disadvantaged racially minoritized and economically deprived groups for centuries and will continue to harm their lives even in the presence of interventions that focus on alleviating inequality at the individual and community level . While our study highlighted possible opportunities for proximal intervention that will be discussed, we recognize that true eradication of the stark disparities in maternal health outcomes will require multiple additional interventions aimed at addressing the policies and practices that uphold structural racism and drive poorer outcomes for women of color. Based on our findings, we provide the following recommendations for intervention to reduce morbidity and mortality for publicly-insured women of color during the postpartum period: 1. Improving continuity of care Extensive studies indicate that a regular and usual source of care can improve health outcomes, particularly for vulnerable populations . Postpartum care is both essential for obstetric outcomes and can serve as a vital link between pregnancy and well-woman care. Policies and strategies to improve continuity of care should focus on team-based approaches that encourage coordination of services for women. 2. Incorporating early postpartum visits Prior studies have shown that women at increased risk for postpartum emergency department use may benefit from earlier scheduled postpartum visits . By providing earlier opportunities for care, our findings suggest that there may be an increased ability to catch serious complications early, before the patient needs to go to the ED. 3. Adapting teaching materials Due to high rates of ER use within 1-week of delivery, interventions focused on discharge planning and post-delivery education among obstetric patients may reduce ER visits. While recent studies have called for increasingly standardized discharge instructions to prevent unnecessary ED use , our findings suggests that educating women on warning signs/symptoms during the delivery hospital stay is not optimal. In addition to the standard provider-patient education, accessible short videos for postpartum women and their family members about what to expect in the postpartum period and the warning signs that should prod them to contact their medical team are critical to accessible patient education. 4. Using evidence-based methodology to reduce providers' implicit bias Addressing structural racism requires a multi-pronged approach from the systems-level down that is rooted in equity frameworks. However, some progress can be achieved with provider-level training on recognizing implicit biases and providing accountability structures that encourage providers to address racism. Examples of accountability structures include providing openly available standardized assessments that allow patients to document their treatment quality, which ultimately should be linked to models of care that are associated with payment. This study had two main limitations. The first limitation is a small number of participants. Due to COVID-19, we altered our recruitment strategy and focus group setting from in-person to virtual. Consequently, we anticipated that we may experience difficulties meeting our recruitment targets due to factors such as varying rates of technology access and fluctuating availability of participants. However, during thematic analysis we concluded sufficient themes were saturated and provided rich information on which we were able to draw recommendations. Additionally, our study was conducted at a single, academic institution in an urban setting and should be further investigated in other contexts such as non-teaching and community health settings. In summary, this study offers evidence for understanding postpartum ED visits as a metric that can be useful in assessing quality improvement. By identifying risk factors for postpartum ED visits and implementing early intervention, the excess burden of morbidity and mortality experienced by low-income, underserved women can be reduced. However, future interventions should focus on addressing the structural forces that shape the contextual experience and influence women of color's ability to lead healthy lives. --- Conclusion Our study provides an important perspective into the postpartum experiences of publicly-insured women of color that can be helpful in identifying areas for intervention to reduce the need for postpartum emergency department usage. Our focus groups elucidated that some risk factors are structural in nature and addressing the patient-level factors is not enough. Addressing some of these factors will require high-level policy interventions. However, some proximal interventions addressing structural elements of care delivery such as enhancing continuity of care to increase healthcare access, adapting teaching materials to improve preparedness and engage caregivers, and training providers to reduce implicit bias, may also be beneficial. --- Data Availability The data underlying this article are not publicly available to protect the privacy of study participants. --- Code Availability The data that support the findings of this study are not publicly available due to them containing information that could compromise research participant privacy/consent. --- Author Contributions EH conceptualized, designed, and supervised the study. EH, TJ, AB, KG, and TH contributed to the implementation of the study. TH, JP, and LG facilitated focus groups. TH, KG, and AB developed code and completed data analysis/interpretation. Author TH wrote the manuscript with support from KG, AB, and TJ. All authors reviewed and approved the final version of the manuscript. --- --- Ethical Approval

The purpose of this study was to explore the postpartum experiences of publicly-insured women of color, and identify how postpartum care can be improved to reduce hospital emergency department usage after delivery. Methods We conducted four focus groups with 18 publicly-insured women who primarily self-identified as Black and/or Latina and gave birth between June 1, 2019 and May 1, 2020. We used inductive qualitative analysis to identify prominent themes from focus group discussions.We identified four domains: (1) lack of access to and communication with a medical team; (2) lack of preparation; (3) value of social support; and (4) participant-identified opportunities for improvement.This study describes the postpartum experiences of publicly-insured women of color with the objective of identifying areas for intervention to reduce postpartum emergency department usage. Our findings suggest that focused efforts on enhancing continuity of care to increase healthcare access, strengthening patient-provider communication by training providers to recognize unconscious bias, increasing postpartum preparation by adapting teaching materials to an online format, and engaging women's caregivers throughout the pregnancy course to bolster social support, may be beneficial.

INTRODUCTION Depression is one of the most common mental disorders [1], adversely affecting people's well-being, physical and mental health, and life expectancy [2,3]. Estimates of its lifetime prevalence vary widely across countries [1]. The prevalence of depression is more common in women [1], those with less education [4], and a lower socioeconomic status [5]. Depression is also related to social functioning, for example, lower relationship satisfaction [6], higher divorce rates [7], and adverse parenting performance [8,9]. Importantly, the causality of these associations is unclear. For instance, is a lower relationship satisfaction the consequence or the source of depression? Further, social environments have diverse effects on depressive trajectories-they can play a protective role or serve as a mental health risk factor [10]. Positive assortative mating or homogamy can be observed for socio-demographic , psychological , and physical characteristics [11][12][13], even though the evidence is mixed [14]. Importantly, couples, where both are depressed, may be more likely to divorce [15]. While actor effects are better predictors of relationship difficulties than partner effects , the interaction of the actor and partner effects is important as well [6]. Homogamy between spouses can arise through various mechanisms [16], namely partner preferences for self-similarity, geographical and/or social proximity , convergence , or secondary assortment [17]. Testing homogamy, however, has predominantly relied on crosssectional data. This is problematic because data on couples' similarity at one-time point indicates only a little about the causes of their homogamy, and nothing about their long-term similarity. From the perspective of interdependence theory [18], interacting partners affect each other's experiences. One might expect partners' mutual influence especially in relatively plastic traits over time. In other words, the unit of analysis should be couples instead of individuals. Depressive symptoms can change over time for many reasons . Previously, four classes of aging individuals were found with different patterns of longitudinal depressive symptoms: consistently low, consistently high, increasing, and decreasing depressive symptoms [19,20]. Thus, we may also expect that some couples have dynamically changing long-term dyadic patterns of depressive symptoms. Tracking spousal trajectories in depression is important to disentangle the dynamics and etiology of depression in a holistic perspective, considering the most important social factors . Importantly, we expect that not all couples will show the same trajectories, which can explain the difficulties in confirming the underlying mechanisms of assortative mating. Differences in longitudinal depression trajectories can be explored by advanced analytical techniques. Here, in a six-wave prospective cohort study, the Survey on Health, Ageing and Retirement in Europe , we tested the trajectories of assortative mating in depressive symptoms in long-term ageing couples, and its impact on their well-being, physical health, and relationship stability. Using the probabilistic growth mixture modeling approach, couples were divided into classes based on their longitudinal trajectories of depressive symptoms instead of using the classical correlation approach on the undivided total sample. This way, we could identify distinct classes of couples having differing longitudinal dyadic trajectories of depressive symptoms [21]. --- RESULTS Altogether 11,136 couples made up the analytical sample. Homogamy was detected for age , education , and childhood socioeconomic position . After controlling for childhood socioeconomic position, we still found homogamy for education . In addition, homogamy was revealed for wellbeing at baseline , number of limitations in instrumental activities of daily living , number of chronic diseases , extraversion , agreeableness , conscientiousness , neuroticism , and openness . The dyadic growth mixture modeling suggested four classes of depressive symptoms . Class 1 was composed of consistently low depressive symptoms in both partners . Class 2 was composed of couples with decreasing depressive symptoms . Class 3 was composed of only women having consistently high depressive symptoms while men had consistently low depressive symptoms . Class 4 was composed of couples with increasing depressive symptoms in both sexes . Although in Class 2 both men and women had a decreasing level of depressive symptoms, men started at a higher level at baseline than their female partners and their rate of decline was greater than their partners' . In Class 4, men's depressive symptoms were lower than women's at baseline, but men's symptoms increased more sharply . Homogamy in the demographic variables only negligibly differed between the classes. Descriptive statistics of the classes are provided in Table 2 and further details of the analysis are in the supplementary information. When compared to couples that were consistently low on depressive symptoms , both men and women had lower baseline well-being within the couples with mutually decreasing depressive symptoms as well as within the couples, where both partners had increasing level of depressive symptoms . Only women had lower baseline well-being within the couples, in which only women had consistently high depressive symptoms, but men had consistently low level of depressive symptoms . Results were similar when tested for the total number of chronic diseases and limitations in instrumental activities of daily living. The prevalence of bereavement was about 3-4 times higher in Classes 2-4 than in the nondepressed reference Class 1 . The Class 4 = 14.37, p < 0.001) and Class 3 = 4.11, p = 0.043) couples had higher chance of break-up than Class 1 couples. See further differentiating factors between the classes in Table 2 and the detailed results of the multinomial regression in Supplementary Table S8. --- DISCUSSION Depression affects various aspects of people's lives. Here, we explored romantic couples in relation to their mutual trajectories of depressive symptoms and how those trajectories were associated with relationship stability, bereavement, and wellbeing. In a large sample of couples from Europe, we identified four Fig. 1 Dyadic latent trajectories of depressive symptoms across four classes of couples. Measurements taken every 2 years between Wave 1 and Wave 7, except in Wave 3; the scale of the depression scores ranged from 0 to 12; men's and women's mean depressive symptoms were presented in solid and dashed lines, respectively; 95% confidence intervals were shadowed in grey. classes of couples based on their dyadic longitudinal depression trajectories. Most couples were homogamous in their low levels of depressive symptoms; followed by a class, where only women had high depressive symptoms ; 7.83% of couples with increasing symptoms; and 7.18% of couples with decreasing symptoms. Further, we observed that couples were strongly homogamous in age, level of education, well-being, and childhood socioeconomic position, and were moderately homogamous in physical health and personality. Notably, the couples with increasing depressive symptoms had the same direction of change but different rates of change and different endpoints of mean depressive symptoms. The couples with decreasing depressive symptoms also had the same direction of change , but they had different starting points converging to similar ending points. Using the traditional nomenclature [11], we may call the second group "convergent", but the first group could be both labeled "divergent" and "convergent". Even though the direction of change was the same in both partners in both classes, their levels of depression at the end of data collection were different. The traditional nomenclature of couples' divergence only considered couples having similar starting levels and different ending levels in a characteristic over time. However, importantly, we added another dimension to categorizing couples' joint longitudinal changes, the direction of change. We argue that true divergence would mean having different directions of change over time . According to this rationale, we did not identify any class of couples truly divergent in depressive symptoms. Why is couples' assortment in depression important? Evidence suggests at least partial genetic background together with familial effects, and especially with unique environmental effects in depressive disorders [22,23]. Further, the offspring of parents who were concordant in affective disorders were at higher risk for depression than children with only one affected parent [24]. Thus, spousal similarity can rise the prevalence of depression in the offspring because of both genetic and environmental factors and their interactions [25]. Knowing the different patterns couples show in their depression trajectories, it would be possible to better explore how different relationships affect their offspring's mental health. When both partners had increasing depressive symptoms , they had the highest prevalence of bereavement and relationship dissolution during subsequent assessment in comparison to the other classes, perhaps unsurprisingly. Importantly, although men's rate of increase in their depressive symptoms was higher than women's, both had lower well-being, and worse physical health than the nondepressed Class 1. A reasonable explanation could be that their diminishing health puts too much weight on these couples jeopardizing their relationship stability . However, one would need to have a more detailed picture of their day-to-day lives including their psychological coping and living circumstances to be able to better plan possible intervention programs for these couples. Couples in the increasing and decreasing depression classes have a certain synchronization in the direction of their depressive symptoms. However, in both classes, men had higher depressive symptoms in almost all waves than women. Interestingly, in the class where women's depressive symptoms were consistently high, men were not affected by their partner's symptoms. This is in accordance with recent findings that women were more susceptible to the emotional contagion of sadness than men [26]. Apart from the lack of emotional contagion in male partners, other underlying factors may have been responsible, for example different manifestations or types of depressive symptoms in each class. The most common class consisted of both nondepressed, homogamous partners. Remarkably, on the other hand, there was no class consisting of both partners with consistently high depressive symptoms. We speculate that people with mutually high depression probably do not initiate a relationship with each other, or these couples are not stable, and their relationship quickly dissolute. However, we cannot rule out the latter possibility from these data, as couples should be followed-up from the initiation of their relationships. --- Future directions Future research should investigate whether initial assortment and preference for low level of depressive symptoms is responsible for homogamy in having no depressive symptoms. Alternatively, part of these couples started heterogeneously, but after an adjustment period they both converged to have low depressive symptoms. Presently, these possibilities cannot be tested because our sample only consisted of established couples, and we have no data available about their depressive symptoms before their relationship started or from the beginning of their relationship. In addition, our observations were limited to the study being collected in waves at 2-year intervals. Hence, day-to-day mood changes could not be detected, but instead, our study extracted trait-like, long-term patterns. To better understand the mechanisms of these trajectories, one should investigate in more detail the day-to-day dyadic dynamics within the couples including a more detailed analysis concerning the specific time of bereavement and relationship dissolution. This would allow researchers to explore who affects depressive symptoms, which is especially interesting in couples with increasing and decreasing depression. Targeting couples with similar directions of change would unfold the underlying shared environmental factors affecting both partners-even if the effect is differing in degrees between the sexes. --- LIMITATIONS AND CONCLUSIONS Evidence for homogamy in affective disorders is mixed [24]. There were methodological differences across previous studies, which may be responsible for that . This study is based on a cross-national, longitudinal, and representative sample. Such data are unparalleled in relationship research and couples' data. Although we did find some differences between the classes apart from their depressive symptoms and well-being , we did not predict them, and thus, they deserve more attention in the future. As already mentioned above, the two-year time lag may be considered too large to observe moodrelated fluctuations, thus our findings apply only to trait-like longitudinal patterns of change. Further, we know little about important aspects of their relationships and mental health, which imposes constraints on the conclusiveness of our results. The couples were already established when the project started, and their overall low rates of relationship dissolution limited our ability to observe the effects of depression on relationship stability. Although these couples came from 16 European countries with varying cultural and economic backgrounds, they still share significant similarities in being Western, Educated, Industrialized, Rich, and Democratic [27] societies. As they were on average about 61 years of age, we cannot generalize these findings to younger couples either. In sum, we showed different patterns in couples' longitudinal trajectories of depressive symptoms. By defining four different classes, we suggest a new approach when considering assortative mating longitudinally. We detected factors, which were specific for each class, such as well-being, physical health, relationship dissolution, and bereavement rates. Consequently, the different classes of couples might need different intervention approaches. Lastly, we demonstrated that the classification approach and longitudinal data are useful both from the perspective of clinical practice and social sciences as they can differentiate unique longitudinal patterns of couples' plastic characteristics. The study has an interesting setup for the future. It suggests that we tend to treat all couples as the same, which obscures important differences. Identifying between-couple differences is a new frontier of research. Not all couples are the same and not all factors will differ from couple to couple and not all factors which differ will matter. --- MATERIALS AND METHODS Participants Data for the analysis came from a prospective cohort study SHARE, which was previously described in detail [28]. Briefly, SHARE is a multi-centric, multidisciplinary longitudinal study that was initiated with the aim to assess health, social network, and economic conditions of communitydwelling individuals in Europe and Israel. The first wave of SHARE was conducted in 2004, followed by five subsequent waves in ~2-year intervals: Participants were sampled based on probability selection methods. To be eligible to take part, individuals must be at least 50 years old, speak the official language of the country, and not have lived abroad or in an institution at baseline. Data are collected using computer-assisted personal interviewing in the participants' homes. In case the participants have a spouse, the spouses were also invited to take part, irrespective of their age. Therefore, the SHARE study is an optimal data source to study trajectories of depressive symptoms in couples. This study was carried out in accordance with the Declaration of Helsinki. SHARE has been repeatedly reviewed and approved by the Ethics Committee of the University of Mannheim. All participants provided written informed consent. Data were pseudo-anonymized and participants were informed about the storage and use of the data and their right to withdraw consent. --- Analytical sample We restricted the analysis to individuals who participated in SHARE with their spouses, and each had at least three measurements of depressive symptoms, irrespective of from which wave the data came. Inclusion of at least three measurements enabled us to study trajectories of depressive symptoms. From 206,723 individuals included in the SHARE database, 139,556 people completed at least one interview. From them, we identified 125,532 individuals who took part in SHARE with a spouse, therefore, there were 62,766 couples. We excluded 546 same-sex couples and 36,746 couples that did not have at least three measures of depressive symptoms, leaving the final analytical sample of 11,136 couples . This sample size is sufficient to perform growth mixture modeling on data of six waves [29]. Mean age of men was 62.31 and of women was 59.21 . Flowchart is presented in Supplementary Fig. S1. --- Trajectories of depressive symptoms To assess depressive symptoms, SHARE uses the EURO-D scale, a tool that was originally developed to compare symptoms of depression in older adults across Europe [30] and has been used in multiple studies [31][32][33]. The measurement with EURO-D was conducted in all waves except for wave 3. Participants were asked whether they have experienced 12 symptoms during the last month . For each symptom, they received one point, creating a scale ranging from 0 to 12 points, with higher values suggesting more depressive symptomatology. Couples' dyadic trajectories of depressive symptoms were extracted using dyadic growth mixture modeling in Mplus version 8.7. This approach combines the latent classification methods and latent growth modeling, which can thus differentiate latent trajectories of depressive symptoms. The couples that followed similar dyadic longitudinal trajectories at a high probability were grouped into the same latent classes. The final classification model selected for publication was performed according to the recent guidelines [21]. See the Supplementary information for all details of the performed analyses. Following interpretation and supporting criteria, the 4-class model was selected for publication . The 4-class model showed the following patterns of dyads: Class 1. Consistently low depressive symptoms in both members of the dyad , Class 2. Decreasing level of depressive symptoms in both men and women , Class 3. Only the women have consistently high depressive symptoms, but the men have consistently low depressive symptoms and Class 4. Increasing depressive symptoms in both members of the dyad , see Fig. 1. The Figure was created in R Studio version 1.4.1717. In a subsequent analysis, the latent classes were regressed in a multinomial regression on the covariates using the 3-step method. Break-up during the follow-up period as categorical distal outcome was predicted by the latent class variable using the DCAT method [34]. Mplus syntaxes can be obtained from the first author upon request. --- Covariates Five sets of covariates were used in five respective models to study, which factors were associated with the trajectories of depressive symptoms. The first set included the following demographic variables: age , sex , region , education [35], childhood socioeconomic position [32], the number of children and grandchildren, and the type of participant's residence area . The second set comprised the following health variables: well-being [36], number of limitations in instrumental activities of daily living [37], chronic diseases , and drugs for depression . A subsample of individuals who participated in the 7th wave had available information on the third set of covariates, the Big Five personality traits , measured by the Ten-Item Personality Inventory [38]. The fourth model included the death of either men or women during the study [20]. The fifth model included a variable break-up, which was defined as a transition from being in a couple with the studied spouse to being single or being with another partner, irrespective of whether the original partners got back together. Information about covariates was acquired at baseline, which is the wave data on EURO-D was first available, except for the variables for drugs for depression , death , and break-up . Covariates of both sexes were used, except for the number of children and grandchildren. As these were too strongly correlated between men and women, we included this only for women to avoid collinearity. Means and proportions were compared between classes in SPSS version 26. --- --- COMPETING INTERESTS The authors declare no competing interests. --- ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41398-022-01950-w. Correspondence and requests for materials should be addressed to Zsófia Csajbók. Reprints and permission information is available at http://www.nature.com/ reprints Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The occurrence of depression is influenced by social relationships, however, most studies focus on individuals, not couples. We aimed to study how depressive symptoms of couples evolve over time and determine, which characteristics are associated with their distinct trajectories. A multi-centric cohort sample of 11,136 heterosexual couples (mean age = 60.76) from 16 European countries was followed for up to 12 years (SHARE study). Information on depressive symptoms measured by EURO-D scale was collected every 2 years. Dyadic growth mixture modeling extracted four distinct classes of couples: both non-depressed (76.91%); only women having consistently high depressive symptoms while men having consistently low depressive symptoms (8.08%); both having increasing depressive symptoms (7.83%); and both having decreasing depressive symptoms (7.18%). Couples with increasing depressive symptoms had the highest prevalence of relationship dissolution and bereavement. In comparison to the nondepressed class, individuals with any depressive symptoms were less psychologically and physically well. Our results suggest that distinct mechanisms are responsible for couples' various longitudinal trajectories of depressive symptoms.

Introduction Health and risk communication are critical to prevent the spread of the severe, acute respiratory syndrome coronavirus 2 , Health and risk communication are critical to prevent the spread of the severe, acute respiratory syndrome coronavirus 2 , also when it comes to the roll out and effectiveness of vaccines, yet it brings with it challenges and complexity [1]. Only when individuals have an accurate understanding of their risk of getting infected by or dying from the coronavirus disease 2019 will they be able to take responsibility and engage with the range of prevention practices and technologies available [1]. Furthermore, studies have also shown that some migrants have an increased risk of contracting infection with COVID-19 as well as suffering severe consequences from it [2][3][4], including psychosocially [5]. It is therefore essential that information about COVID-19 is made accessible to all population groups, including migrants, giving them an equal chance to find effective ways to protect themselves from infection and engage with guidelines concerning preventive measures in their country of residence [6][7][8][9]. However, migrants are often "left behind" in health risk communication efforts [6,10]. A review of public health communication during the first months of the COVID-19 pandemic revealed that fewer than half of European countries had online COVID-19 advice in at least one major migrant language [10]. In Denmark, official information about COVID-19 in English was not published until the end of March 2020, approximately two weeks after the first national lockdown [11]. Information about COVID-19 has since been made available in other languages [12,13]. Other sources of news also failed to consider other languages. For example, live press conferences frequently held by the prime minister and health authorities were broadcasted without subtitles or translation [13]. As a consequence, and as noted by the World Health Organization 'ApartTogether' survey, many refugees and migrants report relying on social media or news about COVID-19 from their country of origin to keep themselves informed [14]. Failure to consider migrants' linguistic diversity in health risk communication responses leaves them underserved and vulnerable [15] and ultimately put migrants at higher risk of SARS-CoV-2 transmission [3]. For example, a report from the Danish Institute for Human Rights interviewing ethnic minorities with poor Danish language skills during the COVID-19 period showed that some were uncertain about the risk of transmission of the disease as well as the authorities' guidelines, which sometimes led to extreme self-isolation and concern [16]. A small but expanding body of literature is beginning to highlight the consequences of the linguistic challenges faced by migrants in this time of crisis. For instance, research from Nepal [17] and the Netherlands [18] has found language barriers to be associated with numerous consequences such as stress, anxiety, difficulties in accessing relevant services, and the spread of misinformation among different migrant groups [17][18][19]. Furthermore, not addressing migrants' linguistic diversity does not just leave migrants underserved and vulnerable to infectious diseases such as COVID-19, it undermines the broader pandemic response. Much current knowledge, however, is based on quantitative data, commentaries, or editorials. We have little empirical understanding of how migrant groups qualitatively experience their access to health and risk information concerning COVID-19. Furthermore, much of what we do know relates to the information barriers migrant groups face in accessing health and risk communication, whilst little has been done to explore the tactics they adopt to overcome these barriers. We therefore take inspiration from de Certeau's [20] notion of 'tactics' to explore the everyday acts and occurrences that migrant groups adapt to attain some control, or make do, in an information and communication setting deemed constraining. By drawing on the concept of 'tactics', we will be able to not only highlight the challenges experienced by our participants but also disentangle their actions and the supportive structures they form part of. Against this background, we ask: what are the tactics adopted by migrants in Denmark when facing the challenges of accessing information on COVID-19? --- Materials and Methods We drew on data from a qualitative interview study that aimed to investigate the various impacts of the COVID-19 pandemic for migrants living in Denmark. The study forms part of the international 'ApartTogether Study', which is a collaboration between WHO, the UN System, and a consortium of universities [13]. Informed oral consent was obtained from all participants, guaranteeing that they would be anonymized. We used pseudonyms when referring to the participants throughout this paper. --- Study Location and Participants The study took place in Denmark during the spring of 2020, following the first wave of the COVID-19 pandemic. Like many other countries, the Danish government implemented a nationwide lockdown of all non-essential activities. However, at the time of the interviews, most activities had temporarily reopened [20]. Recruitment of the migrants used in this study took place in May and June of 2020. Migrants who had filled out the ApartTogether Study survey and who had indicated that they would be willing to participate in an interview were contacted. Only respondents who had answered the survey in English or Danish were contacted. Interviewees were contacted through either e-mail or a text message depending on the contact information provided. After a few days, a reminder was sent to interviewees who had not yet answered our first request. We recruited migrants of any age, interviewing a total of 18 migrants living in Denmark. The migrants included four men and 14 women. They came from various countries within and outside of Europe. Many of the migrants had lived in Denmark for a year or less, and there was a range between four months and seven years for all migrants . --- Data Collection and Analysis In May and June 2020, the first author conducted 18 semi-structured interviews with the migrants. Due to the COVID-19 restrictions imposed by the Danish government at the time of the interviews as well as the general risk of gathering, all interviews were conducted virtually . A description of the study was provided to the participants followed by obtaining an oral consent. All interviews were conducted in English and lasted between 31 and 76 min. The interviews were steered by a topic guide structured around five different themes: corona and social distancing in your everyday life, experiences and sense-making, worries and vulnerabilities, social support and coping strategies, and the response of the authorities and civil society. The conversations surrounding health risk communication are the center of this article. All 18 interviews were audio-recorded and subsequently transcribed in their full length. In our efforts to disentangle emerging themes, we coded and thematically organized the interview transcripts in accordance with the steps outlined by Attride-Stirling [21]. This was done with the support of NVivo 12, a qualitative data analysis software. A total of 112 codes emerged from this process. However, we do not intend to present all the emerging themes in this paper. Instead, we focus on the 28 codes that relate to health risk communication challenges and tactics . The remaining codes covered themes such as worries about their financial situation, well-being of their relatives, and loneliness. In preparation for this paper, the 28 selected codes underwent a second round of organizing. Codes that related to each other were grouped into basic themes. If the basic themes shared similar issues, they were then rearranged into what Attride-Stirling refers to as more interpretive organizing themes . In the following analysis, we will present and unpack the different themes derived from our thematic network analysis. --- Results Throughout the interview narratives, it was clear that many of the migrants faced several challenges when it came to accessing the health risk communication coming from the Danish authorities. In response to this, however, many of them found other channels of communication. --- Challenges: "I Just Want Some Clear Answers" Throughout the interviews, many of the migrants expressed that it was difficult to navigate the authority's response to COVID-19. This included how and where to find health and risk information, or making sense of the information they found, either because it was in Danish or imbued with cultural reference points. For instance, one common issue among the migrants was not knowing which and when restrictions were put in place. This is illustrated by Leah, who had been living in Denmark for only nine months. In the following quote, she describes her inability to follow the regularly press briefings held by the Danish authorities in the spring of 2020 at the height of the first wave of the pandemic: "It has been really hard to figure out like when Danes know [e.g., timing of press briefings]. Like that the prime minister is going to make a speech. Because it will happen at seemingly completely random times. And then like all the sudden we find out "oh yeah, she has been speaking for half an hour" and we turn it on. But like there was no way to figure out. Yeah, I never, I still don't know how people are finding out when she was going to say important things. I was always like a bit behind". It's probably because we are not counted as well, but there is a lot of things that is about Danish citizens and not about foreigners". These accounts can be read as examples of recently arrived migrants feeling left out of the response due to their migrant status and lack of proficiency in the Danish language. Besides being a great source of frustration among several of the migrants, this sense of not being "counted", as Heather puts it, resulted in confusion and uncertainty about how best to navigate the pandemic response . Another common barrier to the migrants' ability to navigate the pandemic response was linguistic difficulties. Many of the migrants did not speak Danish, making it difficult for them to understand much of the COVID-19-related information coming from the Danish authorities. Matej, who had just started learning Danish, provided an example of his inability to follow the Danish news: "So most of the time I feel like that most of the information is in Danish, and I started to learn Danish recently, but I am not fluent yet". Faiza, who had been in Denmark for less than a year and did not speak any Danish yet, also described how information from the Danish authorities are not accessible for migrants: "Information is not that clear. I mean, it is not eligible for people who are like us, immigrants". Despite Faiza knowing how to navigate the news flows, she was still feeling left behind due to linguistic barriers. Many of the migrants interviewed did not feel that they were addressed in a way that was understandable to them, and this was especially true for those migrants who had not been in Denmark for very long or had moved to Denmark alone. Because of their unfamiliarity with the Danish system, coupled with their inability to understand the Danish language, many of our participants were unable to understand the messages and health advice coming from official sources, e.g., the regular press briefings. Like Faiza, they were lost in translation. Shortcomings by the Danish state to make health risk information available to migrants in a timely manner were, however, partly addressed by other structures, such as workplaces or language schools. This was experienced by a small handful of our participants. For instance, Irene, who had come to Denmark only a year ago, explained that the language school that she was enrolled in not only provided Danish lessons but also valuable COVID-19-related information in the beginning of the pandemic: "When it started to sort of get momentum, the whole pandemic. In school we had a class that was completely dedicated to, you know, "this corona is looming, wash your hands", you know, "sneeze in your elbows". So like this sort of, they basically also explained the basic things about, you know, personal hygiene and safety." For others, the workplace was a way to stay updated on what was going on with regards to changing COVID-19 rules and regulations. This was exemplified by Bogdan, who regularly received essential information about COVID-19 from his manager: "Our general manager would send us e-mails. Keeping us posted with what's going on. . Most of the time actually all the time, we knew what's going on. And they would tell us, you know: "when we gonna know more from the authorities, we're gonna let you know"". However, for most of our participants, such 'back-up' support was unavailable, requiring them to actively seek out information. --- Tactics: Seeking to 'Be in the Know' To get information about COVID-19 in Denmark, the migrants employed various tactics. Tactics can be understood as the formation of new daily practices amongst people in marginalized positions as a result of their absence in more formal, dominant and strategic responses, either to 'get by' or to display resistance [19]. In this case, the marginalized positions are those of migrants in Denmark, who experience exclusion from the authorities' channels of communication. They therefore turn towards opportunities to receive information about COVID-19 provided outside the established frameworks of communication. One tactic was to use a tool to translate, for example, press briefings held by the authorities, here exemplified by Linh: "And everyone was just watching the news, and again I don't speak any Danish, but I just used the translator on my web browser to try to see if, what the authorities say, at least when they are on TV or the newspaper". Other tactics that the migrants employed on their own included reading online news in English or their native language or accessing news from their country of origin. Beyond searching for information on their own, some migrants also relied on their close social networks such as family members or friends who spoke Danish. Faiza, who had a Danish boyfriend, described that he translated the most important news about COVID-19: "So, it's difficult and my boyfriend is happy to translate it. If relevant he translates it to me, but then I would like to process it on my own, you know". Some migrants also used their broader social network to stay updated with the aid of social media. Here exemplified by Blanca, who was a part of a WhatsApp group with her friends also living in Denmark at the time. For example, they regularly shared COVID-19 related information with each other: "When it comes to information from Denmark, we, what we had to do, or didn't do, we, we had a WhatsApp-group with our friends here in Copenhagen, and some of them have been here for many years, so they would send links with important information, and every time that Mette [the Danish Prime Minister] talked, they would just send a WhatsApp "so hey, that is kind of like the summary", so they kind of like made it work for us". As the accounts show, Faiza and Blanca made active use of their social networks to access and share information. However, even though both were able to mobilize their social network to get relevant information regarding COVID-19, they also expressed a desire to assess the information on their own without relying on the help of others or digital tools. In our interviews, information sharing about COVID-19 also happened in much more structured transnational networks. Several of the interviewed migrants explained how they actively used larger groups or communities on social media platforms that targeted migrants living in Denmark. This was well articulated by Ana, who did not speak any Danish and did not have any Danish-speaking relatives and friends to help her with translation of COVID-19-related information. In the following, she reflects on how she stayed updated using Facebook: " So it is a Facebook group. And there are some people there that right after the government announcements, they translate everything into bullet points what is most important, so mainly I would follow all the guidelines from there". Ana continues, elaborating on how the group supported her in accessing COVID-19related news: "The Facebook group actually helps quite a lot, because there are some people that they have been through the struggle of the language, and they know how difficult it is to actually have access to news etc. So you have a lot of support of people that have been living on the country, ehh, for many, many years, and they are actually there to support you and give you information". Having just arrived in Denmark, Ana has not had time to adapt to her new surroundings. Therefore, the Facebook group was particularly important for her to stay updated on the COVID-19 response in Denmark. For newly arrived migrants who still needed to establish strong social networks, the group constituted an important access point for COVID-19-related information, including knowledge about measures introduced by the Danish authorities, and ultimately being able to follow guidelines as they relate to the Danish context. The use of such social media networks can be seen as an example of how the sudden reorganization of established spaces, because of COVID-19, was tactically utilized by migrants-providing them with important information about COVID-19. --- Discussion We set out to explore the challenges and tactics adopted by migrants when accessing COVID-19 information. Our findings show that being away from their country of origin in a time of crisis, and having no or only limited knowledge of the Danish system and language, some migrants find themselves excluded from the COVID-19 response. Most of the migrants participating in our study noted that much of the information available at the start of the pandemic was in Danish and that it was difficult to navigate the news flows. Having little knowledge about the Danish language and customs, many of the migrants were unable to keep up with the changing restrictions and guidelines and the rationale behind these. Our data suggest that this was especially the case for migrants who had only been in Denmark for a short period of time and who had come here alone. Those migrants who had stayed in Denmark for several years appeared to face fewer challenges in navigating health risk information about COVID-19. The feeling of not being included in the response-or repeatedly accessing information with much delayled to uncertainty and frustration among many of our participants. Whilst the Danish authorities and non-governmental organizations eventually made essential health and risk information available in multiple languages, it was difficult for the participants to keep up with changes, often announced in press conferences held in Danish. Our participants therefore relied on adopting various tactics to access and share information. These included the use of translation tools, searching for news in their country of origin and on social media platforms, as well as engaging with different digital migrant-networks, helping each other navigate the pandemic in Denmark. In different ways, digital tools emerged as critical enablers of tactics in accessing and sharing health and risk information outside of established structures and spaces. This finding resonates with previous observations. A study by Park et al. [22], exploring the experiences of migrants living in the Republic of Korea at the time of the MERS outbreak, describe digital tools as "weapons to survive" an emergency when living in a foreign country. They found that digital tools were used to receive and share information among migrants in their closer social networks [22]. In our study, however, information sharing also happened in much larger online social media groups and communities, illustrating new tactical practices. Social media thus acted as a key gateway to receive health and risk information about COVID-19 when official information channels were inaccessible. Despite social media platforms being a much-used alternative source of information on COVID-19 among the migrants in this paper, studies suggest that social media platforms also have the potential to disseminate inaccurate information [22,23]. Furthermore, it has been suggested that increased use of digital tools to get COVID-19-related information, and the use of many diverse platforms, can lead to information overload, increasing the risk of important information to be ignored or forgotten [24]. While digital tools were important for the migrants in accessing health and risk information about COVID-19, the tactical use of such tools often also involved fellow migrants or other people in general. Even though digital tools enable access to information, these digital tools are only relevant because the migrants we spoke to were part of compassionate social networks characterized by information sharing and support. Relying solely on the help of others , however, to access vital health information might leave them vulnerable to disruptions in information flows or misinformation. Nonetheless, given the absence of the Danish welfare state to cater for the health and risk information needs of migrant populations in the early and critical weeks of the pandemic, digital tools proved critical in helping our participants to access essential health and risk information when there was no alternative. Our findings are constrained by two methodological limitations. One, we did not have access to language interpretation. This meant that we could only recruit participants who either spoke English or Danish. As many of the participants did not have English or Danish as their first language, important nuances of their experiences may have been lost. Two, we recruited participants and conducted the interviews through online and digital means, requiring a certain level of digital literacy and access to digital tools. It is therefore likely that we have recruited more resourceful migrants, missing out on the experiences of those who are most vulnerable. Future research ought to include more disenfranchised migrant populations. Whilst our findings highlight the practical implications of social media platforms in helping our participants to access information about COVID-19, they also point to the need for offline alternatives to reach more vulnerable migrant groups. Council estates with significant migrant populations could recruit trusted COVID-19 ambassadors for different cultural language groups and support them go door-to-door making COVID-19 health risk, testing, and vaccine information available. --- Conclusions This paper demonstrates the agency and resourcefulness of migrants in overcoming the specific challenges some migrants face when it comes to their ability to access and understand health risk communication in the country they reside in. Understanding and recognizing the role migrants can play in health risk communication responses is key to support them during this pandemic and in the future. There is a need to consider migrants' resources when developing future information strategies, including involving and engaging migrants in the process. --- Data Availability Statement: The data are not publicly available. Requests to access the data used in this study can be sent to the corresponding author. --- Informed Consent Statement: Informed consent was obtained from all the migrants involved in the study. --- --- Appendix A --- Codes Basic Themes Google Translate Information and news are in Danish Danish media are trustworthy Webpages in English

Health risk communication plays a crucial role in preventing the spread of infectious disease outbreaks such as the current coronavirus (SARS-CoV-2). Yet, migrants are far too often forgotten in health risk communication responses. We investigate the challenges and efforts made by migrants in Denmark-in the initial months of the pandemic-to access information about COVID-19. We draw on 18 semi-structured interviews conducted in May and June 2020. All interviews are thematically coded and analyzed. Our analysis reveals that many of the migrants faced several challenges, including accessing information in a language understandable to them and navigating constant streams of official news flows issuing instructions about which actions to take. However, we also note that the participating migrants found numerous creative ways to address some of these challenges, often aided by digital tools, helping them access crucial health and risk information. This paper highlights that migrants constitute an underserved group in times of crises. They are vulnerable to getting left behind in pandemic communication responses. However, we also identify key protective factors, social resources, and agentic capabilities, which help them cope with health and risk information deficits. National governments need to take heed of these findings to inform future pandemic responses.

When the Government of India's Expert Committee on Tribal Health requisitioned for an estimate of the Infant Mortality Rate for the tribal population in India, the office of the Census and Sample Registration System was unable to provide the same. The nation had not cared to measure death rates in tribal children. The committee, for its Report [1], had to depend on indirectly estimated rates based on the Census 2011 and the National Family Health Surveys 1 to 4. But the generic term 'tribal people' incorporates within it 705 different tribes, each having unique culture, lifestyle, and hence, with a different child mortality rate. The great anthropologist Verrier Elwin had long ago documented the life and culture of various major tribes in India [1]. Who would provide us their IMRs? In the present issue of Indian Pediatrics, Verma, Sharma and Saha, three researchers from the National Institute of Research in Tribal Health, provide us exactly that [2]. From the Census of India data they have estimated the IMR, Under-5 Mortality Rate and Life Expectancy at Birth for scheduled tribes from seven states -Rajasthan, Gujarat, Maharashtra, Madhya Pradesh, Chhattisgarh, Jharkhand and Odisha [2]. They selected the tribes with population size large enough, indentifying 123 tribes which account for 94-97% of the total ST population in the respective states. Using the demographic methods of indirect estimation pioneered by Prof. Brass of the London School of Tropical Medicine and Hygiene, they estimated, for the first time, the IMR, U5MR and LEB for these 123 tribes [2]. They deserve kudos for this contribution. These estimates reveal three diversities -between the total population and the ST population of India, between the seven states, and within each state, between the tribes. Since what the authors have estimated, the IMR, U5MR and LEB, are literally the estimates of life and death, they matter! What these different statistical numbers reveal are huge inequalities for the opportunity to survive -the most fundamental human right. These researchers estimate that whereas the IMR for the total population of India, in 2006-07, was 65 per 1000 live births, it was 76 for the total ST population in the country. But hidden within this number, 76,were huge diversities. The IMR for these 123 tribes varied from the lowest 48 for the Gamit tribe in Maharashtra to the highest 124 for the Birhore and Bharia tribes residing in Chhattisgarh, Jharkhand and Madhya Pradesh. Seventy six additional infants died in these later two tribes per 1000 live births. Similarly, the U5MR among these 123 tribes ranged from the lowest of 57 in one tribe to the highest of 203 in another. It was lowest for the total ST population in Maharashtra and highest in Madhya Pradesh. The estimated LEB for these 123 tribes ranged from the lowest 51 years to the highest 72 years . Two major limitations of these first-ever estimates are, one, they are not actual measurements over a period of time but have been indirectly estimated from a crosssectional data, the Census of India. Second, they pertain to the year 2006-07, in a way, already outdated. But they make a beginning of making estimates for the individual tribes. Hopefully, the estimates based on the next national Census will arrive sooner. The two landmark reports, of the Lancet-Lowitja Global Collaboration on the Health of the Indigenous and Tribal Populations [3] and 'Tribal Health in India -the report of the Expert Committee on Tribal Health' of the Government of India [4], have --- E D I T O R I A L C O M M E N T A R Y E D I T O R I A L C O M M E N T A R Y E D I T O R I A L C O M M E N T A R Y E D I T O R I A L C O M M E N T A R Y E D I T O R I A L C O M M E N T A R Y VOLUME 58 __ JANUARY 15, 2021 EDITORIAL pointed out that globally as well as nationally, the indigenous and tribal people suffer worse health status and chances of survival compared to the general population in the countries. Regrettably, India had the second highest IMR for the tribal people in the world, next only to Pakistan. Now Verma, Sharma and Saha show that even within the tribes, there are large disparities between the states and within the states. So, what do we make of this? One, the policymakers need to appreciate the importance of segregated measurement for the tribal people as a whole and for each individual tribe. The expert committee on tribal health has underscored this need; and some movement in the academia can be seen after that. Will the Ministry of Health and Family Welfare, and the Ministry of Tribal Affairs show more action? Second, the tribal development plans -the tribal subplans -and the health plans of the states should now move further and develop the tribe specific plans. Birhore tribe, whether in Chhattisgarh, Jharkhand or Madhya Pradesh, has the highest child mortality. Each tribe has different challenges, hence needs separate attention and solutions. Third, the pediatricians and policymakers need to assert that the tribal mothers and children receive near complete coverage with the proven health care inter-ventions such as the ANC, institutional delivery, home-based neonatal care, immunization, breastfeeding and nutrition, and finally, treatment for pneumonia, diarrhea and malaria. But the coverage will improve only if measured and monitored separately for tribal children. Niels Bohr was absolutely right - tribal children lives matter, measure them!

any of us will be surprised to know that as many as 104 million tribal people live in India, more than the national populations of 90 percent of the countries. How do tribal children in India live, and more importantly die? How many? Why? What can be done? These questions bother every pediatrician's conscience -sometime or other.

Introduction Numerous studies have been published to date about emergency remote in higher education during the COVID-19 pandemic [1][2][3][4]. Consequences for Higher Education institutions were devastating at a broad range of levels, from a human and also an economic perspective [5]. Up to this date, we find reports from all continents and a huge diversity of countries from East to West and North to South (e.g., Pakistan [6]; Turkey [7]; United States of America [8], Russia [9], South Africa [10], New Zealand [11], India [12], and United Arab Emirates [13] among many others, even multinational comparisons [14]. Both large-scale, even global [1,[15][16][17] and small-scale, qualitative studies [18] have been undertaken. Certain studies have focused on partial, subjective perspectives of the educational participants, either from the instructors' side [9] or from the students' [10]. Some studies have contributed to a better understanding of particular pedagogical techniques and tools in the e-learning context [19,20], while other researchers focused on mental health risks [8] and emotional coping strategies [21]. With our study, we want to contribute to this massive body of knowledge that should help us all together to be better prepared to overcome likely new ERHE situations in the future. In Spanish-speaking countries, there are also a great deal of studies [22][23][24][25][26]. We find, for example, Navarro and colleagues' study [26] addressing instructors' related variables in the successful perception of ERHE, while others [21] focus attention on the student's perspective. Each of these studies, however [21,26], addresses different issues, such as digital teaching competence on the instructor's side and emotional coping strategies on the student's side. They do not contemplate the very same issues for both instructors and students so that their results can be considered cumulative in a certain way but not contrastive to each other. To our knowledge, studies comparing students' and instructors' perspectives are still very scarce and of a limited scope regarding participating samples [27,28]. In that sense, our study responds to a current need, providing a comparable perspective of both participants' sides. --- Conceptual Considerations Regarding HE and ERHE Globally considered, HE institutions already had adopted and mainly integrated information and communication technologies for teaching and learning by 2020. However, this was not necessarily the case worldwide. Up to the pandemic, HE institutions in Latin American countries functioned predominantly traditionally, face-to-face [29]. Hence, the emergency forced the HE institutions in these countries to cope with unexpected challenges at the highest velocity, and all instances of the educational community had to adapt themselves to this new situation. Debattista [30] offers good practice principles for instructional design, such as encouraging student-faculty interaction, promoting inter-peer cooperation, facilitating active learning, offering prompt feedback, emphasizing time on task, communicating high expectations to the learners, and respecting and catering to diverse talents and ways of learning. Certainly, these principles do not differ much from face-to-face quality teaching and learning practices. Previous literature on e-learning and blended learning points to the need for specific so-called techno-pedagogical design [31]. In other words, teaching and learning processes in the virtual context require specific instructional decisions to promote and enhance learning chances, as ICT by itself does not warrant positive learning processes [32][33][34]. Even the assessment processes must be subject to particular design in the e-learning and blended learning contexts [35]. As some underline, managing time is probably the biggest challenge in the online setting [36]. Last but not least, human interaction must be taken care of in the virtual learning context [37] in order to promote quality learning. All of these instructional decisions were compromised in the emergency context. As the global study of Aristovnik and colleagues reported [17], the student population from Latin American countries had less favorable studying conditions and more basic infrastructure, such as fragile and unstable access to the internet, compared to richer countries in the Northern hemisphere. In the case of instructors, the lack of the so-called digital competence, that is, the ability to plan and manage teaching and learning situations mediated by e-learning components, was probably the biggest problem to tackle globally, but especially in those places where ICT e-learning components were not yet common ground [38]. Studies prior to the pandemic situation already reported students' difficulty in communicating with peers in the e-learning context [39,40]. As the literature claims, the online interaction between instructors and students feels easier for the students themselves, compared to the need to interact with peers in a constructive manner. In other words, peer interaction for learning purposes in the synchronous and asynchronous space does not happen spontaneously with ease and benefits from pedagogical guidance [41,42]. However, as some studies report [43], there might be some connection between students' interaction quality and achievement during ERHE, which underlines the importance of this pedagogical guidance. Regarding teaching, previous results report the extraordinary demands that the ERHE put on the instructor's need of re-designing teaching and assessment strategies and coping with the loss of immediate students' feedback in classroom interaction, among other issues [44]. With respect to the variety of components of the pedagogical process, many studies coincide in pointing at assessment quality as the most important challenge during ERHE [45][46][47][48][49], which includes a set of challenges from proctoring measures to prevent students' dishonesty to providing useful feedback and promote meaningful learning from the competence-based perspective of current curricula. Our study specifically pursues the following goals: • To describe the ERHE process during 2020 in three Latin American countries , comparing instructors' and students' perspectives since teaching and learning processes cannot be understood only from a single perspective. --- • To identify some variables of ERHE affecting students' and instructors' perceptions of learning. We will particularly look at sex and educational cycle in the case of students as the main personal variables likely to affect subjects' experience. --- • To identify improvement needs for future experiences of ERHE, considering the previous analyses. --- Materials and Methods A non-experimental quantitative design was employed to analyze the perceptions of students and instructors regarding the remote teaching and learning process. This research is of correlational scope, and the methodological approach used was a crosssectional survey. --- --- Instruments In line with the theoretical framework related to successful virtual education experiences prior to the pandemic, two instruments with similar characteristics were designed for online administration, one targeting students and the other aimed at instructors. Both instruments addressed the same dimensions and aspects in order to fulfill the comparative purpose of our study [50]. Four evaluators with experience in educational psychology and psychometric analysis evaluated the indicators derived from the literature review. The evaluators followed an evaluation guideline that aimed to assess the relationship between each indicator and the theoretical variable associated with effective practices in distance education. The dimensions analyzed were empathy between instructors and students, teaching organization, collaborative work, quality of assessment, and perception of learning. Additionally, they evaluated the quality of the indicators of each dimension in relation to comprehension, length, and formal aspects. For each indicator, the evaluators provided a score ranging from 1 to 5 to represent its relationship with the theoretical dimension. The four evaluators achieved an intra-class correlation index of 0.89. The student questionnaire consisted of 65 items, while the instructor questionnaire contained 56 items. Both instruments included Likert-type items, rated on a five-point scale . The questionnaires included four open-ended questions for subjective evaluation and reflection on the participants' experience to gather qualitative aspects that could enhance the information. Regarding internal consistency, both questionnaires exhibited satisfactory characteristics. Cronbach's Alpha was calculated for each of the ten questionnaire scales. For students, the reliability of the lowest Cronbach's Alpha score was 0.79, the highest was 0.82, and the average was 0.81. For instructors, the lowest Cronbach's Alpha score was 0.72, the highest was 0.83, and the average was 0.77. Table 2 shows these dimensions and the internal reliability of the questionnaire for each set of participants. --- Data Collection Contact with study participants was established through university authorities in Chile, Argentina, Ecuador, Mexico, and Colombia, with which we had academic links. Participation was obtained from universities in Chile, Ecuador, and Mexico, countries located in Latin America that share various similarities in terms of language, culture, and indigenous heritage. Likewise, they face social and economic challenges like inequality, poverty, and the need to improve their education systems. The survey was made available online using Google Forms. The questionnaires were distributed during the second semester of lockdown in 2020, a period of emergency teaching and learning. Students and instructors were invited to participate from their respective universities, which centrally sent out the surveys and invitations to participate voluntarily and anonymously via institutional email. Through the institutional email, we were able to obtain almost 95% of the sample for this study. Additionally, the snowball sampling technique was employed. Participants who completed the survey shared it with their peers in other universities, thereby expanding the survey's reach to the largest possible number of practicing university instructors and students. Through this technique, 100% of the sample presented in this article was completed. In this way, the sample encompassed a wide range of cases within the accessible population from different disciplinary areas, considering both public and private universities within each participating country. The Ethics Committee of Universidad del Desarrollo approved this research in June 2020. The approval certificate ensures compliance with all ethical safeguards of scientific research, such as confidentiality, anonymity, voluntary participation, the right to withdraw from the study or choose not to answer, the absence of risks associated with participation, and the competence of the research team. This certificate of ethical approval was provided to the universities invited to participate in the study. The questionnaire used in the study included informed consent, which explained the research objectives and the ethical safeguards involved. The ethical standards and codes of conduct followed included those of the American Psychological Association , the Code of Ethics of the Chilean College of Psychologists , as well as Chilean legislation on scientific research , and the 1975 Declaration of Helsinki revised in 2000. --- Data Analysis Plan Analysis of Variance was used to test for statistically significant differences between groups, and multiple linear regression was conducted to develop explanatory models. The reliability analysis of the scales was carried out using Cronbach's alpha coefficient. All these analyses were performed using R version 4.3.0. --- Results This study specifically aims to describe the ERHE process during 2020 in Chile, Mexico, and Ecuador, both from instructors' and students' perspectives, comparing participants' experiences and perceptions of the quality of the teaching and learning process. Tables 3 and4 show the global results for both the participants' perspectives, students' and instructors'. In the following subsections, we will address the concrete results for each research goal. Students and instructors report that the weakest aspect of education is the quality of the assessment of learning. This difference between assessment and the rest of the studied dimensions is significant for students = 1365.63, p < 0.001) and instructors = 853.99, p < 0.001). On the other hand, students' evaluation of remote education is more critical across all domains than instructors' evaluation of their ERHE experience . Students rate all analyzed dimensions significantly lower than instructors. --- Remote Education Process during 2020, Both from Instructors' and Students' Perspectives in Chile, Mexico, and Ecuador The results of the three Latin American countries studied in relation to the dimensions of Empathy, Didactic organization, Collaborative work, Quality of Assessment, and Perception of Students' Learning are presented in the following subsections. We will also analyze whether there are differences in these dimensions according to the gender of the participants and the disciplinary areas. In the case of the students, we will also report whether there are differences according to the level of education they are studying . It is important to note that there are significant differences between Chile, Mexico, and Ecuador in relation to the technological and physical resources present at the beginning of ERHE = 18.27, p < 0.001). When making comparisons between student groups through the Tukey-HSD test, results show that students from Ecuador report having fewer physical and technological resources for remote education , while Mexican students are the ones who have the best conditions . On the other hand, students studying degrees related to the area of education and biological sciences report having the least physical and technological resources for ERHE compared to the rest of the degrees = 18.68, p < 0.001). There are no significant differences for students by gender = 0.71, p = 0.50) or the cycle they are studying = 0.03, p = 0.97). Regarding instructors, there are no significant differences. --- About Empathy in ERHE Table 5 presents results concerning the first dimension: Instructor empathy versus students' perceived empathy. We found significant differences in the perception of empathy between Chilean students and those from Mexico and Ecuador = 117.80, p < 0.001). Also, when making comparisons between groups through the Tukey-HSD test, Chilean students report perceiving lower empathy from their instructors during the pandemic . Moreover, there are significant differences by gender = 6.86, p < 0.001); female students perceive greater empathy from their instructors. Significant differences also exist by cycle, with intermediate-level students reporting lower perception of empathy from their instructors = 13.58, p < 0.001). Intermediatelevel students differ from both initial-cycle students and final-cycle students . Furthermore, there are significant differences between disciplines = 38.41, p < 0.001); students in the education and social sciences fields perceive greater empathy from their instructors, while students in engineering and biological sciences perceive less empathy from their instructors. Interestingly, in all three countries, students in the education and social sciences fields are the ones who report having the most empathetic instructors = 16.64, p < 0.001; Ecuador: F = 5.71, p < 0.001; Mexico: F = 4.94, p < 0.001). Regarding the instructors, there are no significant differences between countries = 1.60, p = 0.20). However, there are significant differences by gender = 3.04, p = 0.002). Female instructors perceive themselves as more empathetic than males. On the other hand, there are significant differences between disciplines = 5.09, p < 0.001). Instructors in education majors perceive themselves as significantly more empathetic than instructors in the field of biological sciences . --- About Didactic Organization in ERHE Table 6 presents results regarding the second dimension of the questionnaires: Didactic organization. Significant differences are observed in the experience of students from the three countries = 72.28, p < 0.001). When making comparisons between groups through the Tukey-HSD test, it is observed that Chilean students report the remote educational process with lower didactic organization , while Ecuadorian students report a better perception of this variable . There are significant differences by gender = 6.16, p < 0.001). Female students perceive the didactic organization of their virtual classes more positively . Additionally, there are significant differences by cycle = 19.90, p < 0.001): intermediate-level students report a more negative perception in this area. There are also significant differences by disciplinary area = 11.17, p < 0.001), with students in education majors reporting a better experience in this variable and students in engineering fields reporting the lowest . Regarding teaching strategies, students report that the most frequently used ones in ERHE are technological expository , written response , dialogic, and traditional expository . The least used strategies were collaborative and gamification approaches. There are significant differences in the perception of students from Chile, Mexico, and Ecuador regarding students' participation = 47.30, p < 0.001). When making comparisons between groups through the Tukey-HSD test, Chilean students have the most negative perception . Regarding the instructors, we found significant differences between Chile and the other two countries = 6.49, p < 0.01). Chile has a lower mean . There are no significant differences by gender, nor are there significant differences by discipline area. Furthermore, through a Chi-squared test of independence, teaching strategies reported by the instructors were compared. It was found that Chilean instructors report significantly fewer dialogical activities than other instructors = 6.00, p = 0.04) and fewer written response activities = 22.25, p < 0.001). Ecuadorian instructors report the most technological expository activities = 1.75, p = 0.42) and traditional expository activities = 3.93, p = 0.14). Finally, Mexican instructors report the most significant collaborative activities = 9.96, p < 0.01). --- Collaborative Work in ERHE Table 7 presents the results of the third dimension: Collaborative work. We found significant differences between Ecuadorian students and the rest of the countries = 115.00, p < 0.001). When making comparisons between groups through the Tukey-HSD test, results show that Ecuadorian students report a higher presence of collaborative work than the others . Also, there were differences by gender = 5.73, p < 0.001), with women reporting a higher experience of collaborative work. Additionally, there are significant differences between academic cycles = 16.89, p < 0.001). Intermediate cycle students express a significantly lower level of collaborative work. On the other hand, there are also differences among students from different disciplines = 21.28, p < 0.001). Students in the field of education significantly differ from other areas, reporting higher levels of collaborative work. Regarding the instructors, there were significant differences between Chile and the other two countries = 17.53, p < 0.001). The Chilean sample has the lowest scores. There are also gender differences = 3.90, p < 0.001), with female instructors having a higher perception of the implemented group work. There are no significant differences between disciplines = 0.18, p = 0.97). --- Quality of Assessment in ERHE Table 8 shows the results of the fourth dimension of the questionnaire: Quality of Assessment perceived by students and instructors. We found significant differences between Chilean students and students from Ecuador and Mexico regarding the assessment of learning = 184.90, p < 0.001). Using the Tukey-HSD test, it is observed that Chilean students have the lowest score, indicating the most negative opinion regarding the quality of education . On the other hand, Mexican students are the ones who show the best appraisal . We found significant differences between men and women = 6.14, p < 0.001), with female students having a more positive perception of the quality of assessment of learning. Also, there are significant differences between academic cycles = 12.29, p < 0.001). Students in the initial cycle presented a more positive opinion than the other two cycles. Finally, there were differences between students from different disciplines = 56.54, p < 0.001). Students in the fields of engineering and biological sciences have a significantly more negative opinion of the quality of assessment compared to students in the Education field. Through the ANOVA test, significant differences are observed between Chile and the other two countries in the case of instructors = 31.79, p < 0.001). When making comparisons between groups using the Tukey-HSD test, we noticed that Chilean instructors have the lowest mean . In other words, they have the most negative opinion about assessment in ERHE. There were also significant differences by the gender of the instructor = 2.71, p = 0.007), with female instructors having a generally better opinion of the quality of assessment. Additionally, there are significant differences by disciplinary area = 3.39, p = 0.005), with instructors from education-related fields reporting a better experience in this variable and those from engineering areas reporting the lowest. Using the Chi-squared test of independence, we observed that Chilean instructors report higher use of open-ended items than the rest = 4.81, p = 0.09), and Ecuadorian instructors significantly use more closed-ended items = 65.62, p < 0.001). There are no significant differences in the use of performance-based tasks among instructors from different countries = 1.79, p = 0.41). On the other hand, female instructors report fewer closed-ended responses and more open-ended responses and performance-based tasks. Instructors from the education area significantly use more open-ended items and performance-based tasks, while instructors from biological sciences use more closed-ended items. --- Perception of Students' Learning in ERHE Table 9 presents results concerning the fifth and last dimension of the questionnaire: Perception of learning during ERHE. We found significant differences between Chilean students and students from Mexico and Ecuador = 190.30, p < 0.001). Chilean students show a significantly lower score in their perception of learning. Additionally, there are differences between Mexico and Ecuador and in favor of Ecuador. There were also significant differences between men and women = 6.68, p < 0.001); women had a higher perception of learning. There were also significant differences between cycles = 28.58, p < 0.001). Intermediate cycle students perceive significantly lower learning. Finally, there are significant differences between discipline areas = 13.54, p < 0.001). The perception of learning differs among students in education and health-related careers, on the one hand, and students in engineering and biological science fields, on the other. The former group perceived achieving higher learning outcomes. Regarding instructors' appraisal, there were significant differences between Chile and the other two countries = 37.25, p < 0.001). When comparing the groups using the Tukey-HSD test, it is observed that Chilean instructors reported a lower mean. In other words, they perceive that their students learned significantly less than the students from other countries, comparing the other instructors' perspectives. There are no differences by gender. --- Predictive Model to Explain Perception of Student Learning from the Perspective of Both the Instructor and the Student Multiple linear regression was performed to create explanatory models of student learning perception as viewed by both instructors and students themselves. Table 10 displays the outcomes from the teacher's standpoint, while the student's perspective is presented in Table 11. Table 10 shows that the participating instructors perceived that their students learned more in the ERHE education when the instructional organization of their synchronous classes was effective, the evaluation methodologies employed were reliable, the students were not in the intermediate cycle, and they did not belong to Engineering or Biological Sciences undergraduate areas. Table 11 exposes that the participating students perceived more learning in the ERHE when the instructional organization of their synchronous classes was felt effective, instructors displayed empathy towards their students, and peer collaborative work exercises were conducted. Additionally, this trend was observed among female students who were not in the intermediate cycle. Noteworthy, for instructors, pedagogical variables explain their students' learning ; however, for students, in addition to instructional aspects, the instructor's empathy and collaborative work carried out in classes are relevant. These variables encompass undeniable socioemotional components. --- Discussion Due to the pandemic, the implementation of remote education in 2020 was developed in an emergency. In this study, we intended to describe Chilean, Ecuadorian, and Mexican students' and instructors' evaluations on several aspects of the teaching and learning process, paying attention to sex and educational cycle in the case of students, in order to be able to identify needs of improvements for future cases. Initially, educational institutions were concerned with internet access, providing equipment to enable connections, and creating physical spaces to facilitate the distance learning process. However, these basic conditions were not assured. Many students faced challenges due to the lack of access to electronic devices and a stable internet connection, as evidenced by the previous studies [13,16]. In these three national contexts participating in our study, many instructors had to learn on the fly how to use technologies for an online teaching and learning process. Initially, this remote education was a transfer, adaptation, or exact copy of face-to-face education. There was no time for special preparation. As instructors gained experience, nevertheless, it became increasingly clear that a different approach was needed [30,33,36]. The first point of discussion of our results is that students' and instructors' perceptions differ regarding the same aspects, and this pattern is consistent across the three studied countries. In general, instructors tend to have a more favorable perception than students in the five dimensions examined: empathy, didactics, collaborative work, assessment, and learning. Students, as recipients, revealed themselves as more critical than instructors and showed more disagreement with ERHE. These findings align with those of [15], who demonstrated that many students reported a loss of quality in online education compared to face-to-face education. This also corresponds with the findings of [6], who highlighted that students faced challenges in staying focused on online assignments and exams due to distractions. Additionally, they missed the face-to-face interactions with instructors and classmates, which impacted their ability to ask questions and receive real-time feedback, which reminds us of the need to carefully design even the smallest details of inter-peer interaction to facilitate a successful learning experience [37]. On the other hand, one issue on which instructors and students agree is that the area of assessment of learning is the least successful. Assessment remains the Achilles' heel, even more so in online education contexts, as pointed out by [26]. The assessment system generates mistrust among all participants. On the one hand, there were concerns about the possibility of dishonest practices , which raises doubts about the academic integrity of students. Indeed, there were doubts about whether the grades obtained by students truly reflect what they have learned. This is associated with the research findings of [9], where they identified several problems in assessing online learning. One of the main issues they found was the lack of adequate tools and methods for assessing online learning. Instructors also pointed out that it is challenging to assess student participation and engagement online, as well as the authenticity of work submitted electronically. Additionally, instructors mentioned that the absence of face-to-face interaction with students hinders effective feedback and assessment of learning. Pre-pandemic literature already warned about the need to take specific measures for assuring learning assessment in the online context [35], thus guaranteeing that the online space goes beyond a medium of delivery of students' solved assignments or a means of test-taking but evolves into a safe space for effective feedback. In this regard, studies have shown [48] the importance of moving towards more authentic, contextualized, and challenging assessments, where students are required to apply knowledge, make decisions, and demonstrate performance. This entails incorporating more frequent performance-based tasks and the use of open-ended items. Instructor feedback plays a key role in remote education, as it allows for monitoring of learning and encourages student self-regulation. Unfortunately, the main preoccupation of instructors during ERHE was in proctoring measures. The good news is that remote education has allowed some progress to be made. A systematic review by [49] analyzed how the COVID-19 pandemic affected assessment in HE. The study found that traditional methods, such as paper-and-pencil examinations, have become less common while project-based assessments have become more popular. Additionally, it showed that online assessment methods can be more efficient and effective than traditional methods. The cross-country comparison also reveals certain trends. In general, both Chilean instructors and students tend to be more critical of their experience in emergency remote education compared to Ecuadorian and Mexican participants. Chilean students, in particular, perceive less empathy from their instructors and have a more negative perception of remote education's didactics and assessment. Moreover, Chilean instructors are the ones who perceive most that their students learn less, and they also consider themselves less dialogical in their teaching compared to their peers in Ecuador and Mexico. In contrast, in the Ecuadorian sample, students, despite having the least technological conditions for remote education, perceive greater empathy on the part of their instructors. Instructors stand out for engaging in more tutoring activities. Finally, Mexican students are the ones who have the most technology for ERHE and are most likely to engage in performance-based tasks for assessment. Instructors stand out for engaging in more collaborative work activities. It is important to mention that in Chile, the pandemic occurred immediately after the social upheaval experienced since October 2019. This meant that many university students had their first experience of remote education during a time of social unrest without the necessary conditions, and the results were not very positive. Therefore, the ERHE pandemic possibly started with a negative predisposition due to previous experiences and also because of the emotional impacts caused by the political and social polarization during those times. Our second goal pointed at differences between participants' views considering sex differences and study experience in the case of students. Regarding sex, we found noteworthy results. In most of the dimensions assessed, females, both students and instructors, tended to make more positive evaluations of ERHE than their male counterparts. We found no previous study that could help us interpret these results; we would want to underline them. The previous worldwide study of Aristovnik and colleagues [15] did not throw significant differences between students by sex, which might indicate that differences could be linked to Latin-American cultural particularities in addition to sex. Regarding study experience, there is a negative trend in the evaluations of ERHE by students in the intermediate cycle compared to students in the initial or terminal cycle. This could be interpreted based on the experience and comparison that third-and fourth-year students make between face-to-face and remote education during crucial years of their professional training. Such a comparison cannot be fully made by students in the initial cycle since they lack enough university experience and can hardly compare to other kinds of study practices at HE, and students in the final cycle take it less seriously as they are close to graduating from their degree course. Regarding comparisons between disciplines, there is also a trend that has been previously discussed by [25]. Students studying education-related careers tend to share a more positive evaluation of the various dimensions of ERHE. The opposite occurs with engineering-related careers, which tend to have a more negative view. The additional difficulty of communicating mathematically in the online context could help understand these results [31]. It would be interesting to delve into what factors contribute to women in general, and students studying education-related careers in particular, having a more positive impression of the implementation of ERHE. Similarly, understanding why engineering students, male students, and those in their third and fourth year of studies have a more negative evaluation of this process. Future studies could be carried out to further investigate these aspects. The results of this research enable us to identify areas for improvement in future experiences of ERHE, as was our third goal, particularly generalizable to the Latin American context. For instance, the instructional design should be modified to be more dynamic, more dialogic, and to give greater prominence to students. Attention should also be focused on the instructor-student bond, including aspects like containment, visibility, and dyadic relationship, as it has a significant impact on learning, particularly during emergencies. Additionally, it is crucial to ensure clarity in instructions and to consider the needs of the learner, as discussed in the article by [14]. There is a need to enhance dialogic, collaborative, gamified, and technology-based teaching activities, especially considering that didactics and dialogic interaction are two variables that significantly predict remote learning, according to [43]. Additionally, it is known that communication between students and instructors was negatively affected during ERHE, which had an impact on the quality of learning, as stated by Senanayake and colleagues [13]. Technology-based remote education is here to stay, offering many opportunities and positive aspects in terms of learning and outreach. Prior to the pandemic, ref. [40] explored students' perceptions of learning regardless of the course delivery method and online environment. The results indicated that perceptions of face-to-face learning were higher than those of online learning in terms of social presence [37], social interaction, and satisfaction. However, there is always a place for contrasting views. In previous studies, some students even felt very comfortable with online learning, as it gave them the opportunity to be innovative through the use of computer technology. On the other hand, Bustamante and colleagues [18] identified benefits of ERHE, such as the flexibility and accessibility of online learning. Students reported greater ease in accessing learning materials and organizing their study time, enabling them to learn at their own pace and on their own schedule. Additionally, some students mentioned that online education allowed them to save time and money by avoiding daily commuting to the university. Instructors also reported greater flexibility in lesson planning and the possibility of reaching a wider audience. It is, therefore, necessary to gain a deeper understanding of how students perceive and react to ERHE, as the perception and attitude of students and instructors are critical to motivation and learning. The pandemic has had a significant impact on the lives of HE students and instructors around the world. Students reported higher levels of anxiety, stress, and depression during the pandemic, negatively affecting their emotional well-being, learning, and quality of life [15]. Future research on remote education should consider these aspects. --- Conclusions and Practical Ideas Remote education can have significant benefits and opportunities for both students and instructors if we learn from the experience of ERHE during the pandemic. The possibility of saving time on transportation, having greater flexibility in the educational process, studying at one's own schedule and pace, and promoting a more sustainable education are certainly benefits in today's world and for the future. However, technology-based HE cannot be reduced to these logistic aspects. We know that the instructional design of ERHE should be different from face-to-face education. Logistically, it becomes necessary to consider, for instance, shorter synchronous classes and ensure students have the minimum and basic conditions for internet access and technological devices for distance learning. Emotional aspects are also crucial. There is a focus on the importance of stronger connections between instructors and students, as well as among students themselves, as face-to-face interaction and social relationships are missed. Additionally, showing empathy towards students in their new learning process is essential. Pedagogical aspects cannot be overlooked. Instructors must strive for increased student participation and dialogue, prioritize the curriculum, and clarify what is essential. Undoubtedly, implementing assessment is a priority. Students need meaningful and realistic assessments that allow them to apply their knowledge. It is equally important to provide opportunities for feedback from instructors to clarify doubts, learn from mistakes, and reinforce their successes. Both students and instructors perceive instructional organization as a crucial variable for explaining learning. In this sense, technology should be at the service of learning and the user, transforming learning. Not just transferring the face-to-face classroom PowerPoint to the online environment. To achieve this, we must pay attention to the protagonists of the teaching and learning process, students and instructors, in order to respond to instructors' need for professional development and to the learner's perspective and learning needs. --- Data Availability Statement: Access to data shall be offered by the authors by request. --- --- Author Contributions: Conceptualization, A.R. and V.V.; methodology, V.V.; validation, V.V. and A.R.; formal analysis, V.V.; resources, V.V.; data curation, V.V.; writing-original draft preparation, A.R. and V.V.; writing-review and editing, A.R. and V.V.; supervision, A.R.; project administration, V.V.; funding acquisition, V.V. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

The COVID-19 pandemic forced Higher Education institutions all around the world to revise their praxis and update tools and numerous procedures. This study offers a comparative analysis of three Latin American students' and instructors' perspectives on a selection of both the pedagogical and emotional aspects of higher educational life that were affected. The report is about the participants' perception of others' empathy, their evaluation of the organization of teaching and learning and of collaborative learning experiences, their appreciation of the quality of learning assessment practices, and, eventually, their perception of learning. An exploratory study was carried out, based on survey research in Likert-scale form, responded to by 2742 students and 926 instructors. Significant differences were found among subsamples, with Chilean students and instructors having less favorable views in all dimensions compared with their Mexican and Ecuadorian counterparts; also, differences were found with respect to educational levels, discipline areas, and participants' gender. This study contributes with a double-sided view of both protagonists' perspectives (students' and instructors') to reinforce the importance of instructional design, instructorstudent bond, meaningful and realistic assessments that allow for the application of knowledge, and opportunities for feedback in post-pandemic virtual education.

Introduction Given the inequalities in health status that continue to prevail in European societies, it remains crucial to develop effective policies in response. People in lower socioeconomic groups die at younger ages and suffer from a higher prevalence of health problems such as chronic conditions, mental health problems, and disabilities [1]. In Europe, two types of health disparities create concern: the disparities among social groups and those among member states. For instance, a 2010 report from the World Health Organization compared the life expectancy rates of European member states that joined the EU before and after 2004; the report discovered that for those joining after 2004, the life expectancy was 5.6 years lower [2]. This alarming situation has made improved health a key priority of the Europe 2020 strategy to promote at smart and inclusive growth. In the inclusive growth goal, the EC highlights the need to develop a major effort "to combat poverty and social exclusion and reduce health inequalities to ensure that everybody can benefit from growth" [3]. Health disparities are particularly worrisome among vulnerable and socially excluded groups such as the Roma who have been the focus of our research. As various international institutions have reported, it is crucial to acknowledge the blatant exclusion that the Roma people experience within the European Union, particularly regarding health. Specifically, their estimated life expectancy is ten years shorter than that of the general population [4]. The European Commission and the WHO report that this statistic is directly related to the exclusion of the Roma from society [5]. Researchers have traced the health disparities among different social groups to a range of socioeconomic factors including differences in educational level attained, occupational class, and income; gender and race also play significant roles [6]. The associations between these factors are also important; for instance, low income affects educational attainment [7]. In addition, inadequate housing, poor nutrition and health-related behaviours [8], and other factors such as social ties in the local community [9,10] are all powerful determinants of the poor health conditions that affect millions of people around the globe. The WHO's Regional Office for Europe, in its first Report on the Social Determinants of Health and the Health Divide discusses the most relevant social determinants that impact individuals' health: childhood living conditions, education, working conditions and employment, welfare and social protection, poverty, communities, and health systems. The WHO has repeatedly noted that the rates of infant mortality are higher among the children of mothers with no education, especially in certain countries. For such women in Bolivia, for instance, the rate is greater than 100 deaths per 1,000 live births [11]. A 1999 WHO report stated that the poorest and least educated people not only live shorter lives but also are more likely to suffer from disabilities [12]. In particular, the educational attainment of mothers has been found to determine child mortality ratesmaking female education an urgent focus in reducing infant mortality [13]. Available health-related data on rates of premature mortality show that the rates are significantly higher among people with lower levels of education, lower occupational classes, and lower income [1]. Thus, educational level is a key factor influencing health conditions. International institutions such as the WHO and the OECD have provided evidence for this direct relationship [11,14,15]. Indeed, "education is positively associated with a variety of social outcomes, such as better health" [16]. --- Health among the Roma People The European Union acknowledges that the Roma people are one of the most marginalised social groups, "facing deep social problems related to low levels of education, high unemployment, inadequate housing, poor health, and wide-ranging discrimination" [17]. In particular, they face serious health situations because of their disadvantaged social position. In addition, many researchers note that few rigorous systematic data are available on the Roma community's health condition [18][19][20]. The data on Roma childrens' health statuses are inconsistent, due to the challenges of monitoring ethnic groups in Europe, but fragmented evidence indicates high levels of maternal and child mortality and morbidity [5,21]. Furthermore, data from the European Survey on Health and the Roma Community [22] show a high dependency rate among the Roma in Europe: 62 not economically active people for every 100 active ones, as opposed to 48.7 for the EU27. Moreover, the available data support the fact that the socioeconomic situation of the Roma affects their health conditions. Along this line, the UNDP [23] report that hunger is a problem in the poorest Roma communities. Of the Roma surveyed by the UNDP in south-eastern Europe, 53% reported going hungry in the previous month, compared with 7% of the non-Roma living in close proximity to these Roma settlements. The Roma population has also shown a high prevalence of chronic diseases such as migraines, hypertension, and arthritis; dental problems and difficulties with vision and hearing are also widely reported. Again, the European Survey on Health and the Roma Community lists factors that explain the poor health conditions of many European Roma citizens; among them are a lack of resources, poor housing conditions, inadequate employment opportunities, and difficulty gaining an appropriate level of education. The Roma also have inadequate access to prevention services: over 25% of Roma children do not adhere to vaccination schedules and 40% of women over 15 have never been to a gynecologist for reasons other than pregnancy or labor [22]. In contrast, particular cultural traits among the Roma can function as health assets. Among other traits, solidarity within the community, strong family bonds, and cultural norms and traditions surrounding health might help explain-and improve-the health conditions of the Roma people [24,25]. There is a lack of quantitative data specifically associating the health statuses of the Roma with their educational levels. However, the general association between low educational level and poor health status is especially threatening in their case, as the rates of exclusion that the Roma face in education, among other areas, are particularly alarming. One study that focused on the Roma population over age 15 in a selection of European countries with significant percentages of Roma people found that 44% of Roma individuals had not completed primary school, and only 24% had at least a secondary education. Moreover, young Roma have a higher prevalence of risk factors for certain diseases linked to lifestyle and low educational status [26]. Thus, educational factors are recognized as having a significant impact on basic health indicators such as perceived health status. One of the few sources providing data about health conditions and education among the Roma is the report on The health situation of the Roma communities [27], which indicates, for instance, that 15% of Roma participants over age 14 who have finished post-secondary school suffer from chronic diseases, compared to twice that many , among those who have no formal education. Researchers have found that, among other social factors, reduced inequality correlates with better health status and life expectancy [28], which is essential to overcoming the alarming health conditions in Roma communities [29]. In this context, studies have revealed the need to more strongly emphasize social inequalities [30] and to promote multi-sectoral policies that aim to reduce poverty by improving educational levels [31] to reduce the health disparities among the Roma communities and other ethnic minorities and vulnerable groups. In this regard, the academic community, along with the Council of the European Union and the European Commission, emphasize that substantial and effective efforts must be made to reverse the educational exclusion facing many Roma people across Europe to help reduce the health inequalities they experience [17,32]. --- Health Condition and Educational Level Not surprisingly, people's health improves when efforts are made to improve healthcare systems and to address other socioeconomic factors that have impacts on health, particularly education [33]. Indeed, the European Commission and other institutions highlight education as one of the four pillars of the social and economic integration of the Roma; they describe the potential of education to improve health and the impacts that health-related behaviors have on educational attainment [17]. Studies addressing the role of education in improving health have analyzed this relationship from the perspective of the educational level attained [11,16,[34][35][36]. Given this focus, studies tend to neglect other important educational interventions or activities that could influence health. Although opportunities for lifelong learning have been recognized as helping to determine individuals' health conditions [37], no systematic analysis has been conducted on how adult educational interventions could impact adults' health conditions in OECD countries [38]. This research gap, in turn, indicates a potential gap in the way education interventions are analyzed: do they consider the relationships between education and health/wellbeing? Indeed, it is widely accepted that education can directly influence health: greater knowledge contributes to better health. The influence might also be indirect: better jobs and higher incomes promote better health. These potential relationships are described in the WHO Levelling up report on European strategies for tackling social inequalities in health [37]. This report suggests that other educational elements might also play key roles in reducing health hazards, but those elements cannot be identified in analyses that focus exclusively on individual educational levels. That is, although the literature clearly shows the relationship between education and health quality, other factors related to involvement in educational activities might also be influencing those health improvements. For instance, a study on ways to overcome educational inequalities found that active participation and community engagement in decision-making in schools can help reduce health-related problems [39]. For all these reasons, then, further research is needed on the associations between education and health outcomes, especially accounting for groups at various socioeconomic levels [38]. Such research could reveal educational actions that can improve the health of particular vulnerable groups with low levels of education such as the Roma community addressed here. In this article we address this need and explore the particular ways in which two successful actions that are focused on encouraging Roma people to participate in their local schools are improving health statuses. The integrated INCLUD-ED project defined SAs as evidence-based actions in various social areas that are universal and transferable, and have been effective in addressing social inequalities and fostering social cohesion [40]. --- Methods --- Data Collection The analyses presented here stem from the INCLUD-ED Integrated project , the largest research project on schooling funded by the 6th European Commission's Framework Programmes for Research [40]. The project's general goal was to identify educational strategies that either contribute to social cohesion or lead to social exclusion. To this end, the project conducted a comprehensive and integrated analysis of European school systems, educational reforms, and educational practices that increase educational success. The project also explored the impacts that educational inclusion has on other areas of society such as employment, housing, health and political participation. The study was designed to focus on specific vulnerable groups to analyze the actions and policies that could promote their inclusion through education. In the context of this larger project, 26 case-studies of successful schools were carried out in different European countries. To select the cases, starting with data available from governmental sources, we selected those that most closely met three main criteria: schools that were located in low SES areas with a significant presence of the vulnerable groups defined within this project, obtained better educational results on standardized tests than average schools in similar contexts and had high levels of family participation. The findings we present here are part of a four-year longitudinal case study [39] of a school located in La Milagrosa, a ghetto neighbourhood in the city of Albacete in southern Spain with a majority Roma population. This neighborhood is characterized by high levels of poverty; data from the Spanish Ministry of Education, Social Policy and Sports reflect that, in 2008, 35% of the inhabitants of this area were social welfare recipients, 7% were illiterate and 79% had not completed elementary levels of education. All of the INCLUD-ED research, including this case study, was carried out through Communicative Methodology [41], which is based on thinking from contemporary social scientists, including the communicative rationality as understood by Habermas [42] and the process by which humans construct knowledge through dialogue, which Mead calls symbolic interactionism [43]. This methodology was developed to create the conditions for egalitarian dialogue between researchers and the individuals they are studying; such dialogue breaks with the traditional interpretative hierarchy. This dialogue makes it possible to integrate existing knowledge from the academic community with the life worlds of the social actors using a communicative process that continues throughout the research: from the design of the research to the interpretation of the final results [44]. The new knowledge created through this process is relevant to the actors involved and, thus, is more useful for throughout society. This approach is especially valuable in research with vulnerable communities such as the Roma people, whose voices have traditionally been excluded from academic research, preventing studies from having a more relevant and useful impact [45]. To conduct the longitudinal case study the project team collected both quantitative and qualitative data. More specifically, each yearly round included 13 standardized open-ended interviews ; 13 communicative daily life stories to end-users ; one communicative focus group with professionals working in the local project; five communicative observations, and two questionnaires completed by the end-users . In particular, the data from the fourth round , which are analyzed here, consisted of 19 communicative daily life stories conducted with Roma people, along with nine open-ended interviews . As in most qualitative works, the participants were selected following the criterion of purposeful sampling [46], selecting the cases according to the stratification based on the previous knowledge in order to select those cases which could help better to respond to the research question. These techniques were the most appropriate for this research since the aim was the comprehension of the meanings that actors gave to their actions, and how they transferred the abilities and knowledge acquired through their involvement in the educational field to other contexts of their daily life, such as health and care. The research followed the criterion of "credibility", "transferability", "dependability" and "confirmability" which, as has been widely argued by qualitative methodology experts, are the base to ensure the validity of the results [47,48]. The data gathered in the phase presented here relates to self-reported health perceptions, which according to literature, is a good predictor of objective health status [49][50][51], as well as qualitative perceptions of education professionals and the other community representatives interviewed. Nevertheless future research could benefit from triangulation of this information with other sources, such as measures obtained from health records and qualitative research conducted with professionals of the primary care unit of the area. The voices of the Roma families and members of the La Milagrosa community are central to this study. Most of these participants have low incomes and little education beyond the primary level. They are all neighbors in the La Paz school community, and when they first became involved in the school, some were literate. These Roma adults had dropped out of school many years earlier, if they had attended at all. Additionally, few had jobs in the formal labor market, and they faced serious economic challenges, including how to feed their children. The community had experienced such a critical level of social exclusion that some had become involved in the vicious circle of drug abuse and dealing among other illegal activities. However, the transformation of the school into a Learning Community improved the situation of the whole community in a range of very important ways. The INCLUD-ED project was based on an analysis of Successful Educational Actions , which are defined as the actions that contribute to school success and living together. In other words, SEAs contribute to overcoming school failure and early dropout, as well as overcoming the risk of exclusion in other areas such as employment, health, housing, and political participation. The SEAs identified have been shown to lead to positive progress in the results in every context in which where they have been implemented and therefore they have already been transferred to other schools and communities to improve school success and social cohesion. Success is measured in terms of reducing early school dropout and absenteeism as well as in terms of the results obtained by students on the standardized tests administered by the Department of Education of Castilla La Mancha, which evaluates ten competencies, including, linguistic communication, mathematics, knowledge and interaction with the physical world, cultural and artistic competences, treatment of information and digital competence, social studies and citizenship, learning to learn, autonomy and personal initiative, emotional competences, and English. --- Data Analysis: Using Communicative Methodology to Include the Voice of Roma End-Users The research team collected qualitative data so that it was possible to identify and analyze the impacts that the SAs had on students' educational levels and on the health conditions of those participating in the activity and of others in the wider community. In this article it is analysed one SA, educative participation, which is described below, and its impact on the health of the Roma community in this neighbourhood. The voices of the end-users are presented here to explore the relationships between education and health that developed among this Roma community. We look at how, and how much, educative participation helps improve the health of Roma people. The accounts analysed here have been selected because they illustrate well the analytical themes that represent the findings discussed. The use of the Communicative Methodology was critical in both developing the data collection process and conducting the analysis. The communicative approach requires that the research subjects participate throughout the entire project. Various elements guaranteed this participation. One such element was the Advisory Committee, comprising members of the social group that is the subject of the research-in this case, representatives of the Roma community, and has the functions to provide feedback at different stages of the research on various materials, to analyze the literature review, to define the fieldwork techniques so that they would tap as much community knowledge as possible, and to review the final report, highlighting the evidence gathered. The analysis revolved around two dimensions, namely, the exclusionary and transformative dimensions. In analyzing a given situation of exclusion, those using the Communicative Methodology identify the obstacles to reversing the situation , along with elements that could help to promote change . The transformative elements are key parts of this methodology; recommendations for ways to transform the situation can draw from both the existing evidence and the reflections and contributions of those who are experiencing the exclusion. --- Results and Discussion The INCLUD-ED Integrated project identified several SAs in different domains that are helping reverse the social exclusion of vulnerable groups. Importantly; the results of this project have also informed This article describes the ways in which family and community educative participation at La Paz school had impacts on improving the Roma end-users' health. Overall, we found that the participation of Roma families and community members in Interactive groups and Family education, both identified as SEAs [52], fostered specific cultural values among the Roma that support better health. Moreover; they transferred some of their learning in the SEAs to other contexts. Both elements helped to improve not only their own and their families' health conditions but also the general level of health in their community. --- Roma Cultural Values as Assets: Improving Health through Educative Participation Two cultural traits of the Roma-family cohesion, and mutual care and protection-play key roles in their health conditions. For the Roma, the value of family includes great respect for the elderly and authority, care for one's family members, and the observance of cultural family norms such as those related to mourning and kinship solidarity [53]. Importantly, the Roma understand family not as the nuclear family, but rather as the extended family and even the broader community. As they organize their lives, they include all people whom they consider to be near their kin, and whom they identify with their homes, their culture, and their way of doing things. Thus care and relationships also extend to a wider idea of family, more similar to that of community. This conceptualization of community or extended family is illustrated by Aroa, a 25-year-old Roma woman who has siblings and cousins in La Paz school. She explained in her narrative that "everybody knows each other in the neighbourhood and if they are not cousins, they are nephews... we are all relatives here". Thus, the Roma people have a strong sense of identity and belonging to the broad and diverse Roma ethnic group, they share identity traits, such as the family and community-based ways of organizing, the networks of solidarity established within these groups, and the value attached to dialogue and the word [54]. All of these values arise from their longstanding experiences of discrimination. Their centuries-long opposition to injustice has become a unifying force instead of a divisive one, contributing to their cohesion and survival as an ethnic group. All of these experiences have led the Roma people to be aware of and care about not only their personal wellbeing or that of their closest family but also that of the other members of the community, regardless of others' actual roles as kin. This sense of common identity motivates the strong bonds of solidarity that the Roma develop among themselves, a factor that can also help to improve their health. This trait is clearly illustrated in the words of Luis, a 42-year-old Roma father in the neighborhood: "If we Roma can be identified by something it is by this unity with this strength [as people], even more so, we are there in the moments of illness, 100%. For instance, if someone needs blood, and if I [meaning the community] can't donate, we buy it somewhere and we provide it. And we ask each other if [my blood] is good or if yours is, and if we are 100 people, the whole 100 are tested to save the one person. This is a point that I believe identifies us very much, very, very much". The analysis of the ways that Roma families and community members participated in Interactive groups and Family education at La Paz school showed that this involvement helped these families to gain access to more resources and promoted relevant skills among them; moreover, it enhanced particular Roma cultural strategies and values that helped them extend these health improvements to the whole community. The next sections explore in more depth how their educative participation helped reduce some of the health inequalities in the community and therefore enhanced their health overall. The INCLUD-ED project identified educative participation as one of the three types of participation that had the greatest impacts on children's academic achievement [55]. Educative participation means that members of the students' families, as well as others in the community, either become involved in instrumental learning activities with the children or participate in education programs themselves. Among these activities, Interactive groups and Family education were considered SEAs because they helped improve children's academic outcomes and were transferable. The analysis of the educative participation of Roma families and community members at La Paz school showed that this involvement enhanced particular Roma cultural strategies and values that helped them extend health improvements to the whole community. Interactive groups have been found to be a successful form of heterogeneous ability grouping and reorganization of human resources within the classroom; such groups allow all the students to perform better academically, and they improve coexistence. When a teacher implements interactive groups, the classroom is organized into small and heterogeneous groups of students, each with an adult who promotes supportive interactions and dialogue among the students. The classroom teacher manages the classroom and provides extra support when necessary [52]. In this strategy, family members or neighbors come into the classroom to support the teacher and participate in the children's learning. Family education activities are addressed to adults and include literacy, subjects that are part of basic primary and secondary education, and a wide set of courses that train individuals for work in such areas as childcare, care of the elderly, and monitoring playgrounds and school cafeterias. The implementation of Interactive groups and Family education in La Paz started in the 2007-2008 school year. Educational results have improved since its initial implementation, according to the Diagnostic Tests conducted by the regional government of Castilla La Mancha in all schools. As is shown in Figure 1, the variations in the pupils' matriculation provide further evidence of school success. Moreover, absenteeism for the 2006-2007 school year was 30% and it reduced to 10% the following school year. When members of families and the community come into the La Paz classrooms that are organized into interactive groups, they are in charge of encouraging supportive interactions among the children. In addition, through the family education in La Paz, the Roma people acquire instrumental competences ; this also affects their children's success in school [52,[56][57][58]. In this regard, the evidence has long shown that family involvement in educational activities has a positive impact on children's academic performance; it changes their school-related behaviors and improves their achievement. Thus, family education enriches interactions among those of different cultural and educational backgrounds, making that diversity an asset that can further improve both academic results and the coexistence among children of different backgrounds. Among the other impacts are the increased interactions between family members and the school as family members follow up on or ask for help with the child's school activities. Through both interactive groups and family education, Roma parents become active agents in their children's learning; in turn they feel increasingly useful, selfconfident, and rewarded for their work [59]. Such participation also helps to create networks of trust among participants which later translate into increased trust in the use of various social services, such as health care [60]. The focus here is how such family and community participation in schools can become a space to support health improvements among the Roma. Source: Self-compilation based on registration data and standardized tests elaborated by the Department of Education of the Government of Castilla y León. --- Increased Emotional Wellbeing through Educative Participation "I feel much happier, much more satisfied and even more valued… because I wake up with a purpose". Luis' initial involvement at the school was not his own decision. He started volunteering at his children's school while he was still in prison to comply with the requirement for community service, however, his participation in Interactive groups at the school radically changed his life. When he was released, he offered to continue as a volunteer at the school. Luis' experience is like that of many Romani parents at La Paz. As the school opened to the community through its process of becoming a Learning Community, many opportunities became available for the local Roma people who had, until then, lived at the margins of society. Luis did not simply participate in the classroom; in the Interactive groups he became a relevant resource in several ways, improving the learning processes of all of the children in the community, as well as his own. His participation in appropriate educational activities, such as the interactive groups, made him feel valuable and useful. The Roma people were interviewed agreed on this point: as soon as they became involved in the school's daily educational activities, they became happier and more motivated, and active in the regular activities of their daily lives. In addition, other people observed Luis' improvement, including the teachers, his relatives, and the children. Luis explained, --- "To be appreciated by someone is nice, to be appreciated by many people and children, and not only children but teachers, even though the appreciation of the children is important, too because before you got up without a purpose. Now you get up with another, you know, with another perspective...". New routines, habits, resources, and even sources of motivation have all improved the lives of these Roma participants and are directly impacting their personalities, ways of thinking, expectations for their future and that of their families, and ways of communicating. These attributes also affect their health and wellbeing, as described by Luis: "Well, as I said, it is because of coming here, as I said, the way of talking, the way of thinking, the way of relating to another person Now I have a rhythm, I come in the morning, in the evening, but I truly like it". For these Roma parents, the very act of participating in activities in which they see their children learning becomes a trigger that increases their emotional wellbeing; it creates meaning that improves their self-esteem. Rosario, a Roma mother of six children who was in jail for dealing drugs, explained this when talking about her involvement in interactive groups: --- "You could see how children learned, how useful you felt… teachers ask your opinion. In the classrooms you are like if you were also a teacher; they ask you to do something with the children, and then you feel useful, you feel well, you feel valued. And now even more!". They also see that being involved in activities in which they help future generations learn improves their own self-esteem, because they feel useful and see themselves doing something good for others, -in contrast with some of their past personal experience. In this vein, when discussing the impact of her involvement in the school's educational activities, Rosario explained that it had helped her as well as the children. Now she feels that she is a better and more useful person: --- "It has helped me a lot! Because in addition to helping my children, I am also learning, as it's stuff that you have forgotten and also I personally feel very useful and I feel I am a better person when I come here...". This involvement by Roma adults in the interactive groups or family education activities has given new meaning to their daily lives. They now do something that they know is positive, both for themselves and for others in the neighborhood, and this leads them to see more worth in themselves. For many in La Milagrosa, these motivations to collaborate have helped to prevent mental illness in the forms of depression and drug or alcohol abuse. --- Raising Awareness about Health Issues In addition to feeling more motivated and important in others' lives, these Roma adults also indicated that participating in the school's educational activities made them more aware of health problems in the community and the need to intervene. When they are involved in the classroom, the mothers become more aware of health conditions among their own children and the other children, leading them to improve their own attitudes about health. The teachers especially noticed this impact, as in the example below; Monica, a teacher from La Paz school, explained how family members' participation in interactive groups had an impact on hygiene conditions: --- "In my class, yes I do, because [the mothers] see it. They see how many children come in very clean, and then theirs...you know? So yes, they themselves notice. For instance one came and said: check their heads because there's lice …and when they are in class participating, they see this even more". Thus, the educative participation of these Roma adults has made them more aware of health-related issues such as hygiene, along with auditory, visual, and dental problems among the children. Similarly, people interacting within the interactive groups and family education environments constantly raise and discuss various issues related to children's health. Now, Roma family members can acquire important information that they did not have before; they also learn new strategies and resources to improve the community's health. For instance, they discuss children's vaccinations and ways to prevent and treat head lice, and they share information on dental and eye care during informal interactions with teachers, students, and other family and community members. Thus, family participants learn about health-related topics while they acquire new instrumental competences in various other fields, including math, languages, and science. Perceiving themselves as learners, they adopt a different attitude as they hear details about health care or illness, becoming more interested and more active agents in both the educational arena and their families' health. Now, when they go to the doctor they do not simply comply with instructions; they want to know and understand what the terms mean and why the doctor prescribes one treatment over another. Luis explained: "I go to the doctor, or to my daughter's doctor, and I get information with all the details; if there is anything that I don't understand well, they'll have to explain it well, because I'm interested in it For example, I might not know a particular disease, or they would say, 'you have to go to the otorhinolaryngologist', and I would have said 'yes, yes…'; I was ashamed to ask, and I'd said yes, and then once I was in the street I would tell someone that they had told me to go to the 'otolary' something, and they would say 'but will you go there? ' --- and I'd be 'but where?' And now maybe I say 'ear, throat and nose specialist' because they have explained it to me because I have asked about it. I have asked him, 'and what is that about, what is it for, what function does it have?' And I don't feel ashamed to ask." In sum, the data analysis demonstrates the ways in which educative participation has a positive impact on the Roma health, both because individuals acquire new skills and competencies and because they have generated more confidence in this new learning environment. --- Caring for the Family's Health Dolores is 26 years old and is participating in a family education course to earn her certificate in secondary education. She explained that since she became involved in the school, she has paid more attention to the health of her four younger brothers. She dropped out of school when her youngest brother was born; her mother fell into a deep depression, and Dolores had to take care of both her mother and brother and assume the household tasks. Now, as she takes care of the family, she establishes healthy routines and habits, which she sees as her responsibility. When Dolores began to participate at La Paz, she maintained her family responsibilities but still engaged in other activities that benefited both her and her relatives. Other Roma women told similar stories. Their domestic routines changed as soon as they became engaged at the school; now, they can better organize their everyday lives to make optimal use of their time. They participate at the school, assume family responsibilities, and also take good care of themselves. "There must be time for everything: to get ready, to take care of oneself, to feel well… and not be the typical Roma with the bun and the slippers and the housecoat... no, that has to change! That was in earlier times, in the first century, in the second century… one must become independent and go ahead on your own, as well I used to say that women had to be at home, washing the windows and staying with the kids and nothing else. But look at me four months ago… I don't recognize her! And some tell me, 'you used to say that!' and I [respond], 'well, I used to say that last year, but this is already a new year!' And now Dolores is not the same anymore. One has to update!" Importantly, these transformations show the positive impact on families' health qualities. These women take home all that they learn through their new interactions at the school. Dolores talks about her family's nutritional habits, especially those of her younger brothers, to whom she is now much more attentive: --- [In a joking way, talking about healthy eating habits] "I'm an artist, in that I'm worse than my mother! The kids are disgusted by me because of the vegetables, and I say, "mummy prepares boiled things, no fried stuff, because the kitchen ends up greasy and we don't eat well." [Does she really call herself mummy? She's their sister!] "I try to get my brothers not to eat chocolate pastries. And now, in school, the same -vegetables, and now with the fruit, very well. Milk? No, they want Nesquik [chocolate milk]. Well, none of that! 'You'll eat what we have. Don't ask me for funny things, I'm not going to buy them'… I try, I try with the vegetables." As the adults spend time in the educational environment, hear of issues related to health, and acquire important new skills and competencies, they begin to detect and prevent their own health problems along with those of family members. In addition, they bring the knowledge that they acquire in the school to other areas of their lives, particularly their homes. The importance of family care and cohesion among the Roma, along with communication and an emphasis on dialogue, helps to reinforce the learning that takes place at the school. Other members of the community know that if they need help with or advice about personal or health issues, they can rely on support from those who participate in the school. For instance, Aroa, who is becoming more literate through her participation in La Paz, explained how she helped her aunt, who is still illiterate, to understand what the doctor said in a letter to her: As seen here, the strong family bonds among the extended Roma become a channel through which information flows from the person involved at the school to the rest of the family, including those who have no children at the school, such as grandparents. The emphasis on family wellbeing and the relevance of dialogue and communication ensure that what one person learns will flow to others in the household. Given this phenomenon, the conversational topics within the family change with time, and they expand to include health concerns. Luis explains how this occurred in his family: "Well, I just go there and tell them everything, from an anecdote about some child to the discussions with them, all around the clock, from when I get up until I fall asleep...it is all talking about school. Besides, I like to tell my wife, my children… while we eat I explain to them, 'this and that has happened to me'. It's a joy at home… we used to communicate as well before, we talked but it was more like there was nothing to talk about..." "[ Family education courses in La Paz serve two purposes. First, the Roma participants acquire literacy skills and other competencies that they can use to capitalize upon their existing knowledge, gaining skills, for instance, in assertiveness and communication to more effectively respond to doctors' instructions or the pamphlets that come with medications. Second, they transfer their newly acquired knowledge about health issues to their families, where others then use it to make more appropriate decisions about their own health. The following quote from Luis illustrates how, within an adult education course, participants also obtain access to specialized health knowledge, which has a positive impact on their families: "Even the podiatrist has come to the adult learning school because maybe you think that you need to get surgery on your foot, and then this man tells you that your shoes are not right, and maybe you don't need any surgery Well, this man told us that we have to walk like this and that, and I went to see my kids and I told my daughter 'walk, walk,' and I said 'you have to walk like this,' as he showed us, about the movement, and so on And after seeing the podiatrist, I came home and I said 'let's see whether my daughter has this curve to her feet,' but if you don't have this information… well, then that's it." Along the same lines, Rosario, who is involved in interactive groups and family education in La Paz, explained how she is taking better care of her children; she says the same is happening in other families and that people are changing in deeply transformative ways. Her involvement even helped her to overcome her addiction to drugs, which led to deep changes in her own life and those of her family members: "By being involved in school... for instance, there are mothers whose children came to school very scruffy, and by coming here [to the interactive groups], you change your image completely. You take better care of them. This has happened to me. There was a time when I was very bad, I was on drugs. I was in a center for nearly seven months. And at that time I did not take care of my children. On the contrary, they got up on their own, got dressed, and came [to school]. Later on, when I left the center and I was well, it was me and my husband. And I have not fallen back into drugs, thanks to the school, coming to school, getting involved... for me, it is very important, very important." Overall, through such involvement, these Roma people have begun to care more about their own health conditions and those of their families. Their participation has deeply transformed their lives and future prospects at different levels. Not only do they capitalize on the knowledge they acquire and share it to benefit their families; with transformed attitudes and increased confidence, they can also better address health issues and develop more informed relationships with healthcare staff. Their particular cultural values, with the prevailing relevance of family traditions and norms, play significant roles as they improve their own health-related habits and those of their family members. Importantly, these cultural values help to extend the benefits of this participation beyond themselves to their families and even beyond their closest kin, as shown in the following section. --- Caring for the Health of the Extended Roma Community Because interactive groups and family education emphasize dialogue among all participants, they can take optimal advantage of the elements of Roma culture, such as unity and group cohesion, which, in turn, improves Roma health. For instance, when one Roma adult learns that any child in the school has a health problem, she or he reacts quickly, talking with the child's parents or other relatives to find a solution that will improve the child's wellbeing. For example, because of her involvement in an interactive group, Aroa was able to identify why Miriam, a Roma girl in the classroom, was nervous and fidgety every morning. Aroa observed Miriam's behavior and asked her about it: "Why are you so nervous? And Miriam answered, 'because I don't have chocolate milk for breakfast. My grandmother gives me coffee because she doesn't have chocolate milk.' So then… I will, of course, it's the girl's health; she is just 8 or 9 years old and drinking coffee The grandmother was misinformed..." When Aroa came to understand why this girl was so nervous, she spoke with the mother about the problem she had noticed. As a member of the Roma community, Aroa shared more cultural codes with the mother than the teacher did, so she and the teacher decided that she, rather than a school employee, would raise the issue with the mother. In talking with the mother, Aroa was able to inquire more about the girl's health; she then learned that Miriam's younger sister, only five years old, also had this unsettled behavior because of her inappropriate breakfast. Aroa's intervention was invaluable; it led to both girls joining the Morning Class, a school initiative in which children from families in need receive healthy breakfasts before starting the school day. Aroa's account vividly exemplifies several relevant elements grounded in Roma culture that are positively associated with improving health conditions and overall care among the Roma community. We see her concern about the fidgety girl and how she addresses the issue with both the girl and her mother through open and confident communication. This Roma commitment to the health of others in the community serves to increase the trust among other Roma parents who are less involved in the school; they feel relieved that their children are taken care of by the women who volunteer there. Aroa expressed that she could see that other Roma participants were also committed to the wellbeing of other children in the neighborhood. She also realized that even though she had no children of her own in the school, she was more concerned with the other children's health and wellbeing because of her involvement in interactive groups and family education: [About health] "I am more interested… I remember it and I see it. When I arrive at the school, I see that they are also interested in the health of children. I do not have children myself, but I have other relatives in school, and I see that and whether I want it or not, I also get it there too, it's transmitted. I mean, I also want to be interested in them, being family, mother or whatever." These mechanisms have allowed the Roma people to transfer their new knowledge and skills about health beyond the school and the family to the rest of their community. The experience of increased solidarity among various community members enhances relationships among all the neighbors, as they encourage each other to adopt new positive habits. Juana, another Roma mother, explained how the environment had become friendlier as families in the neighborhood cared more for each other: "Yes, yes, there are fewer conflicts. Yes, because the mothers come to the school, they maybe have a coffee, become more friendly, trust each other more, and when you are on the street you say, 'hi, so-and-so.' Maybe before we didn't say 'hi' to each other and now we do, and 'are you going to school, or have you gone?' or 'why didn't you go today?' This stuff." In addition, this positive social control has extraordinary consequences for some of the participants, as solidarity and community care have led them to engage more deeply. Vanessa, a local teacher, explains another side of Rosario's story; her life changed through her participation in the interactive groups, encouraged by community members and the teaching staff. This change deeply affected her health habits and routines as well as those of her family: "And, well, we kept on encouraging her, saying, 'come on, just come one day and try, if you like it, you can offer your ideas'… and from the moment she came, she liked it, and she keeps on coming. Everything improved. In school, her kid is much better, both in learning and behavior. At home, she has started to organize herself, her house is now cleaner . Now she has habits and routines that allow her children to get around and look for work. She is also pursuing training herself, getting involved in the education of her children and taking more care of everything. One thing led to another…" These women's stories show the deep changes in their lives as a result of their involvement in the interactive groups and family education, which then led to changes in their families and the community. Indeed, they recommend that other Roma become involved at La Paz as an effective way to improve their children's futures as well as their own circumstances and their families' health. For example, when a friend of Aroa's was depressed, Aroa urged her to become involved: "A girlfriend of mine has had deep depression, and I have tried to encourage her to take a course [in the school] or be a volunteer in the school, to keep her mind busy with something meaningful." Also increasing the impact on the community is the fact that a wide variety of Roma adults from the neighborhood are involved in the interactive groups and family education. When health issues arise, they are granted a kind of authority among Roma children and youth that even the teachers do not hold. This authority, another cultural element, also helps to improve health habits. The teachers understand how this can be an asset and use it to improve the health of all the children. Luis explains this clearly: "They caught one or two kids in the schoolyard with a cigarette, and they said to me, 'look, Luis, what if you talk to them, as a Roma and as a man? They will not shout at you or swear at you'. So I did talk to them Yes, yes, if it had been the teacher [female] they would've told her whatever, but if it's me, there's almost no need to say a word. They would get quiet, no swear words, no yelling, nothing." This vignette illustrates yet another finding: by spending more time in the school with the children, the Roma adults become able to identify problems among the children that their families often do not notice or cannot control. The cultural emphasis on community members as authorities increases the value of having Roma adults present within the school; they can exert positive social control while reinforcing the bonds of trust among Roma boys and girls and the Roma adults. --- Conclusions Researchers have taken individuals' educational levels as positive indicators of their health conditions, assuming that people with no or little education are less likely to be healthy [16,61]. This association makes even more sense for vulnerable groups such as the Roma, who receive far less formal education, on average, than the mainstream population [26]. The results that have been presented here indicate that educative participation is a successful action that has a positive impact on the health conditions of Roma people with little education. The analysis of this Roma community has shown that participating in such SAs can lead to improved health, even among people as vulnerable as the Roma. Moreover, we found that the positive impact on health by involvement in these activities transfers to other domains, including the prevention and early identification of mental illnesses, such as depression and anxiety, and health issues arising in early childhood, such as those related to hygiene, dental health, and hearing, among many others. Additionally, participating in these SAs can help people acknowledge and overcome drug abuse. The participants provided a wide range of reasons to explain this improvement. Some of these reasons are applicable to other social groups; for example, nearly anyone will feel more vitality and motivation for living when they engage in a meaningful and socially useful activity. However, given the extreme health inequality facing the Roma, our focus here has been on the explanations that are most applicable to this ethnic group. The Roma participants spoke of strategies relevant to their cultural tradition that were enhanced by their educative participation. The experiences of Aroa, Luis, Rosario, Dolores, and Juana illustrate how Roma people, even those with little formal education, can build on particular cultural norms and traditions that are supported in the context of this SA to improve their health-promoting habits. When Roma participants in the school feel that the children's health is also their concern and responsibility, they become even more involved in their own personal care and in identifying health problems, whether physical or psychological, that affect their children or other people's children. In doing so, they gain knowledge and resources about health that others in the community may not have; however, because the interactive groups and family education foster family and community communication, which are also key Roma values, these Roma adults transfer their new knowledge to others in the community. We also found that when these people faced new health problems, the solutions they generated were based on their own cultural values and strategies, such as involving members of the extended family and engaging in dialogue with their children and families. Direct communication and open dialogue appear to be the core of community-based problem resolution among the Roma people due to the strongly rooted value within Roma culture that helps reduce health hazards and increase health-promoting behaviors. Both their deep emphasis on family cohesion and their strong sense of belonging to the same ethnic group are reinforced when Roma people participate in community-based learning spaces, such as interactive groups and family education. Participants in these SAs use strategies grounded in Roma cultural values, which ultimately benefit not only their own families but also the extended Roma community. Importantly, these two educational actions, interactive groups and family education, share two key principles that overlap well with the health improvements in the Roma community: the goal of optimizing participants' instrumental learning and an emphasis on egalitarian dialogue, in which anyone is welcome to contribute to the dialogue, regardless of their social or academic status. The two principles work together to enrich the bonds of solidarity within the extended Roma family and beyond; in turn, this solidarity helps participants transfer what they learn about health in the school to other spaces, reaching the broader Roma community. Future research is needed in order to complement the data obtained. Particularly, it would be interesting to contrast the information obtained with available health records of the population of La Milagrosa as well as with fieldwork conducted with primary health professionals of the area. Evidence that educative participation can effectively improve health conditions among the Roma people is a positive contribution to the goal to reduce the health inequalities from which they suffer. Thus, in addition to simply helping people gain more education, these other educational interventions can help them to overcome health inequalities. Furthermore, our findings emphasize the need for researchers to include the voices and cultural strengths of the Roma and other vulnerable cultural and ethnic groups when designing educational interventions that aim to improve their health conditions. Finally, policies intended to reverse the health inequalities facing the Roma people must be grounded in empirical findings, such as those provided by the Roma community in La Milagrosa in Albacete; in this community, people have transformed their health-related behaviors, regardless of their educational levels, through the implementation of evidence-based SAs. Future works addressed at the development of health policies for Roma will benefit from the findings of this research. --- Conflict of Interest The author declares no conflict of interest.

On average, the Roma in Europe can expect to die 10 years earlier than the rest of the population, given the health conditions they experience. EU-funded research has informed on successful actions (SA) that when implemented among the Roma provide them new forms of educational participation which have a direct impact on improving their health status, regardless of their educational level. The findings from this research, unanimously endorsed by the European Parliament, have been included in several European Union recommendations and resolutions as part of the EU strategy on Roma inclusion. To analyze these SA, as well as the conditions that promote them and their impact on reducing health inequalities, communicative fieldwork has been conducted with Roma people from a deprived neighbourhood in the South of Spain, who are participating in the previously identified SA. The analysis reveals that these SA enable Roma people to reinforce and enrich specific strategies like improving family cohesion and strengthening their identity, which allow them to improve their overall health. These findings may inform public policies to improve the health condition of the Roma and other vulnerable groups, one goal of the Europe 2020 strategy for a healthier Europe.

Introduction For men who have sex with men and transgender women , disclosure of same-sex sexual practices can be a critical step toward engaging healthcare services to appropriately prevent and treat HIV and other sexually transmitted infections . Disclosure of a concealable and stigmatized sexual identity may lead to improved healthcare outcomes by allowing MSM and TGW to express personally relevant information, garner social support, and influence societal views . Disclosure to family members can lead to acceptance, support, and identity formation that can lead to positive psychological adjustments and, in turn, may promote adherence to safer sex practices and medication, including pre-exposure prophylaxis or antiretroviral therapy for HIV . Disclosure to healthcare providers is necessary to receive anatomically appropriate screening for STIs and to inform the frequency of such screening. While same-sex sexual practices have become culturally more acceptable in many Western countries-with associated positive gains in human rights for MSM and TGW-the same has not been observed in many parts of the developing world, including sub-Saharan Africa, where there can be dire consequences for being identified as homosexual, including stigmatization, assault, imprisonment, and even capital punishment . Therefore, disclosure of same-sex sexual practices often occurs at substantial risk to the person making the disclosure . Exposure to these negative consequences of disclosure may create a hostile and stressful social environment, thereby worsening mental and physical health . In sub-Saharan Africa, two of the most commonly reported manifestations of stigma related to same-sex sexual practices are enacted or experienced stigma, which refers to behavioral expressions of stigma such as physical violence, blackmail, and denial of healthcare services and perceived or anticipated stigma, which refers to a fear or expectation that enacted stigma may occur, such as fear of seeking health care . In 2014, legislation passed by the Nigerian government recommended harsh punishments for MSM and TGW and persons perceived to be promoting homosexuality. Following the passage of this law, MSM and TGW reported increased fear and avoidance of healthcare services . In 2017, Nigerian police raided an event organized to promote HIV testing in Lagos and arrested men who were accused of engaging in same-sex sexual practices, a crime punishable by up to 14 years in prison . Prior research has shown that blackmail and beatings due to same-sex sexual practices are commonly experienced by Nigerian MSM and TGW and are linked to increased risk of genitourinary diseases . This could be partially explained by stigma leading to higher risk sexual behaviors, such as increased condomless sex and decreased engagement with healthcare systems . We have previously proposed a model whereby disclosure of same-sex sexual practices can lead to increased hierarchical classes of stigma, then suicidal ideation, then condomless sex with casual sex partners, and finally increased risk of HIV and other STIs . Understanding disclosure patterns of same-sex sexual practices among MSM and TGW communities in sub-Saharan Africa is therefore critical because these communities are marginalized, stigmatized, and have an exceptionally high burden of HIV and other STIs. Nigeria is the most populated country in sub-Saharan Africa, and its residents account for almost one out of every ten people living with HIV around the world and about 10% of new diagnoses annually ). In some communities of Nigerian MSM and TGW, HIV prevalence as high as 44-66% has been reported . The disproportionate burden of HIV among MSM and TGW is driven partially by physiologic factors such as the increased transmissibility of HIV via condomless anal as compared to vaginal sex and partially by social and cultural factors such as the suboptimal quality and coverage of preventive healthcare services available to MSM and TGW . A structural approach to managing the stigma associated with same-sex sexual practices, including any stigma resulting from disclosure thereof, is necessary to optimize quality of care . In Swaziland, a structural stigma-reduction intervention package has been implemented that encourages disclosure as a means of increasing uptake of evidence-based and rights-affirming care for MSM and TGW that includes couples' counseling, provision of condoms and condom-compatible lubricants, anal cancer screening, and laboratory-based STI diagnosis . While the Nigerian healthcare system is stratified into tiers operated at the local, state, and federal levels with inconsistent harmonization , there is increasing recognition at some levels of the importance of engaging MSM and TGW to curb the spread of HIV . Support in Nigeria is shifting toward nongovernmental organizations that deliver health care in "safe spaces" like TRUST/RV368 to address specific needs of MSM and TGW . In these analyses, we separated disclosure out from previously explored stigma classes to identify the proportion of participants who disclosed their same-sex sexual practices and evaluate whether disclosure could be an avenue for intervention to promote engagement with MSM-and TGWfriendly clinics. We also further explored the relationships between disclosure to a family member and/or HCP with condom use, stigma, and discrimination. --- Method --- --- Measures Upon enrollment, data on participant demographics and sexual behaviors were collected by a trained interviewer using a standardized questionnaire. A study physician performed a medical examination and recorded each participant's medical history. Participants underwent counseling and testing for HIV and other STIs. HIV status was determined using a parallel algorithm of rapid tests with Determine ® and Uni-Gold ® kits and a third tie-breaker with HIV-1/2 Stat-Pak , if needed. All testing was performed according to package inserts. Enrollment evaluations were scheduled across two study visits approximately two weeks apart. Participants were categorized based on their responses to two questions: "Have you disclosed to any member of your family that you have sex with other men or that you are attracted to other men?" and "Have you disclosed to any healthcare worker that you have sex with other men or that you are attracted to other men?" Age, gender identity, sexual orientation, education, occupation, and marital status were each assessed by self-report. Participants were asked to describe the frequency of condom use during insertive anal and receptive anal sex using the categories "always," "almost always," "half the time," "almost never," and "never." They were also asked whether they had ever experienced a variety of potential indicators of stigma and discrimination, such as fear of accessing healthcare services, avoidance of health care, and refusal of health care due to their MSM and TGW status. These indicators reflect types of perceived and experienced stigma related to same-sex sexual practices that have been previously observed in Western and Southern Africa . Data from clinical evaluations and participant interviews were collected on paper case report forms, imported into an electronic database using TeleForm , and verified for accuracy by a dedicated staff member. --- Statistical Analyses Separate analyses were conducted to compare groups reporting disclosure of same-sex sexual practices to a family member and/or a HCP. Demographic characteristics, condom use, and experiences of stigma were compared between these groups using Pearson's chi-squared test or Fisher's exact test for categorical variables and t test for continuous variables. In separate models for disclosure to a family member and to a HCP, Poisson regression with robust error variance was used to estimate the relative risk of disclosure and 95% confidence intervals associated with pre-specified factors of interest such as age, gender identity, sexual orientation, education, marital status, city, and enrollment year . Multivariable models included all pre-specified factors of interest. A two-sided type I error less than 5% was considered statistically significant. To help evaluate separate effects of disclosure to either a family member or HCP as well as the impact of overlapping disclosure to both a family member and HCP, comparisons were repeated across all four possible combinations of disclosure: disclosure to neither a family member nor HCP, disclosure to a family member only, disclosure to a HCP only, or disclosure to both a family member and HCP. A single multinomial logistic regression model was used to generate adjusted RRs and 95% CIs for factors associated with each of these categories as compared to the nondisclosing reference group. For all analyses, missing data were carried backward from the scheduled 3-month followup study visit, if available, or were categorized as unknown. Analyses were performed using Stata 14.2 . --- Results --- Demographics and Disclosure of Same-Sex Sexual Practices A total of 2591 participants were enrolled between March 2013 and May 2019, including 34 who did not answer questions about disclosure of same-sex sexual practices and were excluded from these analyses. Of the remaining 2557 participants included in these analyses, 1885 were from Abuja and 672 from Lagos. The median age of participants was 23 years with interquartile range 21-27 years. Same-sex sexual practice disclosure among participants was generally uncommon, with only 925 reporting some sort of disclosure, including 192 who had disclosed only to a family member, 541 only to a HCP, and 192 to both a family member and HCP. Demographic and other characteristics of the study population, stratified by disclosure of same-sex sexual practices to a family member and/or HCP, are shown in Table 1 and Supplemental Table 1. --- Factors Associated with Disclosure of Same-Sex Sexual Practices In unadjusted analyses, disclosure of same-sex sexual practices to a family member was more common among transgender women, more highly educated participants, participants living with HIV or with unknown HIV status, participants in Lagos, and during specific years of enrollment . Participants who self-identified as bisexual and were married or living with a woman were less likely to disclose to family. After adjusting for all evaluated factors, higher education level, recruitment at the Lagos clinical care site, and living with HIV were each independently associated with an increased likelihood of disclosure to family. As compared to participants with gay/homosexual/other sexual orientation, those reporting a bisexual orientation were less likely to disclose same-sex sexual practices to a family member. In unadjusted analyses, disclosure of same-sex sexual practices to a HCP was more common among older participants, those with other/unknown gender identity, more highly educated participants, participants living with HIV or with unknown HIV status, participants in Lagos, and during specific years of enrollment . After adjusting for other factors, older age, other/unknown gender identity, higher education level, recruitment at the Lagos clinical care site, and living with HIV were each independently associated with an increased likelihood of disclosure to family. Participants who were married or living with a woman were less likely to disclose to a HCP in both the unadjusted and adjusted models. The multinomial logistic regression model revealed similar trends . Older age was associated with greater likelihood of membership in one of the two groups that disclosed to a HCP. Factors associated with increased likelihood of all three combinations of disclosure as compared to nondisclosure included higher than senior secondary education , recruitment at the Lagos site , and living with HIV . --- Sexual Behaviors and Condom Use Participants who disclosed their same-sex sexual practices to a family member had a slightly younger age of coitarche with male partners, which occurred at a mean age of 16.5 years , compared to 17.3 years among those who had not disclosed to family . No significant difference was observed in age of coitarche with female partners between participants who had and had not disclosed their same-sex sexual practices to family . Compared to participants who had not disclosed their same-sex sexual practices to a family member, participants who had disclosed were more likely to report always or almost always using condoms during receptive but not insertive anal sex . Participants who disclosed their same-sex sexual practices to a HCP had a slightly older age of coitarche with male partners, which occurred at a mean age of 17.6 years compared to 17.0 years among those who had not disclosed to a HCP . No significant difference was observed in age of coitarche with female partners between participants who had and had not disclosed their same-sex sexual practices to a HCP . Compared to participants who had not disclosed their same-sex sexual practices to a HCP, participants who had disclosed were more likely to report always or almost always using condoms during both insertive and receptive anal sex . Similar inferences were drawn from comparisons across groups defined by all four possible patterns of disclosure . Participants who disclosed same-sex sexual practices to a family member only were more likely to report condom use during receptive but not insertive anal sex. Participants who disclosed same-sex sexual practices to a HCP, with or without concurrent disclosure to a family member, were more likely to report condom use during both insertive and receptive anal sex as compared to participants who did not report any disclosure. --- Experiences of Stigma Compared to those who had not disclosed to a family member, participants who had disclosed their same-sex sexual practices to a family member were more likely to report fear of seeking health services and avoidance of healthcare services but were no more likely to be denied healthcare services because they were MSM and TGW. Furthermore, those who had disclosed same-sex sexual practices to a family member were more likely to have felt afraid to walk around , experienced blackmail , and experienced assault because they were MSM and TGW, when compared to those who had not made a similar disclosure . Similarly, compared to those who had not disclosed to a HCP, participants who had disclosed their same-sex sexual practices to a HCP were more likely to report fear of seeking health services and avoidance of healthcare services . Unlike disclosure to a family member, disclosure to a HCP was associated with a significantly increased risk of being denied healthcare services Fig. 2 Indicators of perceived and experienced stigma due to samesex sexual practices, stratified by disclosure of same-sex sexual practices to a family and b healthcare providers. Bar height represents the percentage of study participants who reported each indicator of stigma upon enrollment. Pearson's chi-squared test was used to compare the proportion of participants reporting each stigma indicator, stratified by disclosure status. Statistically significant differences between groups are shown in bold vs. .5%, χ 2 = 25.46, p < .01). Participants who had disclosed same-sex sexual practices to a HCP were also more likely to have experienced both blackmail and assault because they were MSM and TGW, when compared to those who had not made a similar disclosure . Similar inferences were drawn from comparisons across groups defined by all four possible patterns of disclosure . Disclosure to a family member was not associated with denial of health care, but denial of health care was more common in both groups that had disclosed to a HCP. A similar pattern was observed with sexual violence. Disclosure to a HCP was not associated with fear of walking around, but more participants in the groups who disclosed to a family member reported this fear. Other stigma indicators were generally increased among participants who had made any disclosure as compared to the nondisclosure group. --- Discussion Disclosure of same-sex sexual practices to a family member or HCP were each relatively uncommon among Nigerian MSM and TGW in this study. In each case, disclosure was associated with a larger proportion of participants reporting condom use, consistent with results from prior studies suggesting that disclosure enhances social support, improves quality of health care, and improves treatment outcomes in MSM and TGW populations . However, these apparent beneficial effects were countered by strong relationships between disclosure and stigma, which may have discouraged utilization of healthcare services. Disclosure of same-sex sexual practices among Nigerian MSM and TGW in this study was generally less common than has been observed in settings with less criminalization and stigmatization of homosexuality. The disclosure rates of about 15% and 29% to a family member and a HCP, respectively, among MSM and TGW in this study were substantially lower than the rates of about 75% and 56%, respectively, which have been reported from South Africa, where MSMfriendly legislation exists . In contrast, research from countries with laws that criminalize same-sex sexual practices among men generally shows much lower rates of disclosure, including disclosure to a family member by 46% of MSM and TGW in Cameroon, to HCP by 31% of MSM and TGW in Swaziland, and disclosure to a HCP by only 17% of MSM and TGW participants in a large crosssectional study in Malawi, Namibia, and Botswana . In addition to criminalization, cultural and societal factors may also contribute to disclosure practices and should be considered when designing affirmative and targeted population-based HIV and STI prevention intervention. In China, a country without anti-homosexuality laws but with extralegal consequences for same-sex sexual practices, the overall disclosure rate to HCP was found to be only 16% . In the African context, cultural stigmatization has created barriers to MSM-and TGW-focused healthcare services that are compounded by criminalizing legislation that creates fear of arrest and brutality . Stigma was associated with disclosure of same-sex sexual practices to both family and HCP in this study. Participants in this study reported both perceived and experienced stigma because they were MSM and TGW, including specific stigmarelated barriers to healthcare engagement. Notably, although denial of health care was generally uncommon, MSM and TGW who disclosed their same-sex sexual practices to HCP were substantially more likely to be denied health care. The stigma and discrimination experienced by Nigerian MSM and TGW are in some ways legitimized by Nigeria's antihomosexuality legislation, which has resulted in increased stigma that acts as a barrier to disclosure of sexual practices . Similar to the observation in this study, same-sex practice disclosure among MSM and TGW in Swaziland was found to be associated with stigma and fear of seeking healthcare services . In Gambia, studies show several associations between disclosure of same-sex sexual practices by MSM and human rights abuses such as physical assaults, arbitrary arrests, and threats of decapitation. While important, decriminalization of homosexuality does not eliminate social and cultural sources of stigma. As the first and only African nation to legalize unions between same-sex couples, South Africa is one of sub-Saharan Africa's most legislatively accepting countries . However, MSM and TGW still experience significant stigma, discrimination, and poor reception from HCP in South Africa . A study that evaluated the impact of decriminalizing laws in South Africa, Botswana, Malawi, and Namibia found that human rights abuses such as police harassment and denial of housing persisted despite legislative changes but blackmail and fear were diminished . Legal and social change must occur in tandem in order to optimally promote community acceptance, decrease stigma, and encourage disclosure of same-sex sexual practices to family and HCP. This could possibly improve healthcare engagement and access to appropriate preventive health services. In this study, certain key factors were found to be associated with disclosure of same-sex sexual practices among MSM and TGW. Increased educational level was strongly associated with disclosure to both family and HCP, underscoring the general importance of education possibly facilitating disclosure. The reasons for disclosure being more common among participants living in Lagos are uncertain. Both Lagos and Abuja are large urban centers with reports of discrimination and human rights abuses affecting MSM and TGW . Understanding the social, behavioral, and environmental factors that positively influenced disclosure in Lagos could inform HIV prevention approaches. Same-sex practice disclosure to either a family member or HCP was associated with more condom use. A prior study from China similarly found that nondisclosure was associated with riskier same-sex sexual practices among gay and bisexual men . Despite the potential positive relationship between disclosure and condom uptake, same-sex sexual practice disclosure was also associated with increased perceived and experienced stigma. Prior data suggest that knowledge and attitudes of HCP are key barriers to disclosure of same-sex sexual practices . Training HCP to demonstrate awareness through a patient management algorithm that incorporates proactive inquiry about sexual practices could reduce fear of disclosure and improve care and treatment outcomes. Tailoring interventions to the needs of individuals and communities who are less likely to disclose their same-sex sexual practices may improve results. National recognition and support of nongovernmental organizations that provide community-engaged care to MSM and TGW in safe spaces may help to optimize the gains from disclosure while minimizing stigma. There were a number of strengths and limitations of these analyses. The respondent-driven sampling technique enabled the enrollment of a highly marginalized and hard-to-reach population of MSM and TGW. The use of standardized questionnaires enabled the collection of detailed information regarding study population demographics, same-sex practice disclosure, and associated stigma. However, potentially sensitive and stigmatizing data collected via self-report may suffer from a number of biases, including response and recall biases. The population included in these analyses was recruited in two large urban centers in Nigeria and may not be representative of populations in other areas. Recruitment at trusted community health centers specially staffed to create safe spaces for MSM and TGW may have biased the study population toward a greater likelihood of disclosure, although this bias was mitigated by the respondent-driven sampling methodology that reached participants who had not previously accessed these facilities. Only individuals who could provide informed consent in English or Hausa were eligible for this study, potentially biasing the study population toward inclusion of individuals with specific educational, geographic, or ethnic backgrounds associated with these languages. There was some overlap of disclosure groups, potentially confounding interpretation of the individual effects of disclosure to either a family member or a HCP; however, almost all inferences were robust to sensitivity analyses that re-categorized participants into four mutually exclusive groups representing all possible patterns of disclosure to a family member and/or a HCP. The primary dichotomous analyses were retained to enable easier visualization and interpretation of the relationships that were confirmed by the sensitivity analyses. While we have provided theoretical pathways through which the factors evaluated may be associated with disclosure of same-sex sexual practices, the crosssectional nature of these analyses limits our ability to infer any causal relationships. These analyses did not distinguish between voluntary and involuntary disclosure of same-sex sexual practices. This study revealed a complicated relationship between disclosure and same-sex sexual practices by Nigerian MSM and TGW. Cultural and behavioral implications of disclosure must be considered in designing interventions to engage MSM and TGW in healthcare and HIV prevention services. This study corroborates prior work indicating that interventions to encourage disclosure must be accompanied by larger campaigns to educate communities, create safe spaces to access health care, decrease stigmatization, and promote policies of acceptance. Improved disclosure practices within safe spaces may enhance engagement of MSM and TGW in healthcare and HIV prevention services. --- Compliance with Ethical Standards Conflict of interest The authors declare that they have no conflict of interest. Disclaimer The views expressed are those of the authors and should not be construed to represent the positions of the U.S. Army or the Department of Defense or the Department of Health and Human Services. Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed Consent All participants provided written informed consent prior to enrollment. The study was reviewed and approved by the Ministry of Defence Health Research Ethics Committee, Abuja, Nigeria; Walter Reed Army Institute of Research Institutional Review Board , Silver Spring, Maryland, USA; and the University of Maryland IRB, Baltimore, Maryland, USA. The investigators have adhered to the policies for protection of human subjects as prescribed in AR-70. 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/. --- Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Disclosure of same-sex sexual practices by men who have sex with men (MSM) and transgender women (TGW) may facilitate appropriate healthcare engagement, including risk assessment for HIV and other sexually transmitted infections (STIs), and negotiation of condom use with partners. However, disclosure may also generate stigma. In these cross-sectional analyses, MSM and TGW were categorized based on self-report of disclosure to family members and healthcare providers (HCP) at enrollment into the TRUST/RV368 study of comprehensive HIV and STI care programs in Abuja and Lagos, Nigeria. Multivariable Poisson regression models with robust error variance were used to estimate relative risk of disclosure with 95% confidence intervals. Pearson's chi-squared test was used to compare condom use and stigma indicators by disclosure status. Of 2557 participants who answered baseline questions about disclosure, 384 (15.0%) had ever disclosed to a family member and 733 (28.7%) to HCP, including 192 (7.5%) who disclosed to both. Higher education, prevalent HIV infections, and residence in Lagos were each associated with increased likelihood of disclosure to family and HCP. Older participants were more likely to disclose to HCP but not family. Participants who made a disclosure to family or HCP were more likely to report condom use during anal sex as well as perceived and experienced stigma that included healthcare avoidance, blackmail, assault, and sexual violence as compared to participants who had not disclosed. Improved disclosure practices within safe spaces may enhance engagement of MSM and TGW in healthcare and HIV prevention services.

Introduction Family caregivers are considered as the backbone of the long-term care and support system [1,2]. Family or informal caregivers offer voluntary and unpaid physical, practical, and emotional care and/or support to a person with a disability in the home environment [3,4]. --- The German Situation Recent figures from the he German federal statistical office show that, in Germany, 76% of all care to those with care needs is provided by family caregivers in their home environment [5]. Three to five Million German family caregivers provide unpaid care to 4-5 Million persons in need [6]. Most of these family caregivers are female, between 40-85 years old, and still in employment when younger than 65 years [7]. Family caregiving can be extremely stressful, especially in long-term care situations, e.g., when caring for a family member suffering from Alzheimer's disease or a stroke [8,9]. Caregivers often experience a reduced quality of life, social isolation, loneliness, stress, fatigue, exhaustion, frustration, as well as a sense of helplessness [10][11][12]. A lack of choice of being a caregiver, lack of support, as well as financial strains and unfamiliar bureaucratic workload may also provoke caregiver burden [6,13]. Ideally, family members need to be prepared and educated before they take up their new role as informal caregiver [14,15]. In Germany, the official conditions for caregiver support are embedded in a law that supports the hospital discharge management and the 'Präventionsgesetz-PrävG' [16]. Currently, different stakeholders provide different forms of caregiver support [6,7]. For example, health insurances provide online support, e.g., a "caregiving coach" [17], or group support, e.g., "caregiving courses" [18]. Communal services offer personalized "caregiving-and environmental counselling" . However, these services are uncoordinated, unsatisfactorily utilized [19,20], or planned too late during the caregiving trajectory [21]. Especially in case of a sudden incident, such as a stroke, the organization of tailored family caregiver support should start already within the hospital or rehabilitation setting, i.e., before the patient is discharged. Acknowledgement of the prospective role of family members as part of the patient care trajectory together with well-timed, personalized support might also prevent the so-called 'revolving door' phenomenon [20,22]. --- Context of the Study The Rhine-Maas-Clinicum is a large hospital located in Wuerselen, North-Rhine Westphalia, Germany, with a strong hierarchical organizational structure. The Rhine-Maas-Clinicum has a large geriatric department, with 64 acute geriatric beds and 28 rehabilitation beds spread over three wards. In 2016, 1550 patients, on average 82.7 years old, were admitted. Treatment, care, and medical support is provided by around 70 professionals, with a strictly patient-centered approach. Staff members at the Rhine-Maas-Clinicum treat family caregiver issues in an unstructured, haphazard manner. The needs of family members, as prospective caregivers at home, are insufficiently acknowledged, although care professionals recognize the value of family members in the care trajectory of their patients. Family caregivers have been shown to play an unnamed but critical role in the geriatric health care system, forming a so-called 'shadow workforce' [23]. They are an important source of information for geriatric care professionals, being the ones able to supply key pieces of medical information about their elderly care receiver in a timely manner [23]. However, they often experience strain, especially co-resident caregivers, and run the risk of getting depressed [24]. In a large meta-analysis, it was shown that receiving support, formal or informal, was related to better caregiver health [25]. Supporting them will secure the care to geriatric patients in the home environment after hospital discharge [26]. Recent recognition of the importance of the family as prospective caregivers for geriatric patients prompted the Rhine-Maas-Clinicum to join an academic research team from Maastricht University, The Netherlands, in setting up a Participatory Health Research project to prepare the implementation of a new support programme for family caregivers, called Vade Mecum [27,28]. Vade Mecum aims to sustain health and wellbeing and prevent social exclusion and health inequality of geriatric patient-caregiver dyads. At the core of the concept is a new job profile: a hospital-based care employee with a special assignment to guide, educate and support family members of geriatric patients in becoming an informal caregiver [27]. A first step in designing a better system of geriatric care in the Rhine-Maas-Clinicum is to identify the status-quo and needs of geriatric care professionals to support them. --- Aim of the Study Before full implementation of the new family caregiver support concept in the Rhine-Maas-Clinicum, a pre-implementation step is necessary to gain insight in the status-quo activities and needs of the multidisciplinary team regarding family caregiver support. Because of the haphazard way of engaging with family caregivers of geriatric patients by individual care professionals, little is known about how and what is needed to best deliver that support in a more organized way and to inform the hospital management on the best way of implementing the new holistic support concept. In this article, we report on both the process and the results of a PHR study on actual state and needs of the MDT. At the core of the PHR approach is the delegation of decision-making power regarding the research process from academic researchers to a co-research team consisting of members from the MDT. This co-research team will be formed at the start of the study and is responsible for choosing the research design, participant profile, data collection methods, data analysis and reporting. The actual research question will be articulated by this co-research team during the PHR process. --- Materials and Methods --- Research Strategy For this pre-implementation step, a participatory research strategy is a preferred choice. PHR has a longstanding tradition in public health research [29], and is viewed with increasing interest by researchers, decision-makers, and practitioners worldwide as a strategy to overcome the gap between academic researchers and professionals working in practice [30]. One of the principles of PHR is a power shift in the conduct of research from academic or 'expert' researchers to those stakeholders in the field that possess insightful knowledge and experience [31,32]. In addition to the generation of new knowledge, this approach also aims to empower those involved in the research process [33,34]. Co-researchers learn methods of scientific knowledge production and use the process to understand their own reality, whereas academic researchers understand the real needs of a practiced-based phenomenon [35]. Engaging members of the MDT as co-creator of new knowledge creates ownership of the outcomes [36]. --- Formation of the Research Team, Roles, Research Activities Three categories of researchers were involved in this study: academic researchers , a critical friend and co-researchers. The AR was responsible for: facilitating and stimulating the PHR research process as "external impulse provider", contributing with scientific, technical and managerial knowledge, drafting the research plan and monitoring the progress, and guaranteeing a constant communication flow. The AR started the study by contacting all members of the MDT and explaining the participatory strategy. During the research process, a 'critical friend' supported the AR. A CF is a trusted professional who is able to ask proactive questions and give collegial feedback [37]. The CF in our study was serving as 'listening ear', stimulated reflection and continuous professional development of the AR [38,39]. In our study, members of the MDT were participating as co-researcher with researchers from outside the setting. Members of the MDT are the ones with experiential knowledge in the setting, have ideas about relevant research questions, the acceptability and appropriateness of the research methods and the interpretation of the findings. In principle, all members from all professional groups of the MDT working at the geriatric department of the Rhine-Maas-Clinicum, were eligible as member of the co-research team provided that they have contact with family caregivers. Research experience was not necessary. The head of the department, a medical doctor, served as gatekeeper. The co-research team together with the AR and CF set the research agenda, formulate the research question, decide on which qualitative research methods to be used to collect the data, formulate selection criteria and recruit key informants in their peer group, collect the data, perform the data analyses, report the outcomes, and carry out all other necessary research activities to answer their research question [40,41]. The proceeding of this process is reported in the Results section Part 1: PHR process, and in Table 1. --- Documenting and Reflection A research management plan was developed to keep track of all research aspects and choices, and document who did what on what level of participation . The AR outlined six phases beforehand: 'orientation', 'setting-up', 'planning', 'data collection', 'analysis', and 'reporting', according to the key principles developed for utilizing participatory methodologies in the co-creation and evaluation of public health interventions [42] The AR took lead in the preparation, organization, planning and communication of the project. As the iterative research process progresses and the roles change, contributions, activities and expected or planned outcomes will be inserted in the management plan to document the emerging study. In each phase, the participation grades of the co-researchers will be determined, and graded by using Cornwall's participation typology, which ranges from 1 'co-option', 2 'compliance', 3 'consultation', 4 'cooperation', 5 'co-learning', to 6 'collective action' [43]. To evaluate the experiences with the PHR approach and critically reflect on the benefits and limitations of this research methodology within the setting, a brainstorming session with mind mapping will be carried out in the final meeting, when the co-researchers are still together as a group. Regular meetings between the AR and the supervisor were organized to discuss important research issues and the progress of the subsequent steps, but also ways of ensuring rigor of the study [44,45]. The CF assisted the AR on keeping a reflective mindset, while the AR kept record of reflective talks with the CF. All research choices, activities and reporting in this manuscript were guided by the Standards for Reporting Qualitative Research [46]. --- Ethical Considerations This participatory project aimed to assess the status-quo and needs of hospital personnel regarding support of family caregivers of geriatric patients, in preparation of their new to develop working practice. Participation as co-researcher or participant in one of the in-depth study parts was on a voluntary basis, no incentives were offered. Verbal informed consent was obtained from all individual study participants in the professional peer groups before data collection. The head of the department enabled participation during working hours and facilitated the research activities . Ethical approval was given by the hospital management, approval from an ethics committee was not necessary. Mutual respect, equality and inclusion, democratic participation, active learning, making a difference, and personal integrity were the ethical principles accompanying the entire PHR project [47]. Additionally, the following 'ethics in practice' principles were respected: beneficence, respect for autonomy, justice, and non-maleficence [48]. * Some fragments of the project up to the phase of data collection were published earlier [27]; ¶ : CR = co-researchers; § : From [47]. --- Results First, in part 1, the PHR process will be presented per phase, accompanied by the management plan . Next, both outcomes will be presented , and finally reflections on the research process will be addressed . The project started in February 2017 and ended in March 2018. The PHR life cycle started with the literature exploration, explorative talks with professionals , and a 2-day field visit by the AR. A list with topics was constructed to guide the first meeting with the prospective team members within the geriatric department. --- Setting-Up The second phase started with an impulse lecture to raise interest to conduct a PHR study and, by doing this, to improve the caregiver support within the geriatric department. Members of the various professions in the geriatric department were invited as co-researchers. Careful attention was paid to ascertain representation of all levels of the hierarchical system within the hospital. Two medical doctors, two case managers, two social workers, two therapists , and four nurses took part as 'internal' stakeholders. The AR with the CF completed the research team. In the first workshop, the team was asked to reflect on their daily challenges in the interaction with family caregivers. System-related problems , as well as caregiver-related problems were brought up. Additionally, communal support offers were perceived as fragmented, uncoordinated, and lacking a holistic approach. Caregivers were occasionally involved as partners in the hospital care process, but mainly for purposes of patient care. Caregiver burdens or needs were not assessed systematically at any time during the patient's hospital stay. Co-researchers also reflected on their own expectations and goals of the new support concept. They expected to allocate more time to support caregivers 'tuned to caregiver's individual needs', scheduled in the normal working hours, to benefit from task simplification, to expand their own professional competence and get appreciation. Objectives were that family caregiver-patient dyads were given longer consultation times, as well as individualized and improved support within the hospital. At last, guided by triggering questions from the AR, the main research question was formulated by the team: How does the multidisciplinary research team of the Rhine-Maas Clinic currently supports family caregivers, and what is needed for offering comprehensive support? --- Planning In the third phase, the co-researchers first decided that a qualitative data collection method is appropriate to answer the research question. Next, they decided, based on the organizational structure within the department, on the groups of participants for the study. Six participant groups were indicated: medical doctors, nurses, physio-/occupational therapists, speech therapists, case management, and social services. The co-researchers chose data collection methods after an impulse lecture from the AR, who preselected and presented five possible methods: interviews, focus groups, storytelling, structured interview matrix and community mapping. For each method, advantages and disadvantages were discussed first, taking the existing experience within the geriatric setting into account. Interviews and focus groups seemed to be the best option. Next, co-researchers were equipped to conduct the data collection within their own peer groups, with an instruction memo and a question catalogue prepared beforehand to guide and harmonize data collection within the different professions [27]. The AR monitored the data collection process. --- Data Collection Study participants for the interviews and focus groups were recruited by co-research team members among their peers. In total, 31 individuals took part in the participatory data collection phase. Structured focus group discussions were organized by the medical doctors among the medical doctors , nurses discussed with nurses and physio-/occupational therapists with their own peers . Semi-structured interviews were held by the speech therapists , case managers , and social workers . Each group organized their own meetings. Field notes were produced per group, and then distributed within the entire research team to be analyzed. --- Analyses A 'data-analyzing team' was formed, consisting of two medical doctors, two nurses, one occupational therapist, one physiotherapist and one social worker. The analyzing process was facilitated by the AR and CF. The 'data-analyzing-team' met three times for approximately 70 min per meeting and used a stepwise thematic analysis strategy [49,50] to analyze both the three focus group datasets as well as the three interview datasets. All members of the 'data-analyzing team' were familiarized with all six sets of field notes. Themes and subthemes were agreed upon per dataset first, after which the outcomes from the six data sets were merged, resulting in one final comprehensive multiperspective understanding of both the actual state and the needs for offering caregiver support within their setting. --- Part 2: Outcomes --- Outcome 1: The Actual State of Caregiver Support From the thematic analysis, three themes emerged: professionals' perception of the family caregivers, actual state of professional caregiver support activities, and weak spots or deficits in current caregiver support. Ad : Professionals perceive family caregivers as an important interlinking person between the patient and the multidisciplinary team, but also as at risk of becoming overburdened by infrastructural issues, e.g., by changes in contact persons, or not being informed about rehabilitation and support offers. The social services said that: "the role of caregivers is underestimated in our department". The physio-and occupational therapists observed that "only 'demanding' caregivers are supported and 'quiet' relatives often fall through the grid". During their daily work, some professionals experience caregivers in general as "under-supported" and/or "insufficiently informed" [nurses; physio-/occupational therapy], while others even experience them as an "encumbrance" or "added task in their daily work" [case managers, medical doctors]. Ad : The intensity of contact, extent and moment of involvement varies greatly between the different professions. The existing caregiver support system is perceived as deficient, unstructured and uncoordinated: "Currently, caregiver support is very chaotic" [physio-/occupational therapy]; "Caregiver counselling is rather inadequate, especially on medical and nursing topics, since there are no fixed office hours nor are individual appointments offered" [social services]. However, all professionals provide some support on an individual basis: "One medical doctor offers caregiver consultation talks from 3:00 pm to 4:00 pm" [medical doctor]. Interdisciplinary consultations can be scheduled on request by the caregiver. The knowledge concerning caregiver-specific offers varies greatly among the professionals, and there is hardly any knowledge regarding the quality or acceptance of these offers. Within the hospital, two offers are known: individual dementia caregiver counselling, which is provided by the clinic pastor; and a peer group for dementia patients' caregivers. Some professionals recommend external support offers to caregivers, e.g., health insurances, case managers or practical nursing skill courses provided by the communal services. Ad : For the hospital management, the importance of family caregivers in the patient care trajectory is not recognizable. Caregiver support and counselling activities are not financed. Additionally, the number of available staff is so low that comprehensive caregiver support is difficult to organize. Caregiver support is insufficiently outlined in hospital job descriptions. Professionals feel frustrated and undervalued: "There is little appreciation from the side of the management about our work with the relatives" [medical doctor]. No caregiver support concept or guideline is implemented in the department or elsewhere in the hospital. Professionals lack important resources, such as time, staff, infrastructure , and internal support services. "We have no reserved time slots for caregiver instruction" [physio-/occupational therapy]; "Our office is occupied by three people plus serves as the meeting room for caregiver consultation" [social services]; "Demonstration or training materials on site are either outdated or non-existent." [physio-/occupational therapy]). Professionals recognize caveats in their caregiver-specific knowledge, skills and competences, blaming their vocational or academic education: "We received no orientation on how to conduct a conversation with caregivers in our vocational education. We had to learn this in the job" [medical doctor]; "We possess some experiences, however our knowledge is still expandable" [case management]. Professionals felt that insufficient attention is given to capacity building of staff concerning caregiver issues within the hospital. Although they can participate in external trainings, the conditions vary greatly: "You have to acquire the knowledge yourself" [physio-/occupational therapy]. --- Outcome 2: Needs of the Geriatric Team After careful discussion, four themes emerged from the analyses: the need for an inclusive caregiver support concept, conceptual building blocks, required resources, and enabling conditions. Ad : A holistic, needs-oriented , personal and professional caregiver support concept is required. "We must find a system that involves relatives in a more structured way" [social services]. The values of caregiver counselling must be taught to the entire multidisciplinary team. Counselling should involve: an orientation on caregiver support offers within the department and beyond, e.g., communal support, personalized information transfer, e.g., financial, medical and rehabilitative issues, and practical support, e.g., instructions for care and mobilization: "A mixture of good information about processes, illnesses and supply options via internet, flyers, group offers and individual advice is needed." [social services]; "Relatives should know the duties of each professional group" [physio-/occupational therapy]. Moreover, psychosocial support should be family-centered and better counselling facilities are needed. Maybe a self-help peer group, caregiver information evenings and group counselling sessions would be supportive: "Some caregivers may be willing to meet other caregivers" [nurses]; "Group counselling is in some occasions helpful" [social services]. Ad : To offer to caregivers relevant and valuable information, a number of options were mentioned: a caregiver "welcome folder" containing general information ; profession-specific flyers presenting the different professions, processes and contact persons; information boards within the department 'advertising' caregiver support; and a 'caregiver link' on the hospital's homepage leading to general and specific information and to support offers within the region. Caregiver support needs to be tailored and long-term: "They need most support in the first days during the acute phase and a sound orientation on what to do when the patient will be discharged" [nurses]. Creating a new job in the geriatric department, providing professional caregiver counselling, might be a solution. "A permanent contact person for caregivers is needed" [medical doctors]. This position requires professional experience, strong communication skills, system knowledge as well as special personality traits, e.g., empathy, openness: "The ability to communicate and resolve conflicts, a degree of diplomacy, the ability to keep calm and pay attention to facts as well as active listening is crucial" [social services]. Furthermore, "outreach counselling is needed" [social services]. A family caregiver assessment tool, self-administered by the caregivers before the first counselling, was proposed: "A separate sheet to be completed by caregivers would be useful in many cases" [social services]. The creation of an interprofessional network, which connects all stakeholders providing internal and external support, was suggested: "Networking and participation in external events and organizations, e.g., Dementia Network Aachen, Alzheimer's Society, other geriatric institutions is needed" [social services]; "We must think and act as a team" [physio-/occupational therapy]; "We need to create access to supportive and further-serving institutions" [social services]. Ad : the need for resources were expressed: time, infrastructure , equipment for family education, and more staff. "The social services should be on the spot all day every day, as it has a large role in attending caregivers" [medical doctor]; "Fixed counselling hours should be offered, e.g., caregiver friendly time slots, and the contact persons' name and accessibly should be provided" [social services]; "Adequate demonstration and training tools are required for caregiver and patient training" [physio-/occupational therapy]; "Supervision and case review sessions should be organized" [social services]. Needs-oriented internal or external capacity building activities were advised: "Caregiver support" [social services]; "Communication strategies" [case management]; "De-escalation training or how to best deal with overburdened relatives" [physio-/occupational therapy]; "The nursing school curriculum should extend to cover caregiver issues" [medical doctors]. Capacity building should be financially compensated. Ad : the need for enabling conditions within the hospital was expressed. "Due to the new discharge management law, we have a take care now also for the patient beyond the inpatient stay. Here, more activities are needed across all professional groups within the department, e.g., social services, psychologists" [medical doctors]. To do this, an increased appreciation and attention for caregiver support in the geriatric department by the hospital management is required. "We need increased attention to the special concerns of the geriatric department" [social services]. The management must encourage a caregiver-friendly climate within the hospital and should trigger the development of a comprehensive caregiver concept. "There is a gap in the system that needs to be fixed as soon as possible" [medical doctors]. --- Part 3: Critical Reflection In a separate part of the final meeting, the entire research team reflected on their experiences with the PHR approach. --- The PHR Approach Overall, the co-researchers, CF and AR experienced this PHR approach as 'inspiring and motivating'. The AR experienced the sharing of power as mutual learning process that enabled a true engagement of the co-researchers. However, compared to classic top-down research, conducting PHR for the first time in a hierarchic setting requires a great deal of time and investment in communication and project management. The CF recognized the changing team dynamics and was positively surprised how her comments were stimulating and motivating the team during the different phases. Beside the positive effects, she noticed that it was quite time consuming for the entire team. Initially, co-researchers perceived this project as "black box", especially regarding its outcome and practicability within this hospital. However, the preconditions for the project were set as the head of the department was in favor of the PHR approach. Finally, co-researchers concluded that "participatory research works. It magnifies the knowledge and has changed us" [medical doctors]. --- Participation Grades As can be seen in Figure 1, the participation grades of the co-researchers increased during the study from 'co-option' in the orientation phase to 'collective action' in the 'data collection' phase and onwards. 'Co-learning' was already achieved during the Kick-off workshop . Additionally, it was possible to achieve a higher grade of participation in the course of one activity. For instance, during workshop 1, the grade increased from 'compliance' during the impulse lecture to 'co-learning' when discussing suitable data collection methods. However, to enable co-researchers to participate actively in the next step, it appeared to be obligatory that the AR took the lead by providing scientific knowledge. to participate actively in the next step, it appeared to be obligatory that the AR took the lead by providing scientific knowledge. --- Figure 1. Co-researchers' participation during PHR project, presented per phase in the project and per level of participation [43]. Legend: 1-literature review; 2-field visit, 3-impulse presentation; 4-impulse lecture PHR; 5-exploring expectations and goals; 6-identifying the problem and formulating research question; 7-forming the research team; 8-impulse lecture methodology, 9exploring methodological advantages and disadvantages, 10-determining suitable data collection methods; 11-impulse lecture data collection; 12-determining plan of action and question catalogue; 13-collecting data; 14-impulse session data analyses and decision on 'analyzing team'; 15interactive data analyses; 16-interactive data analyses; 17-critical reflection of experiences with PHR; 18-dissemination activities. Although the AR was responsible for the impulse lectures , the content of the presentations was, to an extent, influenced by the co-researchers. These lectures were necessary as this was the first PHR project in this department. Based on co-researchers´ needs, the AR decided what to include in the impulse lectures. For example, in the second lecture, various qualitative data collection methods were presented together with their benefits and limitations. Based on their input, co-researchers, CF and AR explored the advantages and disadvantages within their setting and decided on a suitable data collection method. --- Benefits and Limitations The productive and constructive atmosphere, the transparency of the research process, the connectedness between the different professions, as well as the reflective nature and being appreciated by other co-researchers, were perceived as positive . On the other hand, PHR was experienced as very time consuming; the constant communication was demanding; the process was sometimes unpredictable which provoked uncertainty; the 'black-box' nature was challenging. With respect to the subject of the status-quo and needs assessment, the co-researchers appreciated the nearness of the research to the reality of their practical work, and the new perspectives they gained through the process. The structured way the AR organized the process was characterized as valuable. The AR experienced the process as motivating and empowering, e.g., understanding the problem from the inside. Giving up control and sharing the responsibility in the various study parts meant that the process required patience and tolerance from all team members, especially in the data collection phase. Additionally, flexibility of the entire team was essential. Co-researchers' participation during PHR project, presented per phase in the project and per level of participation [43]. Legend: 1-literature review; 2-field visit, 3-impulse presentation; 4-impulse lecture PHR; 5-exploring expectations and goals; 6-identifying the problem and formulating research question; 7-forming the research team; 8-impulse lecture methodology, 9-exploring methodological advantages and disadvantages, 10-determining suitable data collection methods; 11-impulse lecture data collection; 12-determining plan of action and question catalogue; 13-collecting data; 14-impulse session data analyses and decision on 'analyzing team'; 15-interactive data analyses; 16-interactive data analyses; 17-critical reflection of experiences with PHR; 18-dissemination activities. Although the AR was responsible for the impulse lectures , the content of the presentations was, to an extent, influenced by the co-researchers. These lectures were necessary as this was the first PHR project in this department. Based on co-researchers' needs, the AR decided what to include in the impulse lectures. For example, in the second lecture, various qualitative data collection methods were presented together with their benefits and limitations. Based on their input, co-researchers, CF and AR explored the advantages and disadvantages within their setting and decided on a suitable data collection method. --- Benefits and Limitations The productive and constructive atmosphere, the transparency of the research process, the connectedness between the different professions, as well as the reflective nature and being appreciated by other co-researchers, were perceived as positive . On the other hand, PHR was experienced as very time consuming; the constant communication was demanding; the process was sometimes unpredictable which provoked uncertainty; the 'black-box' nature was challenging. With respect to the subject of the status-quo and needs assessment, the co-researchers appreciated the nearness of the research to the reality of their practical work, and the new perspectives they gained through the process. The structured way the AR organized the process was characterized as valuable. The AR experienced the process as motivating and empowering, e.g., understanding the problem from the inside. Giving up control and sharing the responsibility in the various study parts meant that the process required patience and tolerance from all team members, especially in the data collection phase. Additionally, flexibility of the entire team was essential. --- Figure 2. Perceived benefits and limitations of the PHR approach within the geriatric hospital setting: outcomes of the brainstorm session in the final meeting. --- Reporting The project was concluded with 'reporting', as a last and ongoing phase. All findings regarding the status quo and needs of the MDT with respect to holistic family caregiver support were included in a final document, which was presented in the hospital by the co-researchers together with the CF and the AR. Based on the outcomes, a 'caregiver orientation' folder was designed for the geriatric department to better inform caregivers regarding the different roles of the team and the caregiver offers within the department. Several co-researchers already presented outcomes and lessons learnt from their PHR experiences in practice on various occasions , whereas CF and AR presented the outcomes and methodological considerations in scientific conferences. --- Discussion Our PHR study resulted in two outcomes, namely a comprehensive understanding of the actual state of the caregiver support within the geriatric acute and rehabilitation care, and the needs for offering comprehensive caregiver support within the department. The geriatric department needs a comprehensive, needs-oriented, personal, and professional caregiver support concept, which includes activities on both individual and system level. Personalized caregiver counselling is required to prepare caregivers for their new role and guarantee the quality of care in the home environment. Practical, accessible, and timely information provision for family caregivers must be developed. Our professionals proposed to assess caregiver's needs before the first counselling with a still to be developed assessment tool. Network building, communication and information materials are needed to enable organizational change. The hospital management should invest in a comprehensive caregiver support that goes beyond the usual economically focused case management routine. Resources , and infrastructure , are required. Investing in health system development on a system level, e.g., offering capacity building activities and investing in intra-, inter-as well as extra-organizational networks, may be necessary. On an organizational level, caregivers must be acknowledged as part of the support team of geriatric patients. Our outcomes are in line with conclusions from a qualitative literature review on hospital discharge planning practices for frail older people stating that there is room for organizational --- Reporting The project was concluded with 'reporting', as a last and ongoing phase. All findings regarding the status quo and needs of the MDT with respect to holistic family caregiver support were included in a final document, which was presented in the hospital by the co-researchers together with the CF and the AR. Based on the outcomes, a 'caregiver orientation' folder was designed for the geriatric department to better inform caregivers regarding the different roles of the team and the caregiver offers within the department. Several co-researchers already presented outcomes and lessons learnt from their PHR experiences in practice on various occasions , whereas CF and AR presented the outcomes and methodological considerations in scientific conferences. --- Discussion Our PHR study resulted in two outcomes, namely a comprehensive understanding of the actual state of the caregiver support within the geriatric acute and rehabilitation care, and the needs for offering comprehensive caregiver support within the department. The geriatric department needs a comprehensive, needs-oriented, personal, and professional caregiver support concept, which includes activities on both individual and system level. Personalized caregiver counselling is required to prepare caregivers for their new role and guarantee the quality of care in the home environment. Practical, accessible, and timely information provision for family caregivers must be developed. Our professionals proposed to assess caregiver's needs before the first counselling with a still to be developed assessment tool. Network building, communication and information materials are needed to enable organizational change. The hospital management should invest in a comprehensive caregiver support that goes beyond the usual economically focused case management routine. Resources , and infrastructure , are required. Investing in health system development on a system level, e.g., offering capacity building activities and investing in intra-, inter-as well as extra-organizational networks, may be necessary. On an organizational level, caregivers must be acknowledged as part of the support team of geriatric patients. Our outcomes are in line with conclusions from a qualitative literature review on hospital discharge planning practices for frail older people stating that there is room for organizational improvement by including the family, improving communication between professional health care employees and the family, better interdisciplinary communication and ongoing support after discharge. Specific contextual needs of family caregivers should be assessed, identification of resources that family caregivers need, and active and early involvement in the discharge process [51]. Our study also fitted recommendations to achieve a whole system change with respect to comprehensive, person-centered assessment and support for family caregivers towards the end of life, both on what needs to be in place within an organization to provide this support, e.g., a protocol for assessing caregivers and responding to the assessment, as well as on what is required to implement and sustain this approach in practice, e.g., a process for training practitioners and available time/workload capacity [52]. In a recent qualitative study from a Norwegian oncology setting, it was shown that health care workers in regional cancer care were also in need of organizational improvement, including systematic involvement of family caregivers in care pathways, and education of professional care workers in caregiver support [53]. The multidisciplinary team at the Rhein-Maas-Clinicum intends to use the findings of their study in the next step: building the new job of 'Geriatric Family Companion' [27]. However, to embed the new professional in the support system inside as well as outside the hospital, input of experienced family caregivers is needed and cooperation of professional support providers working in the area. --- Methodological Considerations A major strength of our PHR study lies in the uniqueness of the applied strategy, as a truly partnered project. Participatory research provides a specific form of understanding that emerges from the synthesis of expertise brought to the partnership and the necessary negotiation of powers within that partnership [54]. This guarantees that the outcomes from the study are relevant, acceptable, and feasible . Apart from the background research that was done by the AR, our co-research team was enabled to complete the whole research project. All function groups of employees within the German hospital setting that were in contact with family caregivers, were represented and contributed to the outcomes. Throughout the entire project, co-researchers were given the power to influence and shape the study itself. Through the active participation of professionals with different backgrounds and professional roles, the research question was explored and answered from different perspectives and its objectives were developed from the real-life experiences [56]. Participation cannot be imposed on research, but should rather be seen as an emergent process, developing from the interaction between the problem, the environment, and the aims, capacities and skills of those involved [35]. Participation depends upon the establishment of an environment of trust within the group [57]. Unfortunately, we did not formally study our group dynamics, e.g., with an instrument [58], so we cannot be sure that everybody's viewpoints and expertise were valued in the process. One of the other limitations of the PHR approach is that is has a gradualist tendency potentially leading to tunnel vision, caused by selection mechanisms [28]: the researchers have no way of knowing if what they discovered is actually true. However, since PHR is practice research, misinterpretations or gaps can still be uncovered and mended afterwards, in the phase of the programme's implementation [28]. Nevertheless, based on the success of our project and the positive reflections by the co-research team, we feel confident that the outcome of the study reflects the actual status-quo and needs of a family caregiver support initiative. The co-researchers felt empowered to continue with their newly acquired skills and knowledge. Our PHR project was terminated in January 2018, but some effects are still noticeable since the final meeting: co-researchers continue to participate in dissemination activities, e.g., authoring articles or presenting outcomes on conferences. In addition, a professional caregiver-working group was founded, and a 'caregiver orientation' folder for the geriatric department was designed. Our PHR study was the first one applied in this hospital setting. As the German health system has been functioning historically through hierarchical decision-making [59], some challenges were expected. Interestingly, looking back on the PHR approach, our co-researchers and participants in the interviews and focus groups described PHR to be "transparent, expedient, interconnecting, 'eye-opening', appreciative, and perspective-providing" . However, the PHR process was resource demanding. It took a vast amount of time and required extensive communication on different levels. In addition, due to the high workloads and understaffing within the department, the composition of the co-researchers in the research team varied over the project life cycle. The different professions in the MDT were represented in each meeting, but not always by the same persons. This situation led to time constraints, communication gaps and "perceived chaos" in the 'Setting-up', 'Planning' and 'Data collection' phases. With respect to the higher management level, a lack of interest and willpower to be involved was perceived. Despite several attempts, co-researchers were unable to schedule a meeting to present their outcomes to the management team. Even with the frustration caused by this, the higher hospital management was kept informed about the progress and invited to the meetings. PHR seemed to be an effective strategy within this hierarchically organized hospital. However, conducting PHR requires methodological experience, elaborated communication skills, a flexible time investment, and an open-minded atmosphere within the setting. --- Conclusions Through a participatory research methodology, a comprehensive understanding of the actual state of family caregiver support within the geriatric acute and rehabilitation care was gained. The geriatric department requires a comprehensive, needs-oriented, personal, and professional caregiver support concept, which includes activities on both individual and system level. To prepare family caregivers for their new role and guarantee the quality of care in the home environment, personalized caregiver counselling is necessary. Practical, accessible, and timely information provision for family caregivers must be established. To enable organizational change, network building, communication and information materials are needed, but above all, caregivers must be acknowledged as part of the support team of geriatric patients. On a system level, investing in health system development may be necessary. Participatory health research appears to be a feasible and effective strategy within a hierarchically organized hospital. However, conducting a participatory methodology requires methodological experience, elaborated communication skills, a flexible time investment, and an open-minded atmosphere within the setting. ---

Because of societal changes, family caregivers are becoming vital in long-term care provision for geriatric patients after discharge from hospital. Hospital-based geriatric care teams need tools to prepare and support family caregivers for their future caregiving role in the home environment. To explore the actual state and needs for implementing a suitable family caregiver support concept in a large geriatric hospital in Germany, a Participatory Health Research methodology was chosen. An academic investigator, assisted by a critical friend, facilitated all research steps. Geriatric care professionals joined as co-researchers and performed qualitative data collection using semi-structured interviews and focus group discussions. The entire co-research team took part in the thematic analyses. The existing family caregiver support was perceived as uncoordinated and incomplete, and a lack of knowledge about support programmes in the community was apparent. The needs regarding a comprehensive family caregiver support concept that acts on both individual caregiver as well as on system level, but also connects the two levels, were formulated. High grades of participation of hospital-based co-researchers could be achieved. A critical reflection on the research strategy revealed that the participatory methodology, although time-consuming, was perceived as a useful strategy within the hierarchically organized hospital.

H UIMIN JIN : To begin with, I should say thanks for your accepting my interview, Professor David Morley. As one of the leading figures of British cultural studies, represented, perhaps we could say, chiefly by the so-called Birmingham School, you should be, I believe, in a good position to clarify some issues about which I am quite curious, concerning the historical development of this school. More importantly, I am rather keen to know from your own perspective your special contribution to British cultural studies, either in a Hall-centred mode or in a broader sense. Well, my first question is about popular culture. Recently, scholars with Manchester Metropolitan University have shown a tendency to narrow cultural studies down to popular culture studies. By doing so, they foreground and intensify popular culture as a primary arena of cultural studies. This is related to the British tradition of cultural studies, from Williams's definition of culture as a way of life, via the Centre for Contemporary Cultural Studies subculture researches, to the course on Popular Culture run by the Open University from 1982 to 1987. You were invited to contribute a course unit on 'Interpreting Television' for this historically significant programme. As to the ever-expanding field of popular culture studies as an academic enterprise, my concern is given to such questions as why academics should take it seriously in the British context, and if it has something specific to do with a politics of culture, say, to the turn to Gramsci, or to, anthropologically, our everyday life, in which it has an enabling, transformative force. David Morley : I see no reason at all why cultural studies should now concentrate exclusively on popular culture. Indeed, in my view, it is crucial that cultural studies also attend to middle-brow and high-brow culture, as these forms and their changing characteristics can only be understood in relation to each other. At the point of cultural studies' inception, however, the position was very different. At that moment, it was widely assumed that popular culture was not worthy of theorists' attentions. So the initial focus on popular culture was a polemical move, in a particular context, designed to show that popular culture was a field that was indeed well worth studying. However, nowadays, that point is generally accepted so it doesn't need to be stressed so much. So, to narrow cultural studies' focus down to only popular culture would be a very regressive move in the contemporary situation. You're right to point to the important shift towards perspectives derived from anthropology as a way of understanding culture as a way of life. But that argument doesn't have any specific or particular relation to the British case. It would apply, I think, to the study of culture in all societies. What is most specific to the development to cultural studies in Britain, in particular, relates to the moment of the late 1970s and early 1980s, the moment of the 'Popular Culture' course which ran at the Open University from 1982 to 1987. That was a moment when several key factors combined, which led to the stress on the politics of popular culture. Let me explain. At that point in the early 1980s, Mrs Thatcher had just been elected and Stuart Hall and others argued that her victory was founded on a form of populism . Hall rightly argued that Mrs Thatcher's victory could only be understood ^and effectively opposed ^on the ground of culture. His point was that she had successfully articulated forms of 'authoritarian populism' which genuinely resonated with the feelings and aspirations of the dissatisfied white working class of Britain. The argument was that the core explanation for the overwhelming working-class electoral vote for her was their support for her vision of the return to what she called 'traditional Victorian values' . They supported her rejection of the post-war liberal progressivism of the 1960s and 1970s. Hall's argument was that, if that was so, then the understanding of the cultural dimension of politics, especially in the form of popular culture, becomes critical. At the same time, a variety of dissident groups in Britain ^women, gays, black people, disabled people, etc. ^were also insisting there were cultural dimensions to politics. They insisted that their oppression often took cultural forms, for instance, negative stereotypes of them in jokes or in popular entertainment, and thus, they argued, the transformation of these forms of popular culture was an important dimension of political struggle. That was all part of the 'turn to Gramsci' and that was the reason for the corresponding stress on the cultural dimension of politics, in which context the analysis of the dynamics of popular culture was then seen as the key task. The further point here is that you have to understand that this was an argument which was made in the face of a rather hard-line, leftist, Marxism, which said that culture was not really important at all, and attributed everything to economic factors. However, that analysis completely misses out the extent to which Mrs Thatcher's political success worked precisely by transforming British culture ^and by transforming what were understood to be British values. She had to transform those things, in order to gain the popular support to go on and win the later political and economic fights that she got into ^e.g. with the miners when they went on strike for a whole year. So her first political victory was at the level of culture: in shifting values and reasserting the sense of individualism, individual achievement, individual ambition and individual responsibility in all spheres of life. Having won that cultural battle, she was then in a position to win the political and economic battles. That was, in a sense, a cultural revolution in Britain: she achieved a revolution in the culture of everyday life and shifted everybody's assumptions a long way to the Right. The long-run significance of this can perhaps be seen now, after the end of the period of the 'New Labour' government. Despite the claim that New Labour had a different approach to politics, in a very significant sense what the New Labour government did was simply to live out its life under the hegemony of the terms of reference established by Mrs Thatcher. I think you could very reasonably say ^as many people have done ^that Tony Blair represented Mrs Thatcher's 'true heir', because the cultural battles that she won ^about individualized responsibility, and about the free market ^established the hegemony of a set of ideas which shifted the whole political terrain, and which New Labour did little to question, for fear of losing 'middle ground' electoral support. Politics in the UK today is still being fought out inside the terms of reference established by the cultural victories which Mrs Thatcher achieved. Now the new Coalition government is pursuing that Thatcherite agenda even more radically than New Labour ever did, in its plans to totally dismantle the remnants of the welfare state set up as part of the post-Second World War settlement in the late 1940s and thus establish a new 'common sense' in which the market is taken as the hegemonic form for the modelling of all areas of social life. HJ: There is a popular view that your reputation in the area of cultural studies was established by your television audience researches, which are mainly shown in your two books The 'Nationwide' Audience andFamily Television . From your empirical work, what theoretically, in brief, have you developed or reinforced? I should say sorry if you would think I am over-concerned with the theory per se, which may be contradictory to the tradition of British cultural studies. DM: It's quite true that my own reputation within cultural studies was largely founded on the audience studies that I did in the 1970s and 1980s, published as The 'Nationwide'Audience and Family Television. However, I was never, in fact, particularly interested in either television or audiences themselves, as objects of study. What I was interested in was the question of cultural power. The choice to make an empirical study of television audiences was simply a way to 'operationalize' a study of the extent of ideological or cultural hegemony, as manifested in the forms of media consumption. As for your further question, concerning what, theoretically, my empirical work has achieved, my answer would be that it has produced a far better model of the operation of media power than would have been made possible by continuing to make abstract speculations about media effects of the sort offered by such scholars as Adorno and Horkheimer in the early period, by McLuhan in the 1960s, or by Baudrillard in the postmodern version. In my view, despite the evident theoretical sophistication of much of this work, it nonetheless still operates with what is, in the end, a very simplistic motion of media effects. Even the latest version of 'new media' theory is still flawed by the simplistic assumption, carried over from Adorno and Horkheimer, that the media have automatic and unavoidable 'effects'on people. Apart from anything else, that is inadequate because it is, ultimately, a technologically determinist argument. I have developed this line of critique of the severe limitations of so-called 'new media theory' most fully in the section of my last book titled 'Rhetorics of the Technological Sublime' in relation to the problems posed by the foreshortened a-historical perspective of these theories and their oversimpli¢cation of the crude binary divide which they draw between the 'old' world of the analogue media, with their audiences of supposedly passive 'couch potatoes', and the marvels of the new digital era, in which it is assumed that 'we' are all more signi¢cantly active. Clearly, my own inclination here would be to go back again to Williams' argument in Television, Technology andCultural Form , which very carefully shows that technologies do not have effects built into them, because it's always a question of how particular technologies come to be institutionalized in particular ways. There are many social and cultural forces which act to determine that. As my own work has shown, along with that of people such as the late Roger Silverstone, new technologies don't simply have effects on people, just as media don't have direct effects on their audiences. The question is how particular people, in particular contexts, perceive the relevance of specific media technologies for their lives, and how they then choose to use those technologies ^or ignore them, or indeed 'bend' them in some way, to a purpose for which they were not intended. In either case, the effects are not directly produced. If my work has been about one thing it has been about how to develop a more satisfactory model of the power of the media. I'm not in the business of denying the power of the media, or denying that particular technologies allow new things to happen. My point is to better understand how that power operates, in conjunction with the fact that people do make choices, and do make their own interpretations of material provided to them by the media, whether we are talking about the mass media of broadcasting or the micromobile media of today's world. One of the interesting points you raise in your questions is the status of 'Theory', and its position within the tradition of British cultural studies. I think it is a very interesting question and it does concern me deeply. It's very common in contemporary debates, especially among people whose background is in sociology or philosophy, rather than in cultural studies, for people to make a claim for high status by presenting themselves as 'Theorists' . However, that is an approach which, in terms of the role of theory in the tradition of cultural studies defined by Stuart Hall, can only appear as extremely na|« ve. Hall's version of cultural studies is not at all resistant to theory as such. But that tradition, within which I'm situated, is one in which we use theory in order to theorize some particular, empirically specific conjuncture or situation, not for the purpose of generalized and abstracted speculation. If you look at Hall's 'Marx's Notes on Method: A ''Reading'' of the ''1857 Introduction ''' , the point is made very clearly there. For Hall, theory is not valued in itself, but in terms of its usefulness in theorizing particular conjunctures. The problematic issue about philosophical-style 'Theory in the Abstract' concerns the high intellectual price which must be paid for any moment of abstraction. Of course, theoretical abstraction is a powerful, and often necessary, intellectual tool, as it allows you to condense what would otherwise be a mountain of potentially confusing data, in order to see the underlying patterns. But, just like a power-saw, it is also a potentially dangerous tool, which must be used very carefully if it is not to do more harm than good. Of course, any abstraction or categorization is, of its nature, reductive. The question is always whether this particular formal abstraction is worthwhile, in a specific case. Each time you abstract, you have to ask yourself whether the benefit you will get, in so far as the abstraction contributes to your ability to make some ordered analysis of cultural patterns, is sufficient to make up for the fact that, in making that abstraction, you'll be losing track of some part of the particularity of the situation you are trying to analyse. Coming, as I have said, from a tradition which emphasizes specificity in empirical investigations, I am unsympathetic to what the French philosopher Michel Serres calls lazy forms of 'one size fits all' theory. As Serres says, it is lamentably easy to use that kind of theory to reduce all phenomena to one ultimate set of 'truths' . However, as he observes, a single theoretical 'pass key' will never suffice to open all doors ^rather, as he insists, each time you want to 'unlock' a specific problem, you must forge the specific theoretical key which will be adequate to the problem in hand . I'm interested in 'grounded' forms of theory. So, for example, if we take the case of TV as a medium, I would not want to say that television is a thing which has some particular set of facets ^or that it is in the 'essence' of the medium, philosophically understood, that it should always have these effects. Rather than that very deterministic and mechanical mode of analysis, I'm interested in understanding how things work in particular circumstances or contexts, when media technologies are actually used by different people. Of course, we must find a way to see the deep-seated patterns in our data, and it is no good disavowing all generalizations and ending up in a post-structuralist morass of just telling a million individual stories of infinite difference. But we must, nonetheless, use those generalizations very cautiously and be attentive to the extent to which, every time we make a generalization, we pay a price, in the loss of contextual specificity. The question is one of always keeping one's eyes on the 'price ticket' ^making sure that what you lose in the use of that theoretical abstraction is outweighed by the gains in analytical power that it enables. HJ: When you were developing your theory of active audiences, did you have in mind German reception-theory such as the works of the Konstanz School, say, Hans Herbert Jauss and Wolfgang Iser? Or, you might have had some other theoretical resources available? It looks as if British cultural studies concurs basically with reception-theory in assuming that audiences are never passive receivers of media messages, or that, furthermore, the reception or consumption of television messages can be elevated as a part of the whole television production. DM: Your question puts me in mind of the story that Janice Radway tells in the introduction to the British edition of her book Reading the Romance , which was published here some years after its initial publication in America. There, Radway explains that, when she subsequently discovered British cultural studies work she was astounded to discover how closely it 'fitted' with the perspective that she, independently, had been developing in her ethnographic work on women readers of romance fiction. The situation was rather parallel, I think, in terms of my own relation to the German theories to which you refer. In fact, I only came across the work of Jauss and Iser at some point in the early 1980s, well after my work on the 'Nationwide'Audience had been done. What was interesting, of course, was that, rather like Radway, I then had the experience of coming across a body of fascinating theoretical work which was very close to what I had been doing, even though it wasn't directly influential in the construction of my own perspective. This is perhaps a general point about the fact that, if something is a good idea, it is probably going to be worked out by more than one person, in more than one place, often simultaneously! So Jauss and Iser and I end up not exactly in the same place but on relatively close theoretical paths, although we got there from different routes, and by means of different theoretical resources. In terms of my own theoretical resources, the key ones were sociolinguistics and cultural anthropology. I believe I was the first person to use the now well-worn term 'ethnography' in relation to media audiences, in a paper that I wrote in 1972: 'Reconceptualising the Media Audience: Towards an Ethnography of Audiences' . Where did I get that from? From the work of people like Dell Hymes and Clifford Geertz in cultural anthropology. As for the resources used to analyse the interpretations which people made of television programmes , the key elements came from debates in the sociology of education about the role of linguistic codes in determining children's educational success or failure ^which was a theory which seemed to me to also be applicable to how the availability of different cultural resources might play some part in determining patterns of decoding TV. At that time, in Britain in the early 1970s, the key debate was between Basil Bernstein and his critic Harold Rosen. The issue at stake was to do with the role of class structure in the determination of language abilities and language uses. Bernstein had a rather deterministic approach, which was very important in identifying differences between working-class language and middle-class language, much of which has very close parallels with the work of Pierre Bourdieu in France about the social distribution of cultural capital and cultural competences. Conversely, while Rosen appreciated the power of those analyses, he was very concerned that they were being conducted in too essentialist a way. Rosen was aware that while class is very likely to have some effect on the forms of cultural capital to which someone in a given social position has access, this is not an automatic process. He was also concerned with the way in which people 'inhabit' their membership of any particular collectivity. In a certain sense, already in the 1970s Rosen was making exactly the kind of argument which someone like Judith Butler makes now, 30 years later. Just as Butler says that we are not all living in what she calls the 'prison house of gender' , Rosen is making the same point about class: we are not just 'prisoners' of class ^because there are different ways of inhabiting middle-class or working-class identities. The main problem with Bernstein's model is that it just avoided this issue entirely in its sociological determinism, assuming that the effects of class position were more or less automatic. This is a criticism which some of us would say also applies to the work of Bourdieu ^which has much in common with that of Bernstein. Thus, one could make an analogy and say that if Bernstein was the English version of Bourdieu, his critic, Rosen, was the equivalent of Michel de Certeau, one of Bourdieu's most important critics in France. So, to go back to where I started in my answer to this question, my main theoretical resources came from cultural anthropology, in terms of providing a model for a 'close reading' of particular actions as understood in context. They also came from sociolinguistics ^and especially from Volos› inov's insistence on the 'multi-accentuality' of the sign, his insistence that there is no totally shared language system in a given society, and that one has to be sensitive to the conflicts that are fought out in and through language and culture. HJ: If you hold that 'individual readings will be framed by shared cultural formations and practices preexistent to the individual' , though in a complex and subtle way, how could you demarcate your theory of active, therefore resistant audiences from the Frankfurt School's 'nonresistant' conception of the masses that are passively injected with a prevailing ideology? When the masses begin to decode the televisional texts, they might have been 'pre-mass-ed' by some other factors, social or ideological, which you admit as pre-existent, other than by the modern communications, television included ^by ideology or, precisely, being part of ideological totality or ideological apparatus: a role, whatever it may be, which you seem to hesitate to recognize. To be clear, my inference here is that so long as you assert that individual interpretations are socially, at least partially, determined, you would have already accepted that the cultural industry as a social or ideological force or apparatus does play a role in making the masses. This argument may be supported by the text's role to which you still attach importance: interpretations are not 'arbitrary' but 'subject to constraints contained within the text itself' . DM: Essentially, in The 'Nationwide'Audience I was trying to offer what I think is a better way of understanding media power than is offered by the Frankfurt School's simplistic approach to audiences ^as passive 'dupes' ŵhich was predominant at that time. I'm quite happy to accept that Adorno and Horkheimer make sophisticated general arguments about the way in which the culture industry shapes consciousness. In criticizing their model of the automatic 'effects' of the media on the masses, I'm not trying to replace it by a theory which says that all audiences are completely 'active' and are making just any interpretation they want out of the media materials they come across. In that respect, my perspective is quite different from that of scholars such as John Fiske in what has come to be called the 'active audiences' tradition. Contrary to them, I'm not trying to deny issues of media power. Moreover, despite my criticisms of Bourdieu and Bernstein's overly deterministic model of class, I'm also very interested in the way in which the ability of a person to re-interpret or re-use the media to which they have access is, in fact, to some extent determined by their social position. That's because their social position will limit their access to particular types of cultural codes and cultural capital. However, I'm not advocating some model of 'free-floating' individuals who are just able to do whatever they like with what the media offers to them. Here I would also distinguish my own position from that of people like Ulrich Beck: when Beck says that in today's 'individualized' world, class is no more than a 'zombie' category of little use in analysing social life, I think he is quite wrong. Class continues to exercise profound effects on people's lives ^especially in an era where, in the UK at least, rates of inter-generational mobility are declining: so that class position at birth is now an even better predictor of an individual's probable life course than it was 30 years ago. Of course, class or any other mode of social categorization can be used in an overly deterministic ^and thus, 'zombie-like' ^way, but that is a matter of how exactly you use the categories in your analysis, not whether or not the categories are intrinsically useful or not. I'm also quite prepared to accept that the socio-cultural 'identities' which I use in The 'Nationwide'Audience analysis as explanations of how this or that person interprets a particular programme are themselves socially and culturally constructed. In English, we have this expression, which is a kind of rhetorical figure, or joke, in which one says: 'Which came first, the chicken or the egg?' In one way it is just a silly joke. But, at another level, it expresses a rather profound philosophical dilemma. Of course, in a sense, it doesn't matter which came first. What you have to understand is what their relationship is. The same goes for the question of the role of the culture industry in the construction of consciousness and in the provision of the materials out of which people construct their identities, through the decoding of TV programmes, among other things. The subsequent issue is how you can then look at those identities themselves as partly explaining why this person, in this particular position, will tend to be affected by this particular media, or how they will tend to interpret some particular media programme. One has to understand that as a dialectical process between determination and activity: to go back to the original, as Marx says, men and women make their history, but not in conditions of their own choice. In the media field, what we are looking at is the way in which social positions go some way to determining the cultural resources available to you, the cultural choices you can make ^out of which you construct your identity. Now that's not to say that you construct your identity on a 'tabula rasa' in a world of completely free choices ^that you could decide to just be anybody, or do anything at random, or interpret the media just any way you like ^you couldn't. We all have only a certain limited repertoire of possibilities available to us, and yet nonetheless, within that circumscribed repertoire there are still choices to be made, and those choices are significant. One can't explain the determination as total or automatic ^to take that view is just to wipe out the complexity of the dialectic which is at the heart of the process. HJ: Adapting from Frank Parkin, Stuart Hall argues that there may be three 'hypothetical' positions from which a media message is decoded by audiences: the dominant-hegemonic, the negotiated, and the oppositional, among which the latter two, compared with the ¢rst, that rarely happens in practice, are stressed for their frequency. It can be seen that there is a point at which your theory of the active audience encounters this argument: the resistance implicit in all interpretations of media texts. Here my concern is with the questions of what makes a reading active or resistant and why the cultural industry is not able to knead all the receivers simply into the masses. DM: I agree with the proposition, implicit in your question, that in a sense, all readings are active. To make any sense at all of a TV message you first have to interpret the dots and noise on the screen as representing images of people, or voices and so on, and you've got to read the rules of perspective, through which you can reconstruct the image of a three-dimensional world from a two-dimensional screen. Thus, activity is always present in the interpretation of media messages. As for when an interpretation is resistant, that is quite a difficult question. You could argue, for instance, that in The 'Nationwide'Audience, the reaction of the young black students who dismiss 'Nationwide' as irrelevant to them, and who refuse, in a sense, to make any decoding of it at all, is far more radical a move than that of the left-wing trade unionists in the study, who produce a classically oppositional reading of the 'Nationwide' text, but who do so from within a framework in which they recognize the programme as relevant to them. While accepting some of the political terms in play, they make a critical counter-argument to them, and that is a form of opposition, certainly. But you could argue that the reaction of those who dismiss the whole genre of that kind of politics as irrelevant to them is, in another sense, a much more radical form of negativity, and I don't think that Parkin's model is able to deal with that question. That model rather assumes that consciously politicized opposition is the most active, and most radical, and in that sense, the most important form of response to the media. I'm not sure that it is the case and that's why, as my work developed, I begin to 'operationalize' the model in a slightly different way, so as to take account of the ways in which all readings are active , but also to take account of the fact that the forms of activity are themselves various, in ways which Parkin's model is not really capable of dealing with. HJ: Social-contextually, why did the Frankfurt School choose a passive model of media consumers while British cultural studies chose an active one? To put it differently, what are the social conditions that underlie this shift of media studies, if there are any that work? In connection with this, textually, I doubt that mass culture and popular culture are referred by them to the same media texts and/or in the same way, even though they might overlap. I wonder if categorically the masses is more closely associated with the producers of the cultural industry while the popular more with consumers of the products. If it is true, can we argue that this shift reflects a social transformation from production society to consumer society, as Jean Baudrillard has distinguished them, and that your audience theory is rooted in such a consumer society in which people have more choices, in the face of a plethora of commodities? DM: I think I'd put it rather differently! I'm not sure that you can understand the Frankfurt School's model of media power without thinking quite specifically of the contexts of Weimar and Nazi Germany in the 1930s, when those theories were being developed. There you have the context of the emergence of powerful new media ^in that case, radio and cinema ând new forms of political rhetoric, operating in a relation to a population which, in its mobility from rural to urban areas, was becoming 'anomic' and thus more vulnerable to persuasion and propaganda. Those circumstances demand attention and sensitivity to questions of media power and manipulation. That is why Adorno and Horkheimer were interested in the particular questions that they focus on. British cultural studies certainly arises in very different circumstances, but I'm not sure that I would interpret that in the way that you seem to imply, in your question. As I understand your argument, you are suggesting that we face a move from a 'producer society' to a 'consumer society', that the concept of 'the masses' is a characteristic of what you call a producer society ^and that popular culture is then, conversely, associated with the consumer society. From that premise, if I understand you rightly, you see 'active audience theory' as being to do with the extent to which, in this thing called the 'consumer society', people have more choices. I think that's a problematic form of historical periodization and one which is characteristic of a certain type of sociological approach ^many social theorists love inventing new periodizations! For example, nowadays many are people talking about Deleuze and something called 'The control society', without a clear definition of what that is or when it started! Then people speak of 'The consumer society': but that has various definitions. You could talk about the consumer society of post-war Europe in the 1950s, or you could talk about today's emerging consumer society in China, which seems to me to operate a in very different kind of way. Those would, minimally, have to be understood as rather different sub-types of the overall concept: and once you recognize that, it follows that you cannot just assume that the same tendencies will apply in the same ways in these different contexts. Rather than thinking in terms of a total shift in historical periods, from producer to consumer society, as you suggest , my own perspective would be to say that in both situations, there always have been ^and still are ^both aspects of that question. That is to say, in what you would refer to as the 'producer society' of the era of mass culture, there were still choices to be made. People still did choose between this or that aspect of mass culture and, indeed, reacted to it, or interpreted it, in different ways ^so they were still active, even then. And conversely, even now, in this so-called consumer society, where people do get to make more choices, we mustn't confuse choice with empowerment in any simple sense, nor with a loss of producer power. I might have a larger number of channels from which to choose what to watch on my television set these days, but it's still a menu set by someone else, within which I have to choose. To put it another way, I'm more interested in the continuities and crossovers between these so-called different 'periods' or different 'types' of society than I am in just accepting such clear binary divisions unproblematically. In that respect I'm influenced by Derrida's insistence on deconstructing and destabilizing those kind of binaries ^and by his insistence on the need to pay attention to the ambivalences which cut across the kind of binary divisions of which sociologists, in particular, are rather too fond of for my liking! HJ: Did the postmodernism, for instance, of Fredric Jameson, Michel Foucault, Roland Barthes and Jean Baudrillard help to shape your television studies, and in the first place, of course, your audience studies? If so, in what ways did it work? You know, postmodern declarations such as the death of the author, or of subjectivity, were quite provocative, and they tend to emphasize the role of readers in the process of reception. DM: Let me take those authors in stages. Barthes was certainly influential for me, at an early stage, in setting up the kind of semiological model with which I was working. But I rapidly moved towards Volos› inov's critique of that rather rigid form of semiology. So Barthes was influential, but perhaps Umberto Eco more so, especially in his essay 'Does the Audience Have Bad Effects on Television? ' . But Volos› inov was the key resource, in producing an analysis that was attentive to the mobility of signs; to the conflict between signs; to what he calls the multi-accentuality of signs; and to the way in which different people will use the same sign or word or image for very different purposes. As for postmodernism, the story is rather different. The simple fact is that postmodernism came much later ^Baudrillard, Jameson and Foucault only came to have influence in cultural studies circles in the UK long after I produced the 'Nationwide' and Family Television studies. Again, as far as these periodizations go, just as I'm not much enamoured of the notion of mass society, or consumer society, or production society, or control society, likewise 'postmodernity' seems to me a rather weakly defined category. The notion that ours simply 'is' a postmodernism era, seems problematic to me, unless I know when it began, which in turn all depends on which definition of modernity or modernism you are using. If there is such a thing as a postmodern era, it's hard to imagine that it started on the same day and exists in the same form in Tokyo, Los Angeles or in some village in rural Uganda. Overall, I'm more interested in the notion of how differential temporalities often coexist in any one historical moment. In this respect I'm influenced by historians of the Annales School, such as Fernand Braudel, and his ideas of differential historical time or, at a simpler level, Raymond Williams' notion of the coexistence of 'emergent' and 'residual' dimensions in a culture, alongside whatever is 'dominant' at that particular time. In that context, one interesting issue, it seems to me, is to do with the different sequences in which particular theorists are read in different cultural contexts. What I have in mind here is how all of this might possibly seem to you, and to a Chinese audience, reading about these things in the 21st century, 25 years or so after I was principally involved in doing this kind of work on TV audiences, in a quite different context. Let me put that point another way. In a public debate at a conference, I was once asked what the difference was between British cultural studies and American cultural studies, and I had to come up with an instant answer! In that setting, my answer was that it was the difference between Foucault read through Gramsci, and Gramsci read through Foucault. The point is that, in Britain, people had been reading Gramsci in cultural studies, and then discovered Foucault, and began to rethink Gramsci a little bit, in terms of what difference Foucault's intervention made to that perspective. Conversely, in America, a lot of people in cultural studies came across Foucault first and only came to Gramsci subsequently, at the moment when British cultural studies 'imported' Gramsci into the USA. So what you get in much American cultural studies is a framework that is fundamentally Foucauldian, but which then works in a little bit of Gramsci. My point is that this difference of sequence, in terms of theoretical influences, has considerable significance. One has to pay attention to the modes of cultural transmission, and be sensitive to the complex routes through which particular theories come to dominance in different circumstances. One mustn't presume that which is, or was, useful or significant at one moment in a particular place will work in the same way or have the same significance in a different set of circumstances. Let me give you another example. I once did some work with the late Italian semiologist Mauro Wolf, who died a few years ago. Before his death, he was working on a book in which he was going to translate George Gerbner's works on 'cultivation analysis' into Italian, in order to develop a model of the long-term effects of the Italian media. What was interesting about this was the potential significance of Gerbner's work in Italy at that particular moment in the very early 1990s. As far as people in the British or American cultural studies tradition were concerned, by then, Gerbner's work seemed rather old fashioned ^and very much 'displaced' by the work in Italian semiotics of people like Eco. However, Wolf could see that if you started from semiological premises, as Italian media scholars naturally did given the strength of that tradition's development in Italy, then importing Gerbner into the debate at that point could prove very useful. I'm making an elementary point about the way in which the significance of any particular theory or theorist may vary immensely according to its context in different times and different places, and according to the very different sequences in which intellectual life develops in particular countries. As for the question of the 'death of the author', and the 'birth of the reader', again, rather like what I said in relation to your counter-position of 'producer society' and 'consumer society' , I don't think that you have to 'kill' the author in order to give birth to the reader! I think we still have to think about both authors and readers. I quite agree with the polemical thrust of Barthes' original proposition, that we have to find a way to allow more space to the reader than was allowed by classical literary theory. But I don't believe that, in order to do that, we need to get rid of the idea of authorship, not least because texts do still have authors, even if those authors are now sometimes collective agents rather than solitary individuals. To give you a different example, there's a wonderful moment where, in his debate with John Searle, Derrida claims his authority over his own texts very strongly. Thus, there's a moment in which he criticizes John Searle for having 'misread' one of his texts. And Derrida says: ' As the author, I claim the authority to tell Searle what I meant.' So there you have Derrida, the doyen of deconstructionism, claiming authority by virtue of authorship. That perhaps indicates that authors are not yet quite dead, and indeed, they don't need to be dead in order for readers to breathe. HJ: Don't you think the ways in which television programmes are structured, or rather, de-structured, are postmodern-styled? You know, in his article 'Cleverness is All', which you quoted to deal with the topic of postmodernism, M. Ignatieff finds that in television programmes narrative is replaced by flow, connection by disconnection and sequence by randomness . Besides, as you noticed, Dick Hebdige also points to the characteristics of such texts: 'Popular culture o¡ers a rich iconography, a set of symbols, objects and artifacts, which can be assembled and reassembled by di¡erent groups in a literally limitless number of combinations' . My further question, then, is whether this new kind of text, appearing ¢rst and foremost in television programmes, was a driving force, among others, that helped to open your horizon of the active audience. --- DM: No, my analysis was developed in the mid 1970s, long before this notion of postmodern 'flow', disconnection, or randomization really had much significance in the world of the media. The media I was analysing, both in the 'Nationwide' project and in the Family Television project, were fairly conventional forms of narrative, and were characterized neither by disconnection or randomness. But, beyond that point, I have some conceptual problems with the question. First, I'm not convinced by the generalization that television are 'nowadays' more disconnected and random than they were at some earlier moment. I'm not sure how powerful that proposition is. It may be true at a high level of abstraction, but there really isn't very much which is true about 'Television in General' . So, from my point of view, as I explained earlier, it's difficult to contrast the 'television of today' with the 'television of yesterday' unless you fill in at least some of the gaps and say which television, where? Otherwise, the generalization is so large as to hide rather more than it reveals. Indeed, one might well argue that there has always been much more variation within television at any given moment than there is between the television of one era and the television of another. Second, I don't think that audience activity depends on disconnection; people can be very active with very closed texts. Umberto Eco is very interesting about the contradictions that one can see at play, in that situation, in relation to the manipulation of 'closed' texts. I also think there is an interesting question to do with historical development. In the 1970s, you get a very important moment when structuralist analysts criticize the traditional forms of content analysis, insisting that you cannot understand a television programme by simply counting the disaggregated bits, in the way that traditional content analysis does. Their argument is that a programme is a structured whole, and so you can only understand the meaning of one bit of content in relation to the other bits with which they are combined. That analysis, first developed in film theory was then largely taken over from film to television studies, so then many scholars stopped doing the traditional kinds of content analysis of television, because they understood that they needed to look holistically at the whole structure of a programme. But then a further difficulty came with the ethnographic evidence that gradually accumulated, showing that's not how people watch television, especially once you have conveniences such as the automatic control device ŵith which a person can change channels without moving, by simply flicking a button. What you begin to get then is a new mode of viewing, in which people don't necessarily sit down and watch whole TV programmes. In fact, most people, most of the time, only consume bits of different television programmes ^they might go out to do something else in the middle, they might change programmes and watch something else and then come back later to the one they started with. At that point, the structuralist critique of content analysis loses its force. One has to recognize that, in so far as viewing modes are increasingly fragmented, it might be the case that the kind of 'cumulative effects' of media patterns of images and statements, of the type researched by Gerbner, might have more relevance than the structuralist critics of the 1970s have supposed, precisely because that critique was premised on the audience displaying a mode of focused attention on the complete, single text. That premise may be feasible if your object of study is film, shown in special places called cinemas ^but it is not feasible when applied to the study, for instance, of domestic television, where the circumstances generally dictate a rather different, more distracted, mode of consumption. However, one still has to allow for certain exceptions. There are moments, even in contemporary conditions of domestic consumption, in which someone might well sit down and watch a whole text ^e.g. of their 'favourite' programme. Conversely, there may be occasions in which someone goes to the cinema and, for whatever reason, doesn't concentrate on the film all the way through. You can't presume a one to one relationship between a particular architecture of display and a particular mode of consumption. Nor can you presume that there is only one mode of consumption for one particular medium. One can realistically, and usefully, only research the identifiable patterns of consumption of different media in different contexts, and one must always be sensitive to variations of the sorts I have indicated. HJ: In the field of popular culture studies, how has British cultural studies responded to French post-structuralism or, broadly, to postmodernism? Would you please give me an overview of their interactions or contestations in interpreting popular culture? It seems to me that British cultural studies has never given up its modernist perspective by which popular culture is not depthless, superficial and then meaningless as Fredric Jameson claims, but a site of ideological struggles, negotiations, concessions, compromises, or in short, of exactly that meaningfulness. DM: My own position is that, interesting and stimulating as both Jameson and Baudrillard can be, there is something fundamentally elitist about the contempt in which they both hold popular culture and, thus, about the contempt in which they implicitly also hold the majority of the people who are engaged with popular culture. To dismiss this whole field as depthless, superficial and meaningless, as they do, seems to me an act of the most stunning arrogance and that is definitely not a direction which I'd want to follow. I notice that, in a previous question, you referred approvingly to the work of Hebdige. I think that, by contrast to people like Jameson and Baudrillard, he is an elegant example of someone who would never make the mistake of dismissing popular culture as either depthless, superficial or meaningless. Rather like Hebdige, my concern is to understand the forms and modalities of popular culture, but, in doing so, I take very seriously the meanings at stake for its participants. To go back to one of my answers to an earlier question, I think it would be impossible to understand the rise of Thatcherism in Britain without understanding how that battle was fought out on the field of popular culture. I don't think that was in the slightest bit 'meaningless' . I do also think there was a difference between what one could see 'on the surface' and the hidden significances that one could read, at a deeper level, from the visible, 'symptomatic' events. Some of the things that might have seemed rather superficial, in fact, turned out to be of huge cultural, economic and political significance, and it was only by studying these seemingly trivial shifts in popular culture very carefully that it was possible to establish what was happening in British culture and politics in that period. The same would follow now: one needs to analyse the contemporary developments of popular culture with the same degree of seriousness ^a theoretical point which goes back, beyond Stuart Hall, to Richard Hoggart's early work on popular culture in the UK in the 1950s. HJ: There is a view which says that the good days of British cultural studies are now long past. It might be true to the extent that Stuart Hall has been retired for years and the CCCS has been closed, and that, what is more critical, there have been few new themes and core figures emerging that generate as much interest as broadly as before. However, from another point of view, after these changes it is time for us to redefine, with some critical reflections, British cultural studies or cultural studies. There are many other centres for cultural studies both in and outside Britain, that is, cultural studies becomes increasingly international. Concerning the new context, I am quite interested in questions such as how British cultural studies is going on in Britain, in what way you and your colleagues continue British cultural studies, how we can promote cultural studies in a global sense, and last, back to our main topic in this interview, how you think about popular culture that crosses boundaries: popular culture is always not only domestically but also globally popular. DM: Clearly, this is a very difficult and important issue. The story of the influence, internationally, of British cultural studies, is perhaps best understood as a kind of 'export industry' through which a particular set of perspectives, initially developed in Britain in the 1960s and 1970s, addressing the specific problems of British society, were gradually exported to the English-speaking territories of the previous British empire, most notably Australia, Canada and then, at a later point, to America. Manifestly, British cultural studies was designed to analyse the situation of a particular country at a particular time. Of course, there are certain theoretical and analytical positions built into it which can be abstracted and transported to different circumstances, allowing for the relevant variations and distinctions that need to be made. However, the idea that British cultural studies can usefully be exported as a kind of 'ready-made' template, according to which the whole world can be understood, is clearly nonsensical: that would just be another form of cultural imperialism, this time in the intellectual sphere. Cultural studies, wherever it is conducted, has to construct modes of analysis which are relevant to its own conditions of production in a particular place, at a particular time. For instance, it may well be that someone trying to develop a cultural studies perspective in China now, as I understand you are doing, can hope to learn certain things from the history of the development of cultural studies in Britain, or elsewhere. But you have to approach those previous intellectual traditions, located elsewhere, with a strong element of critical intelligence, and you have to think carefully about in what ways, given the many differences between British and Chinese society, you would need to transpose the modes of analysis developed in Britain in order for them to be of any use at all in China. So again, that takes us back to the contextual specificity of cultural studies. One also has to be attentive to the way in which the world around us has changed significantly since the 1970s when my own audience work was actually conducted. The model of cultural studies that we developed in Britain then was one which was premised on the notion of a national society. We were primarily concerned with class, race and gender differences within Britain, and with the British media. Certainly, those still remain very important problems but nowadays, communications is simply not national in the way that it was in the 1970s. We now have transnational broadcast systems, and satellite systems of communication, which mean that messages are mobile, all over the place. Cultures are no longer simply national. That is the force of Appadurai's argument about 'difference and disjuncture ' in the global economy, when he talks about how the 'mediascapes' of the contemporary world now exceed national boundaries. He also talks there about how the 'ethnoscapes' of our contemporary situation also now transcend national boundaries, in so far as we live in a situation in which not only are messages mobile across boundaries, but so too are audiences. In a world of increased migrancy, we are now in a situation, as he says, where 'moving messages meet deterritorialized viewers ' . Now, clearly, that is a very different situation from the one that we faced in Birmingham in 1972, when I began my studies of audiences ^and we have to allow for those differences in any contemporary analysis that we want to make. At the same time, to go back to what I said before about my anxiety about clearly 'binarized' divisions and periodizations, I don't want to suggest that the age of the nation-state or 'the national era' is completely over ^or that we now all live in some 'transnational' period, in which we are all equally mobile and all attending to messages which come to us from far away. That's simply not the case. Some people, in some places, are much more transnational than others, and many still effectively live out their lives within national ^and even more local ^boundaries. That was the problem that I explored in my book Home Territories , which was focused on questions of mobility and the media, but was also insistent on continuing to pay attention to the ways in which boundaries of a local, regional or national kind, are still, in many ways, enforced today. Indeed, it would seem to me that in China, in particular, the nation-state is not only 'still alive' but is both thriving and powerful, if in different ways than within the nations of Europe. In ending, let me say again that we clearly cannot take any mode of analysis, be it British cultural studies or anything else, and imagine that it will automatically help us understand the situation in another culture at a different time, without making all manner of cultural translations and transpositions. However, on the other hand, we have no need to imagine that we live in some totally 'new' world, where all previous theories are now redundant. As people like Carolyn Marvin have shown us, all technologies, all media, were new in their own time: the telegraph, the cinema, the radio, the television, the video game, the computer . There is always a moral panic about the latest 'new' medium and its supposed 'e¡ects', and many of the problems we face today have clear historical precedents which we need to consider. I think that we have to develop a much more serious historical and comparative set of perspectives within cultural studies. And I hope that you and your colleagues in China will ¢nd some parts of what I have to say in this interview of use to you in your project of trying to do that! Thank you. --- heartfelt thanks are due to the British Academy, which approved my research proposal on British cultural studies and generously awarded me a fellowship which, among others, made this interview possible.

British cultural studies, represented perhaps chiefly by the so-called Birmingham School, is much marked by its strong orientation towards the application of grounded theory in the analysis of concrete cases, rather than the development of abstract Theory with a Capital T (in Stuart Hall's words). As a leading figure of the Birmingham School and a key representative of the active audience model in television studies, or broadly, media studies, David Morley stands at a point where this trend was set, as is evidenced in this interview. Questioned by Huimin Jin, Morley puts his audience studies into the contexts of British cultural studies, postmodernism, Marxism, social movements, and so on; and in doing so, he shows the ambiguity, and subtlety of his concepts of how to best theorize the active audience. Only by this method, Jin believes, could Morley launch his version of audience studies, which aims not to invent a general theory of media effects, but to use an interdisciplinary range of theories to explore how people actually respond to a TV programme; and only by this approach to audience studies, furthermore, could Morley develop a theory of the audience's activity, which is embedded in the course of their everyday life and that cannot be thoroughly colonized by discourses. Cultural studies, wherever it is conducted, therefore, Morley suggests, has to construct modes of analysis that are relevant to its own conditions of production in a particular place, at a particular time. This is the tradition, as we know it, but also the future, as Morley envisages, of cultural studies.

Introduction While school-based violence can take many forms, bullying is particularly prevalent. Despite definitions varying, bullying is generally considered to be "intentional repeated overt or covert inappropriate behavior from another that is intended to intimidate and harm the target" [1]. In school settings, a student is bullied "when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other students" [2]. For example, negative actions, or bullying, can be direct or indirect and include physical, verbal, or psychological and relational acts that cause children to be systematically excluded from social activities by their peers [3,4]. In many countries, rates of bullying are high. According to a cross-national study of adolescents aged 11.5 to 15.5 years, the prevalence of bullying victimization ranges from 6% in Sweden to 40% in Lithuania, with an average rate of 17% across 25 countries in Europe and North America [5]. Another cross-national study, the Global School-based Student Health Survey , carried out among middle school students in 19 low-or middle-income countries showed that the prevalence of bullying ranged from 8% in Tajikistan to 61% in Zambia [6]. In the 2011 Progress in Reading and Literacy Study , which draws on a sample of fourth grade students spanning 52 countries and regions, about 53% of students reported that they had been bullied at school, and 20% said that they were bullied "almost weekly" [7]. Bullying has been shown to have significant consequences on a student's educational experience and long-term outcomes [8,9]. Studies have found that being bullied at school causes school avoidance and poor attendance [10], inability to concentrate [11,12], lack of academic engagement [13][14][15], early school dropout [16], and a weaker sense of belonging or connection with school [17]. In addition to academic measures, school environments may have an influence on student creativity [18], and bullying has been found to detrimentally affect school environments [19,20]. As such, in addition to the immediate psychological and social consequences, the negative impact of bullying on educational performance can inhibit human capital accumulation, labor market opportunities, and economic productivity in the long-term [21][22][23]. In China, little is known about school-based bullying. The available research, which draws upon samples that are almost exclusively urban, finds that the prevalence of selfreported bullying victimization varies from 2% to 26% [24][25][26][27][28][29]. Similarly, studies of the correlates of bullying in urban China have varied widely. While some studies show that bullying victims in China tend to be male students [25,27,28], a study in Tianjin finds no indicative gender pattern in bullying victimization [30]. Other evidence suggests that the gender of teachers was associated with their responses to student bullying behaviors in school [31][32][33]. In addition, some research reports that students from single-parent families had significantly higher rates of bullying victimization than children from two-parent families [34,35]. However, another study shows that students who were dissatisfied with their parental caring, not students who experienced parental absence, were at a higher risk of being bullied [27]. Although existing studies have facilitated an understanding of bullying victimization in urban China, the existing research provides little insight into the extent of the bullying problem in China's vast rural school system, where most of the country's children attend school. This area of inquiry is particularly important given that rural students lag far behind their urban counterparts in academic achievement and attainment [36][37][38]. To the extent that bullying negatively impacts the social emotional well-being of students, as well as their educational performance and creativity , an understanding of bullying in rural schools is an important step in narrowing the ruralurban gap in China's education system. China is currently attempting to grow from an upper middle-income economy to a high-income economy, but insufficient human capital in rural areas threatens this transition [36]. As the government is exerting considerable efforts to improve human capital in both urban and rural areas [39], understanding the various causes of rural academic underachievement is crucial to effectively narrowing the rural-urban gap. A more thorough understanding of the prevalence, correlates, and consequences of bullying victimization in rural China will not only inform efforts to reduce bullying and improve student well-being; it will also offer insights into how to improve human capital accumulation in rural China and other developing contexts. The overall goal of this study is to examine the prevalence, correlates, and consequences of bullying victimization in China's rural schools, focusing on its relationship to educational performance and creative attitudes. More specifically, the hypothesis we would like to test in this paper is the following: bullying victimization is correlated with lower levels of academic performance and lower levels of creative attitudes among rural students in China. To achieve this goal, we pursue three objectives. First, we document the prevalence of bullying among primary school students in rural China and compare this to other countries using an international comparative metric from the Trends in International Mathematics and Science Study and Progress in Reading and Literacy Study survey [40][41][42][43][44]. Second, we identify the student family and school characteristics that are correlated with bullying victimization. Third, we examine whether bullying victimization is correlated with student academic performance and student creative attitudes both before and after controlling for school, student, and family characteristics. The rest of the paper is organized as follows. Section 2 introduces sampling methods, data collection, and methods for measuring bullying victimization, academic performance, and creative attitudes. Section 3 reports the prevalence and correlates of bullying victimization and the relationship between bullying victimization, academic performance, and creative attitudes. Section 4 discusses the results. Section 5 concludes. --- Materials and Methods --- Sampling The data presented in this study were collected from three rural counties in the southern part of Jiangxi Province in China . Although our sample is from one province of China, it is fairly representative of poor rural counties across China in terms of key economic and social indicators. First, all three sample counties are nationally designated poor counties that were identified by the Chinese government in 2012 as areas with extreme poverty [45]. As such, the economic development in these three counties lags behind the national average in China as well as other areas of Jiangxi Province. Per capita disposable income in each of the three counties was less than RMB 8200 in 2015, which is similar to the average per capita disposable income of RMB 9264 in the 832 nationally designated poor rural counties across China [46]. Additionally, more than 80% of the population in the three counties are rural residents [46]. In these respects, the three sample counties are typical of poor rural areas across China, which are home to nearly one fifth of China's total population. To select our sample, we followed a two-step sample selection protocol. The first step involved selecting a representative sample of schools from the three counties. To do so, we used official records from county education bureaus to create a population frame of all rural, public primary schools in the three counties, totaling 458 schools. We then randomly selected schools using a sampling fraction that ensured the total number of schools in each township was proportionally represented in our sample. This led us to randomly select 120 schools, of which 37 were in County A, 25 were in County B, and 58 were in County C . In this way, our sample is representative of the three counties being studied. After selecting schools, we next sampled classes and students in grades four and five. Due to financial constraints, we randomly selected at most two classes in each grade in each school. Specifically, if there were one or two classes in a grade, all classes in this grade were selected. If there were more than two classes in a grade, we randomly selected two classes. We then surveyed all students in the sampled classes. We also surveyed the math teacher and Chinese teacher for each sample class. Our final sample included 10,528 students from 286 classes in our 120 sample schools . This sample is far larger than previous samples that have been used to examine bullying in Chinese schools [26,27,29]. --- Data Collection All data collections were conducted at the end of the school year in May 2015. We collected four blocks of data. The first block collected information on bullying victimization in school. The second block collected socioeconomic information about students, households, and schools/teachers. The third and fourth blocks collected data on student academic performance and student creative attitudes, respectively. --- Bullying Victimization To collect information on student bullying victimization among our sample, we used the "Students Bullied at School" scale. The SBS scale was developed for the Trends in International Mathematics and Science Study and Progress in Reading and Literacy Study [40][41][42][43][44]. Both the TIMSS and PIRLS are international comparative assessments of academic achievement among fourth grade students across 52 countries and regions representing a variety of development and income levels [7]. The SBS survey was translated into Mandarin Chinese and the translation was verified according to the PIRLS translation guidelines [47]. The SBS scale also has good reliability among teachers in rural China with a Cronbach's alpha reliability coefficient of 0.93. The SBS scale asks students to rate how often they experienced each of six bullying victimization behaviors. The six behaviors are I was made fun of or called names; I was left out of games or activities by other students; Someone spread lies about me; Something was stolen from me; I was hit or hurt by other student ; and I was made to do things I didn't want to do by other students. To create a raw score for the SBS scale, each response was assigned a numerical value . The raw scores range from 0 to 18 . A lower SBS score therefore corresponds to a higher level of bullying victimization in school. Following the PIRLS protocol, raw scores were transformed into SBS scaled scores, which were then used to sort students into three categories by frequency of bullying victimization: "Almost Never," "About Monthly," and "About Weekly." Transformed SBS scale scores range from 3 to 13 points. PIRLS guidelines categorize students with transformed scores higher than 10.1 as "Almost Never" bullied; students with transformed scores between 8.3 to 10.1 as bullied "About Monthly;" and students with transformed scores below 8.3 as bullied "About Weekly." Students experience bullying victimization behaviors "about weekly" and "about monthly" are considered to be "frequently bullied". --- Socioeconomic Information The survey team collected data on the basic socioeconomic information of each student, as well as information about each student's family, teacher, and school. Student socioeconomic information included gender, grade, and a seven-item checklist of household assets. The checklist asked students to indicate whether the family owned a car, a microwave, a refrigerator, a camera, a computer, an electric fan, and/or a flush toilet. A value was attached to each asset to produce a single metric of household asset holdings. Summing the value of all household consumption assets then produced our proxy variable for family asset value. Enumerators also asked students about where they lived during most of the school year-at home or in the school's boarding dormitories . In addition, a survey form was sent to each student's caregivers to collect data on parental education levels, and on whether the student's parents were often away from home. Finally, we surveyed teachers and principals to collect data on the gender of Chinese and math teachers in each class, the student-teacher ratio of each school, and the school's distance from the local government seat. --- Academic Performance Tests We conducted a set of academic performance tests, including a 30-min standardized reading test, a 30-min standardized Chinese language test, and a 30-min standardized mathematics test. All sample students were administered the reading test, which was carefully designed to measure student reading skills. The test questions were adapted from those found in the PIRLS test. The test questions were carefully translated according to the PIRLS translation guidelines and reviewed by a panel of experts and local teachers who are well-versed in China's education system. The translated reading tests then went through several rounds of pilot tests in Chinese schools. The results were independently reviewed by a group of test assessment experts and were revised to make sure they were of the highest quality and appropriate for the designated student levels. In addition to the reading test, students were administered mathematics and Chinese language tests. The tests evaluating math and Chinese language were carefully designed with assistance from educators in the local education bureaus to ensure coherence with the national curriculum, and both exams were pre-tested multiple times to confirm their academic relevance and appropriate time limits. Within each selected school, we randomly assigned half of the sample students in each classroom to take the math test; the other half of sample students in each classroom took the Chinese test. In total, 5237 students took the reading and Chinese language tests, while 5291 students took the reading and math tests. Trained enumerators proctored all exams to prevent cheating and enforce the 30-min time limit. For ease of interpretation, we converted all test scores into z-scores using the mean and standard deviation of scores in each grade. --- Creative Attitudes In the final survey block, students were asked to complete the Schaefer's Creativity Attitude Survey [50]. This instrument includes 32 questions designed to assess children's attitudes associated with creativity, such as confidence in one's ideas, appreciation of fantasy, openness to impulse expression, and use of novelty [50]. Students were given unlimited time to complete the survey. The total score ranges from 0 to 20, with higher scores indicating greater creativity. --- Analytical Approach Our analysis is comprised of three parts. First, in our initial analysis, we looked at the prevalence of bullying at school. Second, to understand what kinds of students are more likely to be bullied, we compared the rate of bullying victimization with different individual, family, teacher, and school characteristics. We conducted t-tests to measure for significant differences between groups and to analyze which characteristics correlate with bullying victim status. Finally, we estimated the correlation between bullying victim status and academic performance, as well as the correlation between bullying victim status and creative attitudes. To do so, we used an ordinary least squares regression model, including a set of covariates in a regression on student dropout. We first ran an unadjusted regression : Y ij = α + βBullied i + ϕ j + ε i where the dependent variable Y ij denotes the academic performance and creative attitudes of student j in school k. Bullied i is a dichotomous variable that equals 1 if the student is being bullied at school about weekly or about monthly, and equals 0 if the student is almost never bullied at school; ϕ j represents county-level fixed effects; and ε i is an error term capturing shocks and characteristics that are specific to the student or are unobserved. β is the within-school mean gap in academic performance and creative attitudes between bullying victims and non-bullied students. To control for the potential confounding effects of student, family, teacher, and school characteristics, we ran a multivariate analysis building on Equation above with the addition of a vector of control variables. Y ij = α + βBullied i + γX i + ϕ j + ε i where the vector X i includes student, family and teacher, and school characteristics. Student individual characteristics include binary variables representing student gender, grade, and boarding status. Family characteristics include: household wealth , type of family , whether parents graduated from junior high school, and whether parents are often away from home. We control for Chinese teacher gender when the outcome variable is Chinese or reading performance, for the mathematics teacher gender when the outcome variable is mathematics performance, and for the gender of both teachers when the outcome variable is creative attitudes. School characteristics include the student-teacher ratio , and distance between the school and government seat . We chose to include the aforementioned variables in our equation based on previous studies that identified them as important factors correlated with student academic achievement. Past research has indicated that individual [51], family [48,51], teacher, and school characteristics [52] are all closely related to the academic performance of students. Therefore, controlling for these variables allows us to better isolate the effects of bullying on student achievement and compare the magnitude with other factors. In both Equations and , we computed cluster-robust standard errors . --- Results --- Prevalence of Bullying among Primary School Students in Rural China Figure 1 presents the rates of bullying victimization among students in our sample. Our results show that the prevalence of bullying victimization among primary school students in rural China is alarmingly high. Seventy three percent of sample students were bullied frequently . This is a significantly higher rate of bullying than the average rate across 52 countries reported by the TIMSS and PIRLS survey , in which 53% of students experienced bullying weekly or monthly [3]. The prevalence of bullying in our sample is also substantially higher than previous estimates of bullying in China. Previous studies have found that the prevalence of bullying is around 20% in Hong Kong [53,54] and 2% to 26% in urban areas of mainland China [24][25][26][27][28][29], meaning that the prevalence of bullying in our sample is nearly three times higher than the highest prevalence reported in studies of urban China. In addition, among our sample, 41% of student reported being bullied almost monthly, and 32% reported being bullied almost weekly. This is also much higher than the rates reported in the 2011 TIMSS and PIRLS survey, which found that only 33% of students experienced bullying monthly and only 20% were bullied weekly [7]. In comparison to the bullying victimization scale score of other countries and regions, our results show that students in rural China experience bullying more frequently than students in most other places in the world. The average SBS scale score for our sample of rural students is 9.2, meaning rural Chinese students are categorized as "being bullied monthly", according to the TIMSS and PIRLS survey cutoff. This puts our sample region at a rank of 5th out of 52 countries or regions, with only slightly lower rates of bullying than South Africa, Botswana, Qatar, and Trinidad and Tobago . When we examine each of the six bullying behaviors assessed in the SBS scale, we find that certain types of bullying were reported more frequently than others among our sample students. The most common types of bullying that students experienced were "being made fun of or called names", "being stolen from", and "exclusion from games or activities by other students" . Of our sample students, 73% had been made fun of or called names at least a few times in the past year; 73% had had personal items stolen from them at least a few times in the past year; and 58% had been excluded by other students at least a few times in the past year. The three other bullying behaviors included in the SBS scale were also fairly commonplace among sample students: in the past year, 50% of students had experienced lies being spread about them at least a few times; 50% had been hit or hurt by other students at least a few times; and 41% had been forced to do things they did not want to do at least a few times. When we examine each of the six bullying behaviors assessed in the SBS scale, we find that certain types of bullying were reported more frequently than others among our sample students. The most common types of bullying that students experienced were "being made fun of or called names", "being stolen from", and "exclusion from games or activities by other students" . Of our sample students, 73% had been made fun of or called names at least a few times in the past year; 73% had had personal items stolen from them at least a few times in the past year; and 58% had been excluded by other students at least a few times in the past year. The three other bullying behaviors included in the SBS scale were also fairly commonplace among sample students: in the past year, 50% of students had experienced lies being spread about them at least a few times; 50% had been hit or hurt by other students at least a few times; and 41% had been forced to do things they did not want to do at least a few times. --- Figure 2. Comparison of average students being bullied at school scale scores between rural China and other countries/regions. Note: a lower SBS score corresponds with a higher level of bullying victimization in school. To enable comparison across countries and regions, SBS raw scores have been converted to scales scores according to the PIRLS conversion chart. --- Factors Correlated with Bullying Victimization Table 3 presents the results of our analysis of the characteristics associated with bullying victimization. We found that several student characteristics were associated with higher rates of bullying victimization, namely gender, age, and boarding status. Boys were slightly more likely to be bullied than girls . In addition, grade five students reported being bullied at a slightly higher rate than grade four students . Students who board at school also experienced significantly more bullying --- Factors Correlated with Bullying Victimization Table 3 presents the results of our analysis of the characteristics associated with bullying victimization. We found that several student characteristics were associated with higher rates of bullying victimization, namely gender, age, and boarding status. Boys were slightly more likely to be bullied than girls . In addition, grade five students reported being bullied at a slightly higher rate than grade four students . Students who board at school also experienced significantly more bullying in comparison to students who live at home . However, it is important to note that although these differences are statistically significant, the rates of bullying victimization for all groups are far higher than the international average. Regarding household and school characteristics, several variables associated with socioeconomic disadvantage were significantly correlated with more frequent experiences of being bullied. Students in the lowest quartile of family asset value, students from single parent families, and students whose fathers or mothers did not finish junior high school were more likely to experience bullying victimization . Interestingly, there was no significant difference in bullying victimization between the left behind children and students whose parents lived at home . We also found that remote schools were associated with higher bullying rates, and students at schools with higher student-teacher ratios experienced more bullying than stu-dents at schools with lower student-teacher ratios . However, in contrast with previous studies [31 -33], our results find no association between the gender of teachers and students bullying victimization. To further investigate the correlates of bullying victimization, we ran a multivariate regression that includes all student, family, teacher, and school characteristics . The results are overall consistent with that of Table 3. Of the individual student characteristics, boys and boarding students tended to be bullied more frequently . Our finding that boarding students experienced more frequent bullying is consistent with past findings in urban China and in other countries that boarding students show higher levels of bullying victimization and perpetration in comparison to non-boarders [55][56][57]. As Pfeiffer and Pinquart discuss, this may be because students who board at school spend more time with peers, for example, in dormitories [55]. Similarly, Chui and Chan suggest that more peer contact, especially with deviant peers in dormitories, contributes to bullying victimization [57]. This is certainly the case in rural Chinese boarding schools, where students may share a dormitory room with more than ten peers [58]. Moreover, separation from parental influence may reduce the impact of parenting practices meant to reduce problem behavior, leading students to engage in more bullying behaviors [55]. Additionally, the results of our multivariate regression show that students whose families have the lowest family asset values, students in single-parent households, and student whose mothers or fathers have not completed junior high school were more likely to experience frequent bullying . These findings are consistent with previous studies that have found children and adolescents from families of lower socioeconomic status are more likely to be involved in bullying victimization [59][60][61][62]. Low family wealth itself may be one cause of victimization, as students have less access to financial resources and inability to afford material goods [35,62]. Additionally, as Jansen et al. [63] suggest, students from single-parent households are more likely to experience an increased level of stress due to broken families and fewer parent-child interactions. The latter explanation is further supported by the evidence presented in Spriggs et al. that found a reduction in parental involvement and communication to be associated with increased bullying victimization [64]. Among our sample, the higher magnitude and significance of single-parenthood than family asset value implies that parenting characteristics associated with single parent households are contributing factors to the risks of bullying victimization. The literature also supports our finding that students whose parents have lower education levels are at greater risk of being frequently bullied. Research has shown that there is an inverse relationship between parental education levels and child bullying victimization, especially maternal education levels [60,[65][66][67]. Together with family asset value and single-parenthood, parental education levels serve as a socioeconomic status indicator [63]. In addition to financial resources, parents of higher socioeconomic status can provide more time, knowledge, and aid to help their children cope with social conflicts. Both aspects reduce the chances that children will be bullied. This point is further supported by the statistically insignificant relationship between whether the father or mother lives at home and bullying victimization. The mere presence of parents at home, when other family factors are controlled for, does not have a significant impact on whether a child is bullied. Instead, it is the underlying socioeconomic status of the family, determined in large part by parental education, which matters. Finally, we found that students at schools with higher student-teacher ratios tended to be bullied more frequently . Frequent bullying victimization has been found in the literature to be associated with a large school size [68,69] and a high student-teacher ratio [70]. This may be because high student-teacher ratios limit teachers from effectively managing student behavior and preventing bullying victimization [70]. --- Variables Being Bullied --- Correlations between Bullying Victimization, Academic Performance, and Creative Attitudes Table 5 presents the results of our OLS regression analysis examining the correlation between bullying victimization and academic performance. Our unadjusted OLS regression results show that experiencing frequent bullying is significantly correlated with lower academic performance. Being bullied monthly or weekly is associated with a decrease of about 0.26 standardized deviations in Chinese language performance, 0.25 standardized deviations in reading performance, and 0.22 standardized deviations in mathematics performance. Our adjusted OLS regression results tell a similar story: frequent bullying victimization is associated with a decrease of about 0.21 standardized deviations in Chinese language performance, 0.22 standardized deviations in reading performance, and 0.21 standardized deviations in mathematics performance. The magnitude of the effect of bullying victim on student academic performance is equivalent to almost half a year of learning [71]. Additionally, we find a significant negative correlation between bullying victimization and student creative attitudes . Students who were bullied weekly or monthly scored lower in creative attitudes than their peers in both the unadjusted and adjusted models, with coefficients of 0.60 and 0.57, respectively . This corresponds to a difference in creative attitudes of about 0.24 standard deviations. The findings in both Tables 5 and6 are all significant at the 1% level. In other words, the magnitude of these associations and their statistical significance are similar in the nonadjusted and adjusted regressions for both academic performance and creative attitudes. To better understand the relative "importance" of bullying for academic performance and creativity, we compare the magnitude of bullying victimization to other significant factors in our multivariate analysis. Of note, the magnitude of the correlation between bullying victimization and student academic performance is as large as the gender gap and single-parenthood, and it is greater than that of boarding status, low parental education and absent fathers. Moreover, the magnitude of bullying victimization on student creativity ability is greater than all other factors measured in this study, including gender, boarding status, low family asset value, single-parenthood, low parental education levels, remote school locations and high student-teacher ratios. These results confirm our initial hypothesis that bullying victimization has a negative impact on academic performance, consistent with the findings of previous studies [8]. More specifically, the literature has found that when a student experiences bullying, the stress incurred from victimization can lead to school avoidance, poor class attendance, and the inability to concentrate in class, all of which directly impede learning and academic achievement [10][11][12]. Similarly, the creative attitudes of students are hampered when they experience bullying. As previous research suggests, positive and encouraging environments are more likely to foster creativity [72,73]. The stress and reduced self-esteem incurred from bullying victimization might create a hostile environment that inhibits student creativity. Moreover, when we compare the magnitudes of correlation, we find that bullying victimization is negatively correlated with academic performance and creative attitudes at a larger magnitude than many other factors. This suggests that reducing bullying in rural schools may have a larger positive effect on student academic performance and creative attitudes than targeting other factors that have been traditionally considered. --- Discussion To the best of our knowledge, ours is the first large-scale study to document the prevalence of bullying victimization among students in rural China. This study also identifies the student individual, family, and school characteristics correlated with bullying victimization. Additionally, this is the first study to examine the correlation between bullying victimization and student educational performance in rural China, before and after controlling for student and family characteristics. Our study finds that students are bullied in rural Chinese elementary schools at rates far higher than both the international average and the rates found in urban areas of China. About 73% of sample students in rural China were bullied almost monthly or weekly, which is much higher than the international average of only 53%. The prevalence of bullying in rural China is also much higher than that of Hong Kong and urban areas of mainland China, where rates of bullying vary from 2% to 26% [24][25][26][27][28][29]53,54]. The most frequently experienced type of bullying in rural China is being made fun of or called names by other students. We also found several characteristics predictive of frequent bullying victimization, including being male, boarding at school, having less educated parents, having a family with a lower family asset value, having a single-parent family, and attending a school with a higher student-teacher ratio. Perhaps most importantly, even after controlling for student, family, teacher, and school characteristics, student academic performance and creative attitudes are both strongly negatively correlated with frequent experiences of bullying. A student being bullied monthly or weekly is correlated with a decrease of about 0.21 standardized deviations in Chinese performance, 0.22 standardized deviations in reading performance, 0.21 standardized deviations in mathematics performance, and 0.57 points in creative attitudes. In other words, bullying significantly impedes the ability of students to perform academically and think creatively. Why do we see such high rates of bullying victimization in rural China? One factor may be that rural teachers lack the time and resources to intervene and prevent bullying. With China's rapid economic progress and intensified urbanization, more financial and educational resources have been poured into cities. Rural areas, though on the radar of policymakers, still face a scarcity of educational resources [74,75]. Additionally, fiscal decentralization has aggravated the unequal distribution of resources in rural areas, leading to low per-pupil basic education expenditures in rural schools [76][77][78]. This means that rural teachers face lower salaries, worse working conditions, heavier workloads, and limited school budgets for professional development compared to their urban colleagues, all of which lead teachers to move within the public education system to better-paying urban schools, creating a shortage of rural teachers [79,80]. As a result, rural schools tend to have larger class sizes and higher student-teacher ratios, leading to less individualized attention for each student. This can explain in part the observed high prevalence of bullying victimization. Especially in an exam-centric education system such as in China, where the exam scores are the primary metrics for academic performance, teachers tend to focus their limited time and effort on teaching and student academic achievement rather than behavior management. Another key factor may be parenting quality and socioeconomic challenges in rural China. Low socioeconomic power in rural families creates adversities for rural children. For example, many rural parents leave rural areas for jobs in China's cities and leave their children behind in the countryside in the care of grandparents. The absence of parents and reduced parental involvement deprives children of opportunities to learn conflict management skills and other related social skills [61]. Moreover, grandparents often have even lower levels of education, which means that can only provide limited guidance to address the troubles in their children's social lives [81][82][83]. Bearing the disadvantages of both educational resources and parental attention, rural students face a higher level of bullying victimization. The negative consequences such as stress hinder not only their academic performance but also limit their creative abilities. This, in turn, perpetuates the preexisting urban-rural educational divide. --- Conclusions The findings in this paper offer insights into factors that may contribute to student academic performance and creative attitudes in rural China beyond those that have been traditionally considered, such as teacher quality, school funding, and student physical health. While our data do not support a causal analysis of the relationships among bullying, academic performance, and creative attitudes in rural China, studies conducted in other countries can provide guidance for future causal research in China. Specifically, many studies find that bullying has negative consequences for a student's educational experience and long-term outcomes [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23], disrupting their ability to learn, think, and thrive in the school environment. Future research into the effects of bullying in rural China should therefore focus on the causal links between bullying and physical/mental health. The results of this study also have implications for China outside of its academic system. If bullying victimization does in fact has a significant negative impact on student academic performance and creative attitudes, then widespread bullying may not only be hurting individual students; it may also harm China's long-term economic development by slowing human capital growth. Supposing that the proportion of students found to be frequently bullied holds true across rural China, then based on the Ministry of Education's statistic that there are about 30 million primary school students in rural China, 22 million primary school students are bullied frequently in China, at the expense of their academic achievement. Given the important link between education and human capital, the correlation between bullying and academic performance offers a direction for improving human capital in rural areas, where widespread low academic performance and high rates of dropout negatively affect educational attainment and labor market performance among large parts of the population. Given these findings, we recommend that China's education policymakers consider incorporating bullying research and prevention initiatives into their agenda. Bullying is a complex and pervasive phenomenon, and it is necessary to understand the problem in full, especially in poor rural areas where academic performance lags in comparison to urban areas. Therefore, causal studies of bullying and its effects on students are vitally important. In addition, although self-reported data on bullying victimization may reveal the true prevalence and magnitude of bullying behavior, future research should explore the characteristics of rural bullies using direct observation methods, as understanding the profile of bullies is critical in addressing bullying more broadly. Anti-bullying programs should also be piloted throughout rural schools. Furthermore, funding for controlled trials of such programs should be prioritized so that there is a solid evidence base for developing anti-bullying policies in the future. --- Data Availability Statement: Data are available upon request. --- Acknowledgments: We would like to acknowledge our collaborators in China. --- Author Contributions: Conceptualization, H.W. and J.T.; methodology, H.W., J.T.; software, H.W., J.T.; validation, H.W., J.T.; formal analysis, H.W., J.T.; investigation, H.W., J.T.; resources, H.W., S.R.; data curation, H.W., J.T., J.X.; writing-original draft preparation, H.W., J.T., S.-E.D., J.X., M.B.; writing-review and editing, H.W., S.-E.D., M.B., C.C.; visualization, H.W., J.T.; supervision, S.R.; project administration, C.C., H.W. All authors have read and agreed to the published version of the manuscript. Funding: This research was supported by Enlight Foundation . The funding source did not have a role in in the design or conduct of the study; the collection, analysis, or in-terpretation of data; the preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication. --- Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Stanford University . Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. ---

School bullying is a widely recognized problem in developed countries, but remains underinvestigated in developing countries, especially in remote rural areas. In this paper, we examine the prevalence, correlates, and consequences of bullying victimization and its relation to educational performance and creative attitudes. Using data from 10,528 students across 120 primary schools in rural China, we find an alarmingly high prevalence of bullying victimization and that several individual, family, and school characteristics are correlated with bullying victimization. Analyses indicate students who are bullied frequently score lower in Chinese, reading, and math tests and creative attitudes. Taken together, the results demonstrate a need for further research and policy interventions to reduce bullying in schools.

INTRODUCTION Since the dawn of humanity, communication has been a regular activity for passing information, triggering a continuous evolution from the drawings on the walls to the zettabytes of today that circulate around the world. This huge amount of data and the ability to transmit it in real time, is due to Information and Communication Technologies . The World Bank has recognized the importance of ICT for promoting economic, human, and social progress, sustaining that technological advance is the second force after economic progress, given that since 1990 it has lifted more than 10 percent of the world population out of poverty. Likewise, the United Nations highlighted that ICT are the right podiums for knowledge sharing, skills improvement, and some electronic services such as e-government, that can bring welfare in areas like employability, health, and education. The European Commission wants to ensure, until 2020, the growth of an information society across the members of the European Union by sponsoring innovation and knowledge sharing among member states, making the most of ICT, encouraging innovation and smart solutions, and creating the right structure to deal with the social challenges of the European and global society . The EC recognized that Europe is behind her economic partners in matters of growth rate, due in part to the lack of use of ICT, although the ICT sector in 2010 was responsible for more than 5 percent of European GDP and even more of the productivity growth . The novelty of ICT at the beginning along with the fast evolution of ICT in recent years, has led to discrepancies between and within countries, resulting in a technological divide. Research has been conducted to analyze the European digital divide, among or inside countries, but rarely both at the same time ; though "the digital divide within countries can be as high as that between countries" . The goals of this study are to appraise the European domestic digital divide for its 28 members, and to analyze Europeans' demographic characteristics, scrutinizing: the main dimensions of digital divide of the 28 members of EU; the main socio-demographic characteristics that most impact the digital divide in the EU; the main socio-demographic characteristics that potentially contribute most to the digital divide, and those that lead to greater asymmetries in the use of ICT, for each country of EU. In addressing these issues, this paper is structured in seven sections. The next one -section twopresents a literature review about digitalization, from the digital revolution to the current digital society with reference to the challenging digital divide. The third section is dedicated to the domestic digital divide and describes four socio-demographic drivers that influence ICT adoption. The fourth one defines the methodology, describing the data and the applied multivariate data technique, as well the outcomes of the data analysis. Section five presents the discussion of findings as well as the limitations of the study. Section six presents the theoretical and managerial implications. In the last section a synopsis of findings is presented. --- LITERATURE REVIEW --- DIGITAL REVOLUTION With the emergence of the information society, a turning point came in the living standards of individuals, with ground-breaking technologies restructuring the traditional society . The first computers, and later the Internet, enabled boundless information besides data, giving a new meaning to manipulation and exchange of these resources, creating a service society and increasing productivity and standards . The personal computer and the presence of the Internet are recognized in the literature as the key ingredients of the digital revolution, since they not only brought material condition and services that otherwise would be unreachable, but also redesigned both cultural and human nature. Directly or indirectly, this technological revolution influenced society's fundamental elements such as: world population, life expectancy, governance, entertainment, human relationships and others . The ability to use ICT and surf the Internet became requirements to live in contemporaneous society, also known as the Information Age . There are no doubts that in the past IT were already considered "the lynchpin of our modern society" being this revolution acknowledged as one of the most remarkable accomplishments of humanity . ICT is one of the most important General-Purpose Technologies , on a par with electrification. For instance, when comparing the U.S. labor productivity in the electrification era and the ICT era, remarkably similar patterns of productivity growth are visible, even though in different time periods . Still, in the year 2000 the IT revolution was just emerging, with mobile phones and Internet usage growing by more than sevenfold between 2000 and 2015 globally . --- DIGITAL DIVIDE As ICT became better known within societies and the Internet technologies became popular, given the "placeless connectivity" with the "anywhere, anytime" approach, intensifing their adoption . Notwithstanding the high level of adoption, Internet access still evaded some individuals, retaining some doubts about how this access gap will affect them in the coming years. Over the years, these gaps widened and led to the isolation and marginalization of individuals and communities, given the uneven access to ICT. The most common term to describe these ICT disparities is "digital divide", and although fairly old, the following definition provided by the Organization for Economic Co-operation and Development is still one of the most suitable to clarify this phenommena, defining it as "the gap between individuals, households, businesses and geographic areas at different socio-economic levels with regard both to their opportunities to access information and communication technologies and to their use of the Internet for a wide variety of activities" . ICT handling has evolved at an unprecedented speed, generating uneven spread of these technologies, depending on whether countries are industrialized or developing, widening digital disparities and attracting the stakeholder's attention . The United Nations states that individuals were falling behind due to the lack of access to technologies, missing opportunities and personal added value that the online world entails, so it has raised the bar for digitalization, trying to make ICT "universal" and "affordable" (United Nations 2016, p. --- 78). The definition and classification of the digital divide is volatile although this is not related to capriciousness or an inability to be precise. It is not a unique concept and it has varied because technology itself has changed considerably . The continuous evolution of ICT has given rise to different digital divides. Digital inequalities are gaining new forms, shifting from a matter of access to a matter of intensity, frequency, and aim of Internet use . The first-order digital divide, which represents the lack of adoption, the binary divide of "have" or "have not" access to ICT, has become outdated, patronized, and imprecise . The second-order digital divide, representing the ICT usage and the proficiency of ICT usage, has become relevant as a measure of digitalization, allowing the inter-country assessment . Van Dijk analyzed the technological differences and affirmed that in most developed countries the physical access divide is narrowing, leaving behind the digital divide confined to access metrics, alleging also that digital skills vary according to innovation speed of new technologies. In 2012, the same author remarked that there is a shift from the first-order digital divide that represents unequal access to ICT, to a second-order digital divide embracing inequalities of skills and usage. It is known that ICT access is on track, but not all of the problems are yet solved. On the contrary, the problems grow deeper . The second-order digital divide reveals itself to be one that is even more important -and more difficult to bridge -than the firstorder digital divide; separating those who are able to make good use of all of these technologies entailing opportunities and improving their standards of living, from those who are not . The emergent second-order digital divide is about capabilities and abilities, recognized as a knowledge divide, linked to fundamental skills that allow exploring all the benefits of the access. This raises a new challenge, a second form of inequality, that should be fully appreciated . People and organizations that use computers and Internet have a relative advantage, compared with others that do not . This digital competence of handling ICT is relevant inasmuch as it gives responses to other citizens satisfying their societal needs, strengthening socially, economically, culturally, and politically, and opening space for innovative content . Digital inequalities of first and second-order can be used as a measure for digitalization. These digital gaps can be analyzed, whether between nations or within nations . According to Çilan and Sadowski , the level of ICT access and usagefirst-order and second-order digital dividesvary among individuals within countries and digital discrepancies inside nations can justify the differences among them. Ritzhaupt considers the digital divide as a many-sided phenomenon crossing different digital extents such as access, usage, and skills as does Lindblom and Räsänen stating that "the digital divide at the individual level springs from many different sources" which emphasizes that individuals act differently in terms of digitalization. For instance, the most disadvantage individuals such as the elderly, less educated, women, and individuals with lower income widen the domestic digital disparities given their disadvantage in society . --- FIG. I - The scope of the study, crossing different technological divides --- DIGITAL DIVIDE'S DRIVERS Technology on its own, without any other catalyst, may drive deep divides within societies, rather than narrow socio-economic problems. It negatively impacts the sustainability of a society, hindering the competitiveness and the economic growth in the long term , and is vital to comprehend vulnerable groups' behavior, guiding them to the best opportunities toward digitalization, fostering economic and social progress. Van Deursen et al. mention that socio-demographic characteristics affect online behaviors, as does Yu , who explains the variability of ICT access through socio-economic conditions. To understand domestic susceptibilities and their implications in the contemporaneous society, four demographic segments were chosen: age, education, gender, and income. --- AGE Age is one of the most popular drivers when it comes to ICT adoption, with the United Nations stating that the digital divide is also an age matter. The generational rift is particularly important to understand the day-to-day appropriation of ICT by the younger and by the elders -the first group labeled as digital natives that were born and grew up already surrounded by ICT and the second group as digital immigrants that were not, and therefore had to adapt themselves to ICT . There are some previous evidences that this second group can present computer anxiety and technophobia given their age-specific characteristics and their non-intuitive ICT skills and usage, exhibiting social conditions such as: disabilities/chronic health problems, living alone, single or widowed, presenting "lower level of education than ICT users" . There are evidences that the online communication among the older group, people over 65 years, is significantly lower when comparing to the youngsters . It is also evident that "for seniors older than 70 years, the relation between age and Internet use seems not to be linear but rather exponential", likewise "with every additional year of age, the likelihood of Internet usage decreases by 8% in five-year differences when considering a range of 65-90 years or more" . Senior citizens are usually stereotyped as unable or reluctant. However, they simultaneously perceive that technology could enhance their self-sufficiency . By increasing ICT adoption and usage by this group, there would be less isolation via the social networks, strengthening relationships and more medical assistance via telemedicine, increasing social and health care . --- EDUCATION Educational digital divides persist in contemporaneous society affecting digital development, justified by the complexity of ICT, which is considered as "a major obstacle for adoption" . Van Deursen et al. agree that "education is an important predictor for the types of activities that people engage in online activities", which is in tune with Nishijima et al. who consider that the educational attainment affects access to ICT as it is "positively related to ICT usage in both developed and developing countries" explaining most of the disparities in personal skills for ICT utilization. For Kathuria and Oh is necessary a threshold level for education and skills to diffuse faster the ICT. For instance, the United Nations mentions that most websites are in English. For people with low education level, who are uncomfortable with the English language, it is a hindrance for e-inclusion. Rogers assumes that "the complexity of an innovation is negatively related to its rate of adoption". Hence, education can be the turning point for the digital divide, overcoming the ICT complexity, making the difference when individuals are faced with ICT, providing the opportunities to achieve digital dividends, prospects for future employment, making it a competition between traditional qualifications and technologies . --- GENDER The gender-related digital divide reflects gender related inequalities, in which "digital inequalities mirror structural social inequalities" . The main digital disparities are more visible in developing nations, where "the gender gap in accessibility and use of ICTs increases the existing gender inequalities" . This driver plays a role in drawing ICT asymmetries as some authors argue: "the phenomenon of technology itself cannot be fully understood without reference to gender" . Families, institutions, and religions shape gender roles and stereotypes, giving the idea that "computers are toys for boys not for girls", determining and framing conducts . --- INCOME The income level is considered by international organizations as a key component of digital . --- METHODOLOGY --- DATA To measure the digital divide between and within the 28 members of the EU, 14 variables were chosen from Eurostat's Digital Agenda Scoreboard Key Indicators for the year of 2016 having into account their relevance in the digital divide matter. The support and rationale behind each one is provided in Table 2. Each variable is disaggregated per each of the four drivers mentioned in the previous section, resulting in 11 sub-levels. These sub-levels are three for age: Generation Y , Generation X and Baby Boomers ; three for education: Low/No formal degree , Secondary degree , Tertiary degree ; two for gender: Female and Male; as for income three sub-segments were defined: 1 st quartile, 50% around the median and 4 th quartile. For Ireland, Italy, Malta, and United Kingdom, data compiled at income classes is not available. The three levels of education were defined using eight basic ISCED levels, as they allow the control of differences in education levels across countries, and are consistent with Eurostat which states that "at this level of aggregation data are considered as comparable over time for all available countries". As for income, using intranational quartiles, allows for the control of specific national differences like exchange rates, purchase power, ICT and other products and services costs, etc. Moreover, we have one group representing the ICT disadvantaged, one of the advantaged groups, and another which is hypothesized to be in the middle. This is consistent with age and education also, in which "Generation Baby Boomers" and the "Low/No formal degree" represent the ICT disadvantaged, whereas "Generation Y" and "Tertiary degree" represent the ICT advantaged. In this way, we can assess which sociodemographic characteristic is more relevant in drawing digital inequalities. To effectively measure the digital divide, variables such as the usage of ICT and the channel to access it must be acknowledged. Considerable lack of ICT usage continues to exist, given that "the lives of the majority of the world's people remain largely untouched by the digital revolution" . In developing countries mobile phones are the main form to access the Internet and according to Kathuria and Oh "mobiles have diffused the fastest among all ICT". Hence, the percentage of individuals that never used the Internet and the percentage using mobile phones to access the Internet were included. The frequency of usage is also appropriate for the study, affecting the ICT knowledge. Zillien and Hargittai The United Nations mentions that "efforts are being made to ensure privacy and security of personal data", pointing out that privacy is a contemporaneous and growing concern. Thus, a measure of privacy awareness was also included. Finally, despite all these technological progresses, one of OECD's first concerns related to ICT was how to foster the access and adoption of e-commerce among citizens . The single digital market taunted a shift from the traditional decentralized consumerism to a flexible, universal market in which tradinggoods, services or online contentis one click away, prevailing popular in the current society, converging to the single global market. For these reasons, two measures of electronic commerce were included, one regarding individuals who buy online and another for individuals who sell . --- DIGITAL DIVIDE DIMENSIONS Given the multidimensionality of the subject that mirrors the EU countries disaggregated by 11 sub-segments for 14 variables , a multivariate approach -factor analysis -is appropriate to find the underlying patterns and to deal with the complexity of this matter . As the successful use of this technique is grounded on the correlation between the variables, these were the first to be analyzed, concluding that a strong correlation structure appears to exist. This finding is supported by the Kaiser-Mayer-Olkin measure of sample adequacy -0.9 -which is, according to Sharma , "marvellous". The next step is to decide the "best" number of factors, taking into consideration the Kaiser's, Pearson's, and Scree Plot criteria. This choice resulted in a two-factor solution. As is recommended, to increase the interpretability of the factors, different rotations were used, starting with the two most popular and widely used in this context -Varimax and Quartimax. However, although most studies in this and other contexts use these orthogonal rotations, implying that factors will be completely uncorrelated, from a theoretical standpoint it does seem plausible to hypothesize that the different dimensions of the digital divide may be, at least to some extent, correlated. From the authors' understanding, countries performing very well on a specific dimension, should a priori, also perform well in others, and vice-versa. Orthogonal rotations do not allow this to happen, affecting "artificially" countries' performance, especially those that score very well in one dimension and, to keep the zero correlations between the factors, are prevented from performing well also in the other. This was, by the way, noticeable in the results using orthogonal rotations, especially in the Northern countries, which scored very well in one dimension and below-average in the other without a valid reason, meaning that they were poorly represented in the new axes. Hence, an oblique rotation, Promax was employed. If the hypothesis of the analysis is true, the results should lead to factors that have a significantly different from zero correlation, and since "oblique rotation allows for correlated factors whereas orthogonal rotation does not, it would appear that oblique rotation would be preferred to orthogonal rotation on this basis", notwithstanding, "orthogonal rotation is a subset of oblique rotations" . Factors represent the extent to which variables are correlated and an oblique rotation could allow seeing "the extent to which the factors themselves are intercorrelated" . According with Iacobucci , the correlation between the two factors was calculated, yielding a value of 0.63 , meaning that the data are adequate for this rotation, as it is neither below an absolute value of 0.3, which could mean that orthogonal rotation could be enough; nor above 0.7, which could indicate that many factors had been retained. The two dimensions are interpreted: one related to ICT applications and services, labeled as e-Services; and the other representing the social networks and the communication channels labeled as Social Networks, thereby responding to the first research question of this study . In the solution provided by Varimax, the two dimensions found explain 80% of the variance contained in the original 14 variables. The factor structure is similar between the orthogonal and oblique rotations. Note that the reference structure, i.e., the correlations between each factor and variable after the common variance with the other factor has been isolated, provided a clear differentiation between the variables highly correlated with the e-Services, from those with Social Networks. It is noticeable that, given the correlation of 0.63 and looking at the overall correlations, although the structure is still clear, some indicators are relatively well correlated with both dimensions. The factor scores of this analysis, i.e., the values that each country/segment has on each dimension, are examined below from different perspectives, to answer the second and third research questions of this study. --- ASSESSING THE IMPACT OF SOCIO-DEMOGRAPHIC CHARACTERISTICS Multivariate Analysis of Variance was chosen to assess the existence of different digital adoption levels among the 28 EU member states, when looking at different sociodemographic sub-segments. This analysis aims to compute the distances between the centroids of the subgroups, detecting if the distance between them is statistically significant . In MANOVA the null hypothesis assumes that all means are the same, testing the differences in two or more vectors of means . MANOVA tests for each socio-demographic segment , if the sub-segments share the same mean toward e-Services and Social Networks. Also, a post-hoc test was conducted to find which groups, specifically, have significant differences . --- FIG. II -Multiple comparisons of the means of the socio-demographic sub-groups Notes: *** The mean difference is statistically significant ; ** The mean difference is statistically significant ; * The mean difference is statistically significant According to Figure II, within the age segment there is statistical evidence that the Baby Boomers present different levels of adoption for both e-Services and Social Networks when compared to the two younger generations . As for education, all three sub-segments present different levels of adoption for both dimensions, as well as for income, except for the first quartile of income group and the 50 percent around the median group, who do not present statistically significant differences in the level of adoption of e-Services. Among the 28 EU countries, in a developed society such as Europe, gender does not present statistical evidence of different levels of adoption between men and women. MANOVA tests the differences of the means/levels of digital development, without testing if the digital disparities across the sub-segments of the same "driver" are similar. In other words, MANOVA cannot test which driver, if any, present higher digital divide, which is clearly of interest to know given the statement of Conover that "the variance of the population may be the quantity of interest". As for this study, the asymmetries of the four socio-demographic indicators will be assessed, identifying which of them engender bigger/smaller digital disparities in the EU, by using the abovementioned sub-segments . So, given the assumptions of the data, not following a normal distribution according to the Kolmogorov-Smirnov test, a nonparametric significance test was computed. Conover's Squared Rank Test for Variances was applied to evaluate if the digital asymmetries, i.e., the variance within the sociodemographic segments is different and, if so, identify in which of these segments there are significantly different disparities . Regarding Social Networks, age is unanimously the socio-demographic indicator that generates the biggest disparities among the subgroups when compared to the dispersion of other subgroups, such as for education, gender, and income. --- FIG. IV - The main socio-demographic indicators affecting domestic digital divide per EU countrymeasured by the standard deviation --- DISCUSSION --- DISCUSSION OF FINDINGS The EU member states have different backgrounds with diverse levels of expansion and standards of living influencing unequally the socio-demographic groups, and thereby disrupting the equal usage and adoption of ICT. When analyzing the four socio-demographic segments, starting by the age groups, the Baby Boomers are far below the arithmetic average adoption point for e-Services and Social Networks, presenting statistically significant differences against the youngest two generations at a confidence level of 99% . This matches with Quan-Haase's statement, affirming that Baby Boomers are still behind in terms of Internet access, digital know-how, and enthusiasm. Regarding education, the differences between the subgroups' average, for both dimensions, are statistically significant among themselves. Gender in EU28, in opposition to developing countries does not present significant differences in terms of digital adoption toward the two dimensions of the study. For income, on the other hand, individuals with higher income present significant differences against the other two groups of income, on e-Services. As for Social Networks, there is statistical evidence at a confidence level of 99% that all income categories have different levels of adoption. These differences in developing countries would be even more noticeable, as most European individuals, regardless their income sub-segment, have sufficiently high yields that allow ICT acquisition, which is in line with earlier research . Is noticeable that for the most proficient groupsgeneration Y, those with tertiary degrees, or in the 4 th quartile of incomethe mean differences toward the less proficient groups are statistically significant on both dimensions. High education and income levels are recognized as an "entrance ticket" to access and ICT usage, enabling individuals with solid know-how in technology . To identify which of the segmentsage, education, gender, or incomepresent significant asymmetriesaccording to the Conover's variance test there is at least one socio-demographic segment with asymmetries significantly different from the other three segments. For instance, on the e-Services dimension, among the four socio-demographic segment there is at least one pair of them that present unequal asymmetries, statistically significant with a probability of 90% to be true. On Social Networks, the corroboration of significant asymmetries achieves 99% of probability, revealing stronger evidence of disparities regarding Social Networks. For both dimensions , there are no statistically significant differences in digital asymmetries between each pair of segments, except for education and gender ; and income and gender . For Social Networks, age presents statistically significant differences toward all of the other three segments, displaying the largest disparities. Looking at the two dimensions found , education levels draw the most noticeable asymmetries for e-Services adoption, given that the most educated individuals deal better with ICT complexity, taking advantage of online services in their regular day to day activities compared to less educated individuals . For countries such as Denmark, Estonia, Finland, Germany, and Latvia e-Services adoption is more unequal among generations. Spain and Lithuania present more inequalities on both income and education. For Social Networks, it is unanimous among the 28 EU countries that age is the indicator that generates the greatest inequalities when compared to education, gender, and income, largely justified by the detachment between the youngest and the oldest. Coelho et al. enumerate several reasons that can lead to this gap between the elders and the Social Network Services , such as the apathy to ICT, prejudgment of sharing online their privacy, in opposition to the young adults that esteem the online network and the communication associated with it . --- LIMITATIONS The study analyzes the domestic digital divide for a specific point of time, considering data from 2016. Some changes might therefore have occurred and findings must be considered as a proxy to the present reality. Another limitation of this study is that are there are only 14 variables, which can be considered as too few, given the multidimensionality of the digital divide and the constant innovations in ICT. The income segment had missing data relative to income for four countries: Italy, Ireland, Malta, and United Kingdom, and for that reason this segment must be seen with caution. Nevertheless, the main goal of the study is to identify the main socio-demographic characteristics that cause greater disparities, despite the implicit level of adoption. There might exist low levels of dispersion between the subgroups of a socio-demographic indicator, which is positive. However, their digital level of adoption could be very low or very high, which is not considered. --- IMPLICATIONS --- THEORETICAL IMPLICATIONS OECD mentioned that "the digital divide among households appears to depend primarily on two variables, income and education", and was a matter of access that only the wealthier, therefore also more educated, could afford. Scrutinizing the divides of the 28 EU countries in 2016, namely: age, education, gender, and income, there are still some digital disparities that need to be closed. This study specifies the neediest groups with lack of adoption. For instance, e-Services adoption is more related to the education level of individuals, while Social Networks adoption is more determined by age. Income is no longer the main trigger for digital divide for the 28 EU countries when analyzing developed societies, such as those that belong to the EU. Digital divide results should be analyzed with caution, since it is a multidimensional phenomenon and different backgrounds ask for different measures, if the goal is to narrow digital discrepancies in an effective way. Regarding the multivariate technique that was applied, findings show that an oblique rotation when analyzing the digital divide through factor analysis better represents, the reality: e-Services dimension is correlated with Social Networks, sharing 36% of variance within each factor. --- MANAGERIAL IMPLICATIONS Europe is suffering from a growing professional ICT skills shortage and a digital literacy deficit. These failings are excluding many citizens from the digital society and economy and are holding back the large multiplier effect of ICT take-up to productivity growth. This requires a coordinated reaction with member states and other stakeholders at its center . It is necessary to flip the spotlights -EU strategies and efforts -according to the two dimensions, e-Service and Social Networks, ensuring different focuses consonant with the socio-demographic sub-segments. As ICT became a commodity, the effort to narrow within and between digital divides was through the implementation of initiatives that could bridge the generational gap in case of Social Networks and that could harmonize the educational level within countries, raising awareness of the benefits of the electronic services. Public authorities, private organizations, and individuals should make a personalized effort to narrow digital differences. If each EU country would analyze the within country digital divide by socio-demographic segments and act according to its necessities, the EU would benefit more as a whole, narrowing at the same time the international and domestic digital divide, improving digital literacy of Europe. --- CONCLUSIONS The European strategies and recommendations intend to narrow digital disparities, achieving similar conduct between and within countries, but there is still work to be done. This study analyzes the domestic digital divide for a specific point of time, considering data from 2016 relative to age, education, gender, and income, disaggregated by 11 sub-segments: three for age, education, and income; and two for gender. On the 14 variables of the study, two dimensions were identified to represent the digital divide, e-Services and Social Networks. There is statistical evidence that the level of adoption of the two ICT dimensions varies according to the sociodemographic sub-segments, within each socio-demographic segment, presenting a confidence level of 99% when computing multiple comparison. For e-Services, education presents the biggest disparities, and for Social Networks, age classes are clearly the most unequal. Nevertheless, there is statistical evidence that the segments present disparities between them, with a confidence level of 90% for e-Services and 99% for Social Networks, revealing greater evidence of inequalities in this dimension. When analyzing the countries' asymmetries, the main drivers for digital divide in e-Services are mainly education and age. As for Social Networks, age is unanimous as being the driver that affects this digital divide, for all countries.

This paper assesses the digital divide between and within the 28 member-states of the European Union. The analysis comprised four socio-demographic contexts: age, education, gender, and income. Because of the digital divide's complexity, a multivariate approach was applied -factor analysis with oblique rotation, which resulted in two distinct dimensions: e-Services and Social Networks. To test the significant differences of European Union positioning and European Union disparities, Multivariate Analysis of Variance and Squared Rank Test were computed. Findings show that e-Services adoption is influenced primarily by the education level of individuals, while Social Networks adoption is more affected by individuals' age.

Introduction Although religious beliefs are typically considered at the individual level , religion is fundamentally a social phenomenon. Sociologists long ago recognized not only that religion is social in function, but also that it is socially produced and reinforced. For instance, Simmel argued that the ''faith which has come to be regarded as the essential, the substance, of religion, is first a relation between individuals.'' Similarly, Durkheim emphasized that interaction in groups creates and reifies religion by fostering solidarity through the genesis of shared symbols. Historically, religion has been viewed as a creation of groups as well as a source of group -and thus interpersonal -connection. In contemporary terminology, this suggests that religion is a source of friendship selection leading to network homophily . That is, a means by which people come to know, affiliate, and become close to each other . Of course, the processes producing religion socially can also change participants' religion, whether renewing it or inspiring new faith. Ernst Troeltsch , for example, considered religious faith to be socially and contextually, rather than individually, produced. This theme is evident in much of Durkheim's work too. For instance, in his emphasis on how religious interactions generate symbols that preserve feelings of group solidarity, thereby transforming individuals by shaping cognitions and emotions, and stimulating the desire for more religious experiences . In other words, the religious content of people's social worlds can be a source of personal and possibly religious change. To the extent that religion is socially derived and embedded in social networks, the religious content of those networks may feedback to influence the affiliations, beliefs, and involvement of individuals . In this way, the religion of friends can influence that of individuals through socialization processes . In contrast to the focus on the social origins and influences of religion by early sociologists, however, most quantitative research on American religion is now based on surveys of unconnected individuals, with the result that the social context of religion is obscured . More than four decades ago, White criticized the implicit assumptions of the ''psychological consonance'' model that had become, and remains, so dominant in the sociology of religion. As White argued, the prevailing sociological approach to religion is ''peculiarly individualistic.'' We do not argue that all scholars of religion endorse this view so much as they have their hands tied by design and method , so that even as data collections and analytic tools have become more sophisticated, individualistic approaches to studying religion still dominate. In contrast to this individualistic emphasis, we concentrate on the friendship dynamics of adolescent religion. In doing so, we find that religious beliefs, activities, and affiliations are both a cause and consequence of friendships for adolescents who attend small schools. The findings are consistent with Collins' interaction ritual chain theory, which builds off of Durkheim and Goffman to argue that social interactions create powerful symbols -such as religion -that influence and organize subsequent interactions. The analytic approach we adopt uses innovative new social network analysis models that decompose social selection and socialization processes directly. Social selection reflects changes in social networks that result from the religious similarity among actual and potential friends, and socialization reflects religious changes in persons as a function of friends' religion . We use data from 7 to 12th grade adolescents from seven small K to 12th grade schools where friendship networks were collected over time to assess whether religion is uniquely friendship-inducing, a friend-based socializing factor, or if both processes are jointly operative. We build upon this analysis by quantifying the relative contributions of each to the network-religion autocorrelation, while also assessing whether the selection/socialization results derive from additional structural network and background factors. --- Religious individualism in adolescence Individualistic narratives are quite common in contemporary society, as Bellah and his coauthors' outlined in their influential book, Habits of the Heart . The account they present describes the contemporary American preoccupation with ''individual agency'' , even among regular churchgoers . In particular, American youth are thought to be highly individualistic in their religious beliefs and activities . Smith and Denton summarize the contemporary adolescent preoccupation with individualism: . . .nearly all American teenagers believe that they are not influenced by anything at all, religious or otherwise. Like most of the adults who have socialized them, teenagers take for granted an image of themselves as autonomous and self-defining individuals fully responsible for and capable of the formation of their own lives. Many teenagers actually bristle at the suggestion that they are directly influenced by people and institutions outside of themselves. It is during adolescence -when Americans are apparently most resistant to the idea that they are social beings, even while paradoxically entering what is generally the most intensely social phase of their lives -that we examine the social nature of religion. During adolescence youth expand the time they spend socializing with peers as they seek to forge their own futures . They interact together in various settings in and out of school, at church, youth groups, during sporting events, parties, etc., and so come to develop histories with each other that form and reform friendship networks, socializing each other in the process . Though research on adolescent social networks tends to focus on risky behaviors and delinquency , late adolescence is a time of substantial religious and friendship change , making this period, particularly when combined with the strong social and interactional foci provided jointly by school and church ), a key time to study the individual -and social -natures of religion . --- Friends and religion: Homophilous social selection vs. socialization Network autocorrelations reflecting the tendency for friends to be similar to each other in various ways is a normative feature of social relations expressing the fact that interpersonal associations are socially arranged . For the most part, Americans interact with people similar to themselves in terms of age, class, race, gender, and religion . In fact, Americans' social networks are predominantly composed of people with similar religious perspectives, affiliations, and levels of religious participation . This is especially true for women , conservative Protestants, Jews, and the religiously unaffiliated . Yet, the question remains: how does network-religion autocorrelation arise? There are two primary global mechanisms leading to cross-sectional network autocorrelation. The first mechanism, social selection, involves friendships forming among those who are similar to each other . Friendship selection processes have become a widespread concern in adolescent research , albeit one that is rarely but increasingly being studied directly . For our purposes, religion-based selection reflects changes in the ties between individuals comprising the network based on religion. The emphasis on homophilous social selection is evident in research in the sociology of religion. For instance, the church growth literature asserts that people prefer friends with religious beliefs and affiliations similar to their own . Similarly, the ''religious economies'' perspective argues that sociodemographic characteristics play a smaller role than ''homophily of preference'' in structuring religiously homophilous social networks . More recently, Vaisey and Lizardo suggest that worldviews are related to selection but not influence. Sherkat perhaps most clearly exemplifies the emphasis on homophily: ''People choose friends and spouses in accordance with [religious] preferences; hence, valued others are likely to reinforce existing desires rather than arouse new ones.'' Thus, our first research question is: Do adolescent friendships disproportionately form amongst those who are religiously similar to each other, and is this the primary mechanism of network-religion autocorrelation ? The second mechanism, socialization, posits that friends socialize each other and so become similar over time . This mechanism postulates changes in religion as friends adapt to and influence each other , possibly in order to find balance with the perspectives and beliefs of their friends , to impress them , or to gain social acceptance and status . Furthermore, as we elaborate below, socialization may also reflect the internalization of shared religious symbols that increase in emotional salience as a result of socializing together. The influential Lofland and Stark theory of religious conversion emphasizes the role friends play in bringing new members into the fold and is based on a version of a socialization hypothesis. Other research in the sociology of religion supports this perspective, suggesting that social networks are instrumental in introducing people to new religious groups and viewpoints . This view is also a reflection of the general concern in the risk behavior literature that peers negatively influence each other . Our second research question is thus: Do friends' socialize each other's religion over time, and is this the primary mechanism leading to network-religion autocorrelation? --- Friends and religion: homophilous social selection and socialization In contrast to the emphasis on either selection or socialization, it is possible, and perhaps likely, that both selection and socialization operate to produce religious-based network autocorrelation, leading to the question of the relative contribution of each. Collins' ) interaction ritual chain theory provides a framework for conceptualizing and expecting such joint social dynamics. In IRC theory , the ritual, often an everyday interaction between friends or acquaintances, ''is a mechanism of mutually focused emotion and attention producing a momentarily shared reality, which thereby generates solidarity and symbols of group membership.'' IRC theory posits a joint process of interpersonal engagement leading to observable patterns of social selection in friendship networks and peer socialization in religious participation and belief. Key to IRC theory is the idea that interactions are ''ritualized'' in the sense of Goffman so that actors have behavioral and emotional expectancies for their socializing. These interactions both create and sustain symbols that have meaning to the participants. Indeed, religious symbols and ritual styles organize interactions, and thereby form a basis for new and renewed interaction . The result is that even changes in friendship networks outside of religious institutions will reflect religion-based social selection. At the same time, the very experiences promoting group solidarity and vitalizing religion changes participants. When participants become caught up in a mutual focus of attention, they become ''emotionally entrained'' with one another, which heightens the salience of the social experience. Creating emotionally charged intersubjectivity is what religious services are designed to do , and the emotions evoked by religious symbols are qualitatively different from those evoked by other symbols . Similar processes can unfold between peers in friendly interactions, so individual religious views or engagement can change as friends socialize each other. Such interpersonal influences capitalize on communal symbols during interpersonal interactions and through the creation of shared emotionally charged interpersonal symbols created during socializing. IRC theory thus provides the foundation for understanding the microdynamic processes aggregating across individuals to produce homophilous selection in networks and socialization-induced patterns of individual religious change, which can also produce homophily when viewed cross-sectionally. Moreover, the theory is clear on a number of points. First, both processes should unfold concurrently since the two are inextricably interlinked. Social selection mechanisms lead people to form relationships with those to whom they have more emotionally entraining interactions, of which shared religious symbols are likely to be important. Collins also argues that we are ''emotional energy seekers,'' and so intersubjective shared realities experienced during interactions change people. Second, these processes should be evident outside of religious congregations to the degree that participating in religion creates powerful, socially shared symbols that can be reinforced and changed through interactions with others . Therefore, our third research question is: Do both friend selection and socialization work jointly to produce network-religion autocorrelation, and are the magnitudes of both processes comparable? --- Network and background factors Both network processes and background factors could result in spurious estimates for selection and socialization if not accounted for. Network mechanisms reflecting the fact that these processes are sources of change and stability may be particularly influential. For selection processes, accounting for triadic closure may be especially important . Consequently, we control for network closure as a source of friendship change. Network closure, in particular, appears to vary across religious traditions and is correlated with religious participation . In addition, we also explore the roles of popularity and activity on selection since religious youth may have more exclusive friendship groups , and as alternative mechanisms of individual religious change . Finally, because individuals have a profile of background characteristics that may jointly influence the role of religion in selection and socialization processes, we also account for several background factors capturing alternative selection and socialization mechanisms. For instance, previous research suggests that religious-based network homophily varies across religious traditions . Of course, the behavioral outcomes -measures of adolescents' religious participation and belief -should also vary across religious traditions . We include controls for parents' religiosity and education too since parents' religious beliefs and activities and social class are strongly associated with adolescents' religious perspectives and behaviors. Previous research also suggests that social networks and religious participation and belief are correlated with friends' having the same religious affiliation . These factors, and other sociodemographic background factors implicated in adolescent network processes , must be included in the models to ensure reasonable effect estimates assessing the central research questions around which the analysis is organized. --- Data and methods Data come from waves 1 and 2 of the in-home components of the National Longitudinal Study of Adolescent Health . Add Health is a stratified longitudinal study of 7-12th grade youth begun in 1994 with in-school questionnaires administered to approximately 90,000 students in 140 schools. A nationally representative sample of over 20,000 students was drawn from the in-school study and data were collected in-home in 1995 from both adolescents and parents. Another survey was administered again to the adolescents approximately one year later at Wave 2. This longitudinal sample consists of a core probability sample and special oversamples including 16 ''saturated'' school-settings where efforts were made to collect data on all attending 7-12th grade students so that a network sample could be maintained over time. Of these 16 schools, two were large and 14 were small . We used seven of the saturated settings, all K-12th grade schools that are relatively racially and ethnically homogeneous, to construct our sample. Our decision to use these schools was based on several criteria. First, because our analysis requires longitudinal measures of friendship networks, we were limited to the saturated schools. Second, one of the schools was a special education school and another six were 6-8th grade. We chose not to use these latter schools because the 8th graders moved into high schools for which full network data is not available. Third, larger schools capture different macrosettings than the small schools as indicated by the enormous size differences. The grade cohorts of the two big schools are larger than the entire 7-12th cohorts of the smaller schools. The result is that we focus on the social dynamics in a collection of smaller, more homogeneous settings. The joint sample size of the small schools comprising this study is 798 mostly white 7-12th grade students. The largest school contributed 163 students to the analysis and the smallest contributed 61. Three were public rural schools , the remaining four were private , three of which were urban . Network data was present for 70-89% of the students on the schoolprovided roster, rates that are acceptable for social network analysis . --- Measures --- Dependent and focal independent variables We employ six dependent variables that measure religious activity, belief, and affiliation/identification. 1 Religious service attendance is coded from the question, ''In the past 12 months, how often did you attend religious services?'' Youth service attendance is similar in that it asked about attendance at special activities for teenagers at churches, synagogues, and other places of worship. These two variables take on values 1 = never, 2 = less than once a month, 3 = once a month or more but less than once a week, and 4 = once a week or more. Importance of religion is assessed with responses to the question ''How important is religion to you?'' The importance of religion measure takes values from 1 = not important at all to 4 = very important. Frequency of prayer ranges from 1 = never to 5 = at least once a day. Religious identification is captured with two dichotomous variables: whether the respondent self-identifies as a born again Christian or reports no religious affiliation . 2 1 Although some researchers combine measures of religion into scales, such as public and private religiosity , we examine single-item indicators for three reasons. First, since this is the first analysis to simultaneously model selection and influence in the religious homogeneity of adolescents' social networks, we did not want to assume that selection and influence operate the same across different aspects of religion. Second, as recent research shows , individual attributes can impact different indicators of religion in unique ways, which can lead to misleading results if measures of religion are combined into scales. Third, the models we employ are designed to work with ordinal dependent variables making scales more complicated to use. 2 Due to an unfortunate skip pattern in Add Health, adolescents with no religious affiliation were not asked about their religious beliefs and activities. These unaffiliated respondents must be kept in the sample for all analyses to ensure proper specification of the network portion of the model . Consequently, we code unaffiliated respondents as never attending services or youth services, as not being born again, as placing no importance in religion, and as never praying. This coding most closely reflects what we know about unaffiliated adolescents. For instance, according to Wave 1 of the National Study of Youth and Religion, a nationally representative survey of adolescents ages 13-17, 94% of unaffiliated adolescents never attend religious services , only 13% of unaffiliated adolescents say religion is very or extremely important in daily life , and more than half of all unaffiliated adolescents never pray . The last dependent variable, the friendship network matrix, is used to map whom each adolescent views to be a friend over time. The network thus reflects the peers each adolescent views to be a close friend at each wave. This includes ''best friends,'' but is not limited to them since our definition of friendship captures individual views onto their network and not dyadic consensus reflecting reciprocal ties . The adolescent friendship network at each wave is constructed from two sets of variables requesting nominations of up to five male and five female friends from the school roster. The total sample makes use of all available nominations. --- Control variables For controls we include whether the respondent is female , grade , whether the youth is white , and whether the parent is single . Religion is also included in two ways. First, religious tradition is included with the following categories: evangelical protestant , mainline protestant, Catholic, other religious affiliation, and no religious affiliation. This scheme follows the denominational coding outlined by Steensland and colleagues , though we combine the Jewish and ''other'' religion categories due to small number of respondents in these groups. 3 Second, the Wave 1 parent interview is used to construct a standardized scale for parent religiosity for the responding caretaker coded from the following four items: religious service attendance over the prior year, importance of religion, frequency of prayer, and agreement with sacred scriptures of their religion. 4 The parent education of the responding parent is included as a five-value variable with categories ranging from 1 = did not graduate from high school to 5 = received postgraduate training. Finally, we include the number of off list nominations provided by the adolescent during the network portion of the survey. Although the majority of nominations in Add Health are to friends at school, close to 30% are not . In addition, we also include an indicator for whether the respondent was in the restricted nomination sample because some adolescents were allowed to nominate only one male and female friend due to a survey implementation error. The result of this error is that the full friendship network was not captured at the wave 1 in-home survey for 40% of the youth in the sample. We carried the wave 1 in-school nominations forward for these youth5 to preserve the full network so that we could conduct the longitudinal social network analysis. There will thus be greater change in the networks for the restricted nomination than regular sample, so we have constructed this indicator to reflect the fact that overall change in friendships will be greater for these adolescents. --- The model Studying religion-based network selection processes is complicated by the fact that the model must account for tie changes between individuals. Consequently, the model must consider both who is a friend with whom and who is not. Socialization processes reflect how individuals change in response to characteristics of their friends as well as the changing composition of friends over time as interpersonal interconnections change. Changes in behaviors can also feedback to influence the friendship network to the extent that selection is a salient interpersonal process. In order to address this complicated set of interlocking processes with friendships predicting changes in religion, and religion predicting changes in the friendship network, the analyses presented in this paper utilize the new class of Simulation Investigation for Empirical Network Analysis models developed by Snijders ) and colleagues . The model has a number of advantages over traditional analytic approaches . For instance, the model incorporates friendship preferences as well as structural network mechanisms, and direct information on friends in the network allows estimation of how friends influence each other . These models are unique because they are designed specifically to model tie changes and simultaneously link these alterations to changes in behavioral variables so that socialization effects ''control'' for selection, and vice versa . The parameters are estimated by constructing models decomposing the total amount of change in the networks and religion between observation moments into a series of smaller changes, called microsteps in the SIENA procedure. These microsteps reflect one change in either the interconnections or the religious behavior of a focal adolescent that together, across many microsteps, aggregate up to produce the total amount of observed change. In application, this means that the estimated coefficients capture changes in the logit of creating/keeping or terminating one tie in the network selection portion of the model, or the logit of a one-unit change in a religion measure. The sequence of these microsteps is designed to be a Markov process where changes in friendship and religion are linked together and modeled jointly. For more detailed and technical discussions see Snijders et al. and Steglich et al. . Friendship selection processes are studied in the network portion of the model since selection reflects changes in friendships over time that result from prior religious belief, activity, or affiliation, and from structural as well as other factors. This model component specifies the effects of network structure and adolescent's attributes on change probabilities in friendship status . Religious selection is operationalized with three parameters including the influence of religion on the number of friends chosen , the effect of religion on being chosen as a friend , and a dyadic religion similarity effect. Religion similarity ranges between 0 and 1 and expresses how similar the adolescent and their friend/ potential friend are to each other and is the key homophilous selection parameter under scrutiny. Friendship choices can depend on the configuration of the network more broadly, so a number of network structure effects capturing triadic network closure processes are also included , along with parameters for the control variables: the adolescent , potential friend , and potential friend and focal adolescent operationalizations . These effects are described in Table 1. The friend socialization process is captured in the religion dimension of the model since individual changes are motivated by friends' religion and other factors. This component models individual religion with functions of network statistics and the main effects of control variables in a way analogous to logit coefficients from ordinal logit models. The key socialization parameter, a network statistic, is the average religion similarity between the focal adolescent and their friends . As we indicate below, it is possible to include other network effects. However, those we explored using score tests were unrelated to changes in religion, and so have been omitted . Control effects include main effects of the background variables indicating increases/ decreases in religion, as well as the shape parameters, both linear and quadratic, describing the distribution of religion over time. These parameters are described in Table 1. --- Analysis The analysis uses the SIENA software to model friendship and religion changes in the joint combined social network of the schools. Because youth in different schools are unable to select each other as friends, out-of-school elements in the sociomatrices are fixed . All respondents were included in the analysis and were allowed to enter the study later or leave early using the composition change method of Huisman and 6 First, there is a full meta-analysis approach requiring estimation on each network separately. This approach is generally considered preferable because it allows parameters to differ across networks. There were estimation problems due to the small network sizes, model complexity, and limited observations over time, however, so we opted to use this simpler method. In other work with these schools, results have tended to be nearly identical whether network models are grouped as we have done here or the meta-analysis approach is used. The second approach treats schools as different time periods and so allows rate parameters to differ across schools while fixing the coefficients. Inferences were virtually identical to those reported here so we have used the joint network approach since doing so simplified other aspects of the project management. A positive effect implies generalized reciprocity while a negative effect with a positive transitive triplet effect suggests local hierarchies . b. This effect is an inverse effect of network closure so effects tend to be negative, suggesting that indirect connections tend to close through, e.g., the formation of transitive triplets, or else the indirect connections are lost. Snijders . Missing attribute and religion data were treated as non-informative following the method described by Huisman and Steglich . Parameters were tested using t-ratios of the coefficient estimate divided by standard error based on findings indicating that the distribution follows an approximately standard normal distribution . Additional parameters that were tested but not included in the analysis are also presented at the bottom of Table 1. Score tests were used to determine if these parameters improved the model performance against a baseline model including the network structure effects and religion influence and selection parameters . Because these parameters did not improve the model performance, they are not included in the model series we present. Score tests were also used to simplify the model structure with respect to the control variables. Ego, alter, and similarity parameters were omitted from the model specification when they were not statistically associated with half of the outcomes to maintain a consistent model structure across behavioral and network processes. Finally, the contribution of the different processes to the autocorrelation between the friendship network and the religion outcomes is decomposed by the method described in Steglich et al. . The spatial network-religion autocorrelation is calculated using Moran's I across a special model series disaggregating the contributions of different mechanisms. In this way, religion similarity is decomposed into the proportionate contributions of selection, socialization, alternative selection and influence from the other control variables and structural network effects , and general trend effects in friendships and individual religion. --- Results --- Descriptive statistics Descriptive statistics for the religion outcomes at both waves are presented in Table 2. Average service attendance is 3 on a scale of 1-4 with approximately equal proportions increasing or decreasing their scores over time . In addition, the average similarity between friends is over .7, indicating that friends are on average about 70% similar to each other. In general, average similarity at wave 1 is high for all of the outcomes. Youth service attendance is the lowest at .69 and no religious affiliation is highest at .84. These results indicate that friends' religious homophily is already substantial by the later grades for youth who attend small schools. In addition, 30-45% of youth increase/decrease their participation or religiosity over the study period, while much smaller proportions change their affiliation and/or born again status . Although there is substantial pre-existing similarity, the considerable change in participation, importance of religion, and prayer indicates that adolescent religion is not fixed over this time period. The amount of change in the identification variables is much smaller and the total proportion of ''nones'' is also low, which suggests that there is less information for identifying social dynamics of religious identification than for participation, importance of religion, and prayer. Descriptive network and covariate statistics are shown in Table 3. On average, youth nominated 2.4 offlist friends, while sending and receiving approximately 3.3 friend nominations, about 1.2 of which were reciprocated. Moreover, adolescents were in nearly 3.4 transitive triplets, two 3-cycles, and were connected to seven other students through one intermediary. The sample is almost entirely white and only 22% reside with a single or divorced parent. Half of a. These columns present the proportion increasing or decreasing their religion between waves. b. This column presents the combined proportion increasing/decreasing their religion between waves. the sample is affiliated with an evangelical denomination, 23% are affiliated with a mainline denomination, and 11% are Catholic. Finally, the average responding parent in the sample had between a high school degree and some college attendance. --- Model results We present coefficients and standard errors for the service attendance model series in Table 4 in the logit metric. The first service attendance model, denoted S1, includes the religion and structural parameters, in addition to the off list and restricted nomination sample controls. Because these models are complicated and not familiar to most researchers, we begin our interpretation of the model with a discussion of the structural, rate, and shape parameters. The outdegree parameter is negative because the network density is low given that the average adolescent nominated fewer than four friends who were also present in the network. At the same time, the reciprocity parameter expresses the tendency for friendships to be reciprocated . The transitive triplets, 3-cycles, and number distance = 2 are three ways to express network closure. In this case, friendships tend to form among the friends of friends' to the effect that the odds of a friendship that closes a transitive triplet are larger by a factor of =) 1.55, while friendships producing the other two structural patterns are less likely as indicated by the negative coefficients. This suggests that there are local hierarchies so that some youth are more popular in their local network neighborhood than others, as implied by the combination of the positive transitive triplet and negative 3-cycles effect . Moreover, distance = 2 connections tend to close over time or to be removed. These factors thus drive friendship change and consistency through structural closure processes. Before interpreting the focal parameters, we turn briefly to the rate parameters to clarify them, although we do not view them as substantively meaningful. Rate parameters capture the change opportunities in either the selection and socialization models at each microstep of the routine. Thus, there were an average of 12.7 friendship change opportunities, which were .44 higher for the restricted nomination sample reflecting the longer time-lag between networks for this group, and three opportunities to change service attendance. We do not view these as substantive parameters since they reflect the number of microsteps needed for the data-constrained simulation portion of the estimation algorithm to reproduce the total amount of change in the network and religions variable. The shape parameters capture the distribution of service attendance so the other parameters in this part of the model reflect movement along this distribution. The fact that the linear and quadratic parameters are both positive indicates a shape function with small frequencies at low values and high frequencies at high values. In other words, the frequencies increase at an increasing rate, which is consistent with the descriptive statistics showing high levels of service attendance among these adolescents. Finally, the service attendance parameters in the network selection model indicate a tendency for more religious youth to be more popular , that there is no difference in the tendency to nominate others , and, most importantly, that there is homophilous social selection. The odds of having a friendship relative to not having one are approximately = 1.38) 38% larger among perfectly similar compared to maximally dissimilar youth , all else equal. Moreover, average friend similarity is large and positive in the service attendance model, which shows that individual service attendance changes to become similar to that of friends, a form of social influence we have referred to as socialization. The odds of increasing service attendance is =) 6.5 times larger for an adolescent maximally similar to their friends than for one maximally dissimilar, but it is important to note that full coverage over that range is not realistic, particularly when considered in light of the fact that selection is operative and that pre-existing similarities are large. More realistically, each average similarity difference of .1 relative to friends' raises the odds that attendance increases by = 1.21) 21%. These results are thus consistent with the idea that both selection and socialization processes take place simultaneously so that changes in religious attendance is responsive to that of friends', even while it forms a basis for friendship. The remaining models, S2-S5, build the full model by including the background factors in groups. Although the magnitude of the homophilous selection parameter shrinks by Model S5 when parent religiosity is included, service attendance selection remains an important process in the model. The same holds true for socialization, which also remains large and significant across models. The other parameters in the model indicate that friendships tend to form among those of the same gender, same grade, similar parent educational backgrounds, the same family structure, and the same religious tradition. Moreover, whites reported more friends than minorities in these schools and were more likely to be friends . Notably, youth whose parents were religiously similar were more likely to be friends as well. Results for the service attendance portion of the model further indicate higher levels of attendance among those with more educated and religious parents. Notably, the no affiliation effect is significant. This captures a regression to the mean -since the ''nones'' had very low involvement at wave 1, a few increased their involvement, leading to the counter-intuitive estimate. This finding shows a few times over the course of the analysis. Overall, comparing effect magnitudes for the service attendance selection and influence processes is complicated by the meaning of the similarity measures used and the fact that the coefficients are derived from two different probability models. In order to facilitate comparisons, we have decomposed the network-behavior autocorrelations into key model components in Table 5. 7 Results are presented for each of the outcomes, so we will refer back to this table to supplement the additional outcome-specific results . The network-service attendance autocorrelation is observed at .41 and the estimate, at .39, is very close to this value in Model S5. Over 64% of this reflects the ''trend'' of existing homophilous friendships and the consistent pattern of attendance observed among these youth. Approximately 9.2% results from selection, similar to the amount due to the other background factors, including the influence of parent religiosity, and nearly 20% from socialization. These findings suggest that the influence of friends plays a larger role than friendship preferences. A subset of results for the remaining outcomes is presented in Table 6. Two models for each outcome are reported. The first corresponds to model S1 in Table 4, and the second corresponds to S5. Only focal parameters in the network portion of the model are included because the other coefficients are generally similar to those reported in Table 4. Youth service attendance shows stronger signs of both selection and socialization than service attendance does, as indicated by the larger coefficients. Indeed, both key parameters are significant in Y1 and Y2 , once again suggesting the presence of IRC processes. The autocorrelation results in Table 5 corroborate and elaborate this. Approximately 35% and 27% of the .38 autocorrelation results from socialization and selection, respectively. Moreover, the proportion due to trend effects is much smaller because youth service attendance is more socially responsive than regular service attendance. In Y2 the alter effect is also marginally significant, suggesting that more involved youth are also 7 These are estimates of the decomposition and there is uncertainty in these estimates that is not quantified. This uncertainty reflects other factors such as the uncertainty in the parameter estimates themselves. Moreover, as one anonymous Reviewer pointed out, it remains possible that the proportionate contributions do not fully capture the endogenous process resulting in some misattribution of the contributions made by both selection and socialization because of the narrow time frame that the study covers. more popular, all else considered. With respect to youth service attendance, females decreased their involvement relative to males over time. Surprisingly, and in contrast to regular service attendance, parent religiosity was unrelated to youth service participation. Note too that the negative linear and positive quadratic slope for the shape parameters indicates a bi-modal distribution with groups of youth at opposite ends of the distribution. Importance of religion is also a dual process combining elements of social selection and socialization in both the simpler and more complicated model . Interestingly, youth for whom religion has greater importance both receive fewer friendship nominations and report fewer , which, when considered in conjunction with the selectivity parameter, suggests that they are more exclusive in who they consider friends. As shown in Table 5, although trend effects indicating some state dependencies are important , nearly 25% and 17% of the .42 autocorrelation reflect socialization and selection, respectively. Importance increased for those with single parents and no affiliation over time , and among those with more religious parents . Results are somewhat similar for frequency of prayer, at least in terms of support for the idea that prayer is both a source of friendship selection while also being responsive to friends' prayer with both contributing about 15% to the estimated autocorrelation. As with importance, youth who pray more are also more socially exclusive, both in nomination receipt and in who they view as friends. Surprisingly, parent religiosity is unrelated to changes in prayer. At the same time, females, youth with more educated parents, those with single parents, and the unaffiliated all increased their frequency of prayer. The final set of outcomes captures religious identification. Despite the low numbers of religious ''nones,'' and the small amount of total change in this variable between waves, model N2 reports selection and socialization effects with each accounting for over 20% of the .25 autocorrelation . Whereas those who pray frequently and place a great importance in religion were less popular , Model N2 shows that those with no religious affiliation are more popular . Of the background factors, only parent religion predicted no religion: youth with more religious parents were less likely to be a ''none.'' Identifying as a ''born again'' Christian is also related to friendship selection , while also being responsive to whether or not friends identify as born again . Although born again youth are more popular , they report fewer friends . When combined with the selectivity parameter, this suggests, once more, that more religious youth are more cliquish. Mainline and Catholic youth were substantially less likely than evangelical Protestant youth to report being born again, as were those with no and other religions, though these latter two effects were not statistically significant. --- Discussion and conclusions Sociologists have long been interested in the social nature of religion, but directly incorporating different social and contextual processes into the study of religion has proven challenging. We sought to remedy this situation by examining the social nature of religion with the adoption of a dynamic longitudinal social network analysis framework. In so doing, this study offers a number of innovations and unique insights into the sociological study of religion while also extending the adolescent research literature and providing a broad empirical assessment of a key expectation derived from interaction ritual chain theory . When viewed cross-sectionally, the youth in this study prefer friendships to those who are religiously similar. Even so, religious participation, devotion, and identification changed for many, and these changes were systematically related to changes in the friendship network. We have shown that religion, whether measured as participation, devotion, or identification, is pervasively social among adolescents who attend small schools. That is, religion is a source of social attraction influencing who spends time with whom, and is thus a source of the religious homophily that has been noted by others . At the same time, however, individuals' religious participation, devotion, and identification also respond to that of friends', which shows that how people engage with religion is subject to the same social forces as other behaviors and preferences, such as alcohol use and music tastes . In fact, the results show that socialization plays a somewhat larger role than friend selection in explaining network autocorrelation for four of the six measures of religiosity. Overall, then, these results suggest that both processes captured in our first two research questions play an important role among adolescents who attend small schools, but that of the two socialization appears to be a slightly stronger force than selection since we estimate that it accounted for 25% vs. 17% of the autocorrelation, on average. Given the salience of these two processes, and the fact that neither is predominantly important, these results are consistent with the idea that religion and networks mutually influence one another. Overall, we interpret this to be more consistent with IRC theory expectations than the traditional frameworks motivating the first two mechanismspecific research questions. Although scholars have recently begun incorporating social contexts into the study of religion using multilevel approaches to operationalize group-level effects , these studies have a number of limitations that we have begun to address here. First, they tend to focus on congregations as the relevant religious contexts. We have shown that religion can be important and pervasive in other social contexts too. Religious contexts are not selfcontained, and they clearly spill-out to influence social processes in other settings. Collins' ) framework suggested that this would be the case because the powerful symbols religious participation generates create a foundation for interpersonal interaction, which renew old and generate new symbols, and thus lay the foundation for interpersonal influence and changes in religion. Second, group-level effects as typically operationalized and modeled take the whole group to be a monolithic entity in which the interpersonal processes unfolding within them are invisible to the analyst. These processes are undoubtedly important as adolescents compete for standing and acceptance in peer crowds that extend beyond the local friendship groups studied here , and even into the broader adolescent marketplace for relationships . While group effects matter for religion , there is also a great deal of social activity within larger groups . Religion, among the adolescents comprising this study, was an important part of those dynamics. Religion is both consequential for structuring the group itself and is differentially distributed based upon the sub-group friendship structure through interpersonal friend socialization processes. This demonstrates that the difficulties of incorporating social contexts into studies of religion are exacerbated by the fact that individuals influence their environments , so that the flow of influence between ''group'' and ''individual'' levels is nonrecursive. Third, some years ago White critically noted that a ''psychological consonance model'' had become dominant in studies of religion, and we believe it still remains commonplace, albeit often implicitly. According to White , this model assumes ''that theology is the primary source of religious behavior.'' Even so, motivations for studying aspects of the interpersonal dynamics of religion have long been evident in the sociology of religion , though the theoretical focus tends to be on only one of our equations, namely the role of selection or socialization. In contrast, our results suggest that a broader conception of interpersonal social process and religion may be warranted. That is, there is ''jointness'' in the friendship selection and socialization processes leading to social congruence. The role of theology in predicting religious behavior may in fact be less central than has been assumed. Thus, a broader integration of Collins' IRC theory into the study of religion, and more broadly in adolescence since the social forces at play in religion may be similar to those for risky behaviors , should prove fruitful. The approach we adopted is inferentially stronger than others since we measured the network directly, and thus the behaviors of friends were measured directly. Moreover, the SIENA models we employed allowed for both selection and socialization processes to be assessed longitudinally and concurrently. Most research on religion and social networks is handicapped by reliance on the information that respondents provide about their friends' religion, but not their friends' actual religion, or the religion of their potential friends in their personal networks . This information on potential friends was an integral portion of the analysis allowing us to clearly identify how religion is incorporated into selection processes. Consequently, change in our models was not biased by either inaccurate cognitions that arise when individuals impute the behaviors of others , or the ''reflection problem'' that is inherent in multilevel analyses because reference groups cannot be identified . Just as there are strengths to this analysis, there are also a number of important limitations that circumscribe the generality of our findings. First, this study covers only a 1-year period over adolescence. While studies among adults suggest similar social dynamics of mood across the life-course , more studies over longer periods and age ranges are needed. The result of this limitation, however, is that some youth may already be going through transitions in both friendships and religion, and it is not entirely clear if two waves are sufficient to completely disentangle this process. Second, the schools in our sample do not comprise a random sample, and so generalizability, as with most social network analysis studies, is suspect. Third, while we have limited our analysis to friends that adolescents feel closest to, social influences, and possibly even selectivity, may extend out further into the network. This could reflect the desire to participate in crowds like ''jocks'' or ''geeks'' , indirect connections to other youth whom adolescents share friends with , or romantic partners and their friends . That is, we have only captured a narrow -albeit important slice of adolescent social life. Fourth, our analysis is limited to adolescents in small schools, leading to concerns that social processes may differ substantially across settings. Indeed, there are several reasons to believe that the social dynamics of religion may differ by school size. For instance, youth in small schools have more knowledge of each other and they are more connected to each other and to their schools , just as attendees of smaller congregations are more likely to know one another . In addition, there is evidence that network processes vary across schools . School size has a large impact on administrative factors and the structure of the curriculum , which, via course selection, can constrain friendship opportunities in larger schools and thus the knowledge that students have about each other, such as their religious affiliation and involvement. There will also be more opportunities for shared religious experience among adolescents in smaller settings to the extent youth may be more likely to attend the same religious congregations. As a result of these factors, we have focused on this set of smaller schools as the first step in a larger research agenda. Studying broader sets of schools across a deeper set of more heterogeneous settings is an important goal to be addressed in future research. Even with these restrictions, however, the findings we present here are highly relevant and show that, at least in some places in contemporary America -about 770 k students attended small, K-12th schools over the 2009/ 2010 school year -religion among adolescents is very socially dynamic. Future research will shed further light on the extensiveness of these processes across settings, but as we have shown, models of individual religious change should strive to incorporate a richer interpersonal backdrop. This is, of course, not a new proposition. Sociologists since Durkheim, Weber, Simmel, and Troeltsch have argued that religion is social in nature. In this article, we offer new model and empirical evidence to support this proposition that religion is a social phenomenon, both in the sense that religious similarity promotes social connections, and in the way friends influence each other's religious participation, devotion, and identification.

Longitudinal social network data on adolescents in seven schools are analyzed to reach a new understanding about how the personal and interpersonal social dimensions of adolescent religion intertwine together in small school settings. We primarily address two issues relevant to the sociology of religion and sociology in general: (1) social selection as a source of religious homophily and (2) friend socialization of religion. Analysis results are consistent with Collins' interaction ritual chain theory, which stresses the social dimensions of religion, since network-religion autocorrelations are relatively substantial in magnitude and both selection and socialization mechanisms play key roles in generating them. Results suggest that socialization plays a stronger role than social selection in four of six religious outcomes, and that more religious youth are more cliquish. Implications for our understanding of the social context of religion, religious homophily, and the ways we model religious influence, as well as limitations and considerations for future research, are discussed.

INTRODUCTION The past 20 years have seen dramatic rises in suicide rates in the United States and other countries around the world . These trends have been identified as a public health crisis in urgent need of new solutions and have spurred significant research efforts to improve our understanding of suicide and strategies to prevent it . Unfortunately, despite making significant contributions to the founding of suicidology -through Emile Durkheim's ) classic Suicide -sociology's role has been less prominent in contemporary efforts to address these tragic trends, though as we will show, sociological theories offer great promise for advancing our understanding of suicide and improving the efficacy of suicide prevention. Here, we review sociological theories of suicide with the explicit goal of building bridges. We begin where all sociologists must: with Durkheim. However, we offer a more comprehensive understanding of Durkheim's insights into suicide than prior reviews provided by non-sociologists . This is critical. Much of the nuance and richness of Durkheim's insights have been lost in modern suicidology, and yet Durkheim is foundational to understanding sociological theories of suicide, as well as understanding the potential of sociology for suicidology. We also discuss limitations in the Durkheimian approach and how more recent efforts have not only addressed those concerns but have done so by bringing sociology's broader theoretical and empirical toolkit to bare on suicide. These insights draw largely from social network theories, cultural sociology, sociology of emotions, and sociological social psychology. We conclude by making explicit bridges between sociological and psychological theories of suicide and by noting important limitations in knowledge about suicideparticularly regarding the roles of organizations, inequality, and intersectionality -that sociological scholarship is uniquely prepared to address. --- Durkheim Explained The sociological study of suicide remains rooted in founder Émile Durkheim's ) empirical study of suicide, still the disciplines' greatest contribution to suicidology . Durkheim's theory posits two core principles: that the structure of suicide rates is a positive function of the structure of a group or class of people's social relationships and those that social relationships vary according to their level of integration and regulation. Though Durkheim never clearly defined his dimensions, sociologists have generally treated integration as the structural elements of social relationships like the number and density of ties and regulation as the degree to which a collective's moral order controls and coordinates its member's attitudes and behaviors . Additionally, Durkheim articulated two continua and four types of suicide related to integration and regulation: egoistic/altruistic suicides and anomic/fatalistic suicides . Importantly, Durkheim was not interested in the subjective appraisals suicide decedents provided for why they chose suicide, but rather saw suicide, like alcohol abuse or homicide, as a symptom of collective breakdown of society. In turn, rather than focus interventions to reduce suicide on individuals, he argued [like many population health scientists today ] that a more efficacious avenue to protect individual well-being lies in collective public projects to produce protective structural changes. These changes can restore the integrative and regulative functions of the social groups to which individuals belong or lessen the intense pressure on individuals in social groups where integration and regulation have exceeded "healthy" levels. Durkheim was writing at a time of immense political, economic, and cultural change, which in turn motivated his emphasis on the types of suicide predicated on too little integration or regulation over the dangers of too much. Consequently, empirical research examining when and why connectedness or moral clarity might prove fatal to a group's members was sidelined until rather recently; a point we will return to below. --- Integration and Suicide Of the two social factors, Durkheim's integration has had the most profound impact on both sociology and suicidology. In explaining the power of integration, Durkheim argued that the more extensive and denser a collective's social relationshipsi.e., the more integrated the collective -the more enmeshed individual group members become, and, therefore, the more meaning and purpose individuals feel about their lives. He remarked, "The bond that unites [individuals] with the [group] attaches them to life [and] prevents their feeling personal troubles so deeply ." He continues that suffering physically, psychologically, or spiritually, "does not exist for the believer firm in his faith or the man strongly bound by ties of domestic or political society" . This collective belonging protects individuals from what Durkheim termed "egoistic" suicide, or suicides resulting from isolation and a lack of collective belonging. Integration, then, is borne of the recurring social relationships that require tending and care, and which are embedded in larger networks that form groups, communities, or perhaps, even nation-states. This includes being tied to families and neighborhoods as well as communities . These relationships provide members with what sociologists call social capital, or tangible and intangible benefits built on membership . In recent theories of suicidology, integration has been operationalized through perceptions about belongingness and connectedness . However, Durkheim was not interested in perceptions or appraisals, which he argued were subjective. Instead, integration is meant to be a characteristic of the group, not of individuals . Regardless, Durkheim's basic premise -that being highly integrated is protective against suicide -has received consistent strong empirical support across time and space and disciplinary boundaries . Conversely, the flipside of egoistic suicide -suicides caused by too much integration or altruistic suicide -has received scant theoretical and empirical attention . In Durkheim's estimation, tight-knit societies could rob individuals of their ability to make decisions under certain conditions, leading to suicides for the "good of the group." He pointed, for instance, to Hindu Sati, a rare form of suicide in which Hindu widows are compelled to throw themselves on their husband's funeral pyre . Though Durkheim thought over-integrated suicides relics of earlier forms of society, Abrutyn and Mueller have argued they are more common than we think. Pointing to the literature on social capital and on suicide clusters , they argue that in the meso-level of society, we can find numerous examples of communities where social structure can be exceedingly dense, like some religious communities , high schools and neighborhoods , army bases, and institutions like prisons or psych wards . Indeed, many of these places are disproportionately vulnerable to the emergence of suicide clusters . This highlights potential downsides to connectedness, such as groupthink or high costs for non-conformity and cautions scholars from positing connectedness as a purely protective phenomena. --- Regulation and Suicide Durkheim also argued that suicide rates were related to the degree to which a given group's rules and social norms were consensually clear, coherent, and shared. Living in a poorly regulated society or social group resulted in what Durkheim termed "anomic" suicides. In essence, Durkheim posited that humans, as animals, were not inherently moral creatures, but had to acquire morality from without. Notably, "moral" was synonymous with "social" in Durkheim's day, and thus he saw social bonds as having integrative features like intimacy and regulative features like moral obligations and expectations. Thus, Durkheim set up several routes to de-regulation causing suicide. First, societies where norms were constantly changing and or where there was a general breakdown in moral clarity, people's ability to easily identify their purpose would be constantly under attack. Second, regulation could suddenly be weakened, either by a change of status in the individual or by a collective crisis that challenged society's ability to provide clear moral or social guidance. In short, Durkheim saw a sense of shared moral clarity as an independent force providing protection to members of a group. While Durkheim emphasized the societal level, it is important to note that we can also develop moral relationships with a group and an abstract system of norms , which expands the "web" in which a given person may find themselves protected. Like integration, too much regulation may also cause what Durkheim termed "fatalistic" suicide. For Durkheim, fatalistic suicides occurred when members of a group or social category were subjected to intense psychic and physical coercion such that there was no hope for a future without suffering. Though Baumeister has argued, suicide is very often about escape from pain, like other Durkheimian types, fatalistic suicides refer to a class of suicides that are not limited by specific individual motives. To date, few studies have explicitly explored Durkheim's fatalistic suicide, though we can provide some examples of its possible research potential. First, structural inequality or violent oppression within families or communities may render groups of oppressed individuals disproportionately vulnerable to suicides. For example, we know that women in violent relationships often feel trapped and over-regulated ; and are more susceptible to suicidality . Women in rural China or Iran, for instance , may also fit this pattern, as may women of color who emigrate to another country and find themselves in precarious employment situations . Second, suicide bombers are often over-regulated , which may produce the type of structure that delimits options for resisting and expressing one's obligations to their community . --- Durkheim's Limitations Despite the importance of Durkheim's theory to suicidology generally, and sociology of suicide more specifically, Durkheim's theory is not without limitations which have in turn shaped more contemporary sociological theories of suicide. One of the oldest and most notable limitations of Durkheim is methodological. Durkheim fails to adequately address the ecological fallacy of studying suicide rates to understand individual behavior. Durkheim forcefully argued that societal-or macro-level forces caused individual-level behavior , and yet the link between societal-level social forces and individual behavior is challenging yet crucial to document. Compounding Durkheim's methodological limitations was the intellectual climate of his day. As a nascent discipline, Durkheim worked hard to distinguish and legitimize sociology apart from psychology and anthropology. Hence, using social psychological or cultural ideas -two sets of phenomena associated, respectively, with the other disciplines -was impossible. He could not, for instance, think about identity or emotions in sociological terms and, therefore, could not bring sociology into the micro-level of social reality. As we shall see, this limitation, as well as Durkheim's explicit rejection of Gabriel Tarde's imitation theory , has also constrained contemporary sociologists, until rather recently, from thinking about how suicide may spread from one person to another. Finally, Durkheim's own lack of attention to power and inequality, and the legacy it has generated, represents a major limitation. Though Durkheim sees regulation as comforting and supportive, there is a line between moral authority being an anchor in a chaotic storm and it being a source of domination and oppression. This line, as we shall see, has obscured the role inequality, stratification, and oppression play in suicidality. In short, regardless of the importance of Durkheim's basic insights, they fall short of helping us understand why a particular person dies by suicide and the mechanisms through which external social forces get inside someone's psyche generating pain and rendering them vulnerable to suicide. In this next section, we map sociological advances in understanding suicide by focusing on the new structural and then cultural/social psychological approaches that have emerged over the last two or three decades. To be sure, Durkheim's approach continues to loom large over sociology, with a recent review lamenting the sheer lack of new approaches to the sociology of suicide , and thus while we highlight all major scholarship and theoretical contributions as possible, the basic dearth in research programs or teams is a more general limitation of the sociology of suicide. Like Durkheim, these theoretical and methodological projects build on the idea that there are emergent, distinct properties that are not reducible to the individual and her perceptions or decision-making. Yet, they do not deny the importance of intra-personal factors, instead they seek to supplement them. Collectively, these advances have great significance for general theories of suicide and for suicide prevention. --- Structural Insights One of the first big innovations to Durkheim's macro theory was to incorporate advances from structural sociology -and namely insights from social network theories -to elaborate how social integration and regulation matter to suicide. Social structure is a notably elusive concept, but it usually refers to sets of stable social arrangements that evince certain properties regardless of the specific incumbents. Social structures deeply shape individual life chances by sorting us into particular opportunities, experiences, subcultures, social roles and obligations. They can be both easy to measure, as in the neighborhoods we live in or the schools we attend, or complex and intangible. Network theories facilitate the identification of local social structures that are salient to the individual and more closely capture the reality of the social world that surrounds them in their daily lives . One of the greatest advances in sociology of suicide is the social network elaboration of Durkheim's theory. This approach allows for greater specificity of social structures and crossfertilization with contemporary social theory. With Durkheim's "societies" translated into the operation of different networks, solidarity comes from the presence of strong, interlocking social relationships. The power of the external social world is preserved, while situating the individual more realistically in it. Another advantage of a network approach is that it avoids the overly optimistic view of personal ties as always protective. Indeed, a plethora of work within the social network perspective has long demonstrated that the presence of negative ties is potentially more powerful in affecting individual well-being than positiveties . Perhaps most importantly, a network approach highlights how integration and regulation coexist and in fact likely co-determine place-based vulnerability to suicide. An idea that is contrary to Durkheim's four distinct "ideal-types" of suicide . Instead, scholars advanced a curvilinear theoretical predictive plane with four dangerous poles matching Durkheim's types . One dimension, running from left to right, represents integration. Another dimension, running from back to front, represents regulation. Both dimensions run from high to low, and their interaction generates the four types of suicide. When individuals live in social structures characterized by too little integration or regulation, the threads of the social safety net are too far apart to catch them when crises destabilize their equilibrium. Egoistic and anomic suicides are then theorized as "diseases of the infinite" because of the extreme gaps in the societal safety net that normally support individuals during times of individual or community crisis. Conversely, the social safety net closes up when social structures are overregulated or over-integrated. With no flexibility or give in the safety net, individuals who experience crises hit a wall that shatters rather than supports. It is in the center of the net, where ties are moderately integrated and regulated that individuals can be safely caught and restrained from their suicidal impulses ). A recent study illustrates the power of this structural approach. Using novel US data connecting the social profile of individuals to the social profile of the communities where they live, the study draws from social network theory's principles of selective attachment or homophily and differential association . Specifically, researchers examined whether the presence of more "like" or "similar-others" would affect individual suicide risk and found that community "sameness" generally reduces individual risk of suicide. This multi-level examination of individuals' embedded lives provides a glimpse into how "sameness" taps into structural ties, normative climates, and social diffusion processes. In fact, some of the most robust suicide research findings at the individual level are dramatically moderated by a consideration of their social environmental counterpart . Recent research further reaffirms the notion that current gaps in societal safety nets contribute to emotional distress and suicidality during the COVID-19 pandemic . Collectively, these findings suggested a critical and fundamental sociological insight into suicide: connectedness is protective to a point. Where there are too few others at risk , socially supportive ties may be unavailable but when others share the same fate, the sense of individual failure transforms into structural failure , potentially reducing the psychological harm of the experience. But when that sense of despair or fatalism engulfs the community as a whole, the ability to see any future can be restricted in isolated and historically stigmatized communities . These studies illustrate how essential it is to consider the roles of social structure and culture in social interaction, as a core feature of theories of suicide; to not do so contradicts basic contributions to contemporary population health research . It also illustrates that while Durkheim offered foundational insights into suicide, focusing overly on his specific hypotheses rather than the general theoretical propositions or attempting an artificially general theory of suicide, only weakens our capacity to understand how the external social world shapes suicide. And while structural insights into suicide represent major advances, contemporary sociological research raises two theoretical issues that cannot be ignored, and must be synthesized into, the understanding of suicide as a complex phenomenon -culture and contagion. --- Exposure to Suicide While Durkheim presented himself as a general theory of suicide, there are intricate aspects of social interaction that fall outside his purview but are related to how social structures and connectedness impact suicide. A second major line of sociological scholarship examines exposure to suicide through one's social networks and communities and in so doing offers perhaps the clearest example of how social ties can produce harm . Decades of research from a variety of methodological approaches and causal modeling strategies has confirmed that exposure to media reports of suicides -especially celebrity suicides -or personal role models, like parents or friends , is associated with increased risk of suicidality. This line of research emerged from a series of studies by sociologist David Phillips ) that found that suicide rates among audiences exposed to media reports of suicides would spike temporarily. At the time, this was radical in that Durkheim famously denied the roles of micro-sociological processes related to social interaction, as well as diffusion or contagion in suicide. Phillips turned to a forgotten sociologist, Gabriel Tarde, to think theoretically through what he came to call suicide suggestion, a term derived from Tarde, who wrote about the diffusion of ideas and behaviors through social relationships . Tarde was what would be called a social psychologist, but in the late 19th/early 20th century, his epistemology was too close to psychology, and thus Durkheim rejected it out of hand. Durkheim firmly committed to the idea that larger structural forces were causal, and thus, he is usually understood as rejecting the idea that suicide could "spread" or be "socially contagious." And while research has repeatedly found, using more conservative methods than Phillips, an association between media exposure and increases in suicide rates, like Durkheim, these studies fall short in their ability to identify the primary mechanism or mechanisms that link the media exposure to the individual-level actions. Nevertheless, a series of promising studies emerged following Phillips' work, which focused on the consequences of being exposed to a personal role model's suicidality . With the growth of network analysis in the 21st century, suicide scholars in this burgeoning tradition began taking cues from network studies that found many social behaviors, like obesity and smoking, were socially "contagious, " net of individual factors . It became apparent that the structure of a person's social network mattered, as longitudinal research found that adolescent exposure to friends of friends was associated with greater risks of suicidality . Likewise, networks appear to have gendered effects, with girls being most at risk of suicidality when they have exceedingly small social networks or are immersed in exceedingly large ones . Additionally, in a groundbreaking study, Baller and Richardson used spatial analysis to determine how crucial characteristics of place -like the degree of infrastructure -are to the clustering of suicides in places. They concluded that the structure of place and diffusion processes cannot be divorced from each other; once again illustrating the importance of theorizing and modeling the multiple levels of society within which human behavior is situated. Despite these advances, the question remained why and how suicide contagion worked. While this is still an area in need of further exposition, in one unique study, researchers leveraged network data with pairs of adolescent friends to determine whether knowledge of a suicide attempt was necessary for suicide contagion to occur . The study found that youth who did not know their friend had attempted suicide were not at higher risk of suicidality over time, though if they did know they were. Additionally, exposure to a friends' suicidal thoughts was not sufficient to increase risk of suicidal thoughts or behaviors 1 year later. These findings suggest the power of behavioral role modeling. How, why, and when social behaviors diffuse through social networks or contexts is an important and ongoing area of inquiry within sociology of suicide specifically and social network science more generally . --- Regulation, Culture, and Behavior While these structural sociological theories described above offer multiple important advances for the sociology of suicide, they leave several unexplored social scientific questions -specifically, what mechanisms translate structure into meaningful social beliefs and practices that shape our attitudes and behaviors related to mental health and, ultimately, suicidality. Arguably, these gaps in the sociology of suicide can be addressed by drawing on insights from the broader theories in the sociology of culture and sociological social psychology. The incorporation of culture and sociological social psychology matters for several reasons. Eschewing explanations that motivate behavior by intra-personal perceptions, sociologists have generated substantial evidence that individual behavior is motivated -and justified -in reference to the web of social relationships and the broader structures and cultures in which these are embedded . We begin by reviewing theoretical advances that reconceptualize Durkheim's regulation as a cultural force to better elaborate how culture shapes behavior and suicide. Durkheim's choice of regulation as a key causal force was rooted in the idea that collective ways of acting and thinking not only reinforced integration -that is, everyone is or is believed to be doing the same things, and thus share more than they differ -but that they were psychologically, emotionally, and socially healthy . Although Durkheim could not imagine using cultural analysis, his conceptual ideas about regulation square quite neatly with contemporary cultural sociology. Groups of all sizes have cultures, and these cultures are shared -within reason -providing individual members with a sense of who they are, what they are supposed to feel, think, and do under various conditions, and what it means to belong to that group . Culture is activated every time members interact in real life or when one member anticipates or imagines interacting with another member; culture is also activated whenever we come into contact with externalized representations of it , such as a Catholic individual seeing a crucifix. Members watch each other and sanction each other to regulate each other's behavior. However, culture also is internalized in our conceptions of the generalized other: people do not just act because they do not want to be sanctioned by others, but rather are motivated to act by the cultural schema, scripts, and frames they are exposed to and internalize and come to take for granted as normative . This set of insights is fundamental to explaining social behavior of all kinds but has largely been neglected in suicidology, even as structural and psychological accounts of suicide have been criticized since the 1960s for ignoring the role cultural meanings play in understanding and explaining variations in suicidality across time, space, and groups/classes of people . And though it may be tempting to dismiss cultural regulation as a causal mechanism, research on other types of behavior shows culture not only shapes us; it regulates us morally -that is, it may proscribe or prescribe a behavior as a normative option under a shared set of conditions. This is imperative for suicidology for two reasons. First, the last two decades have seen theories attempting to explain how suicide ideation is transformed into action expand dramatically . Second, these theories largely neglect the simple fact that suicide is a social act and therefore is replete with cultural meanings that attempters symbolically externalize to their intended and unintended audiences, who make sense of the suicide via meanings they too have internalized. Put in the language of many current psychological theories, cultural sociology argues that suicide is not just about acquiring the proper cognitive and practical capacities to attempt, but also the normative capacity, or the belief that suicide is a viable and socially acceptable option for expressing outwardly something felt internally . Recent decades have seen a growing body of historical, anthropological, and sociological evidence supporting the argument that culture matters to suicide. Research clearly demonstrates that societies and/or subgroups within those societies carry different beliefs about suicide across time and space and death more generally . These beliefs, ultimately, contribute to notions of when suicide is justified , if ever, and, therefore, erect prohibitions for entire classes of people or may make suicide a normative option . This argument extends beyond whole cultures and applies to subpopulations and their subcultures. For instance, research has shown that how Americans interpret the suicides of men and women is through very different "cultural scripts" ), which has consequences for how their performed suicidality may be expressed and received by both the attempter and her intended audience , and, for which type of person might be at risk of suicide under certain conditions . Other research has found distinctive beliefs and, subsequently, suicidal practices among young Latinas in the United States , in rapidly growing urban spaces in southern India , and some Indigenous communities in the United States and Canada . A second body of research underscoring the role of culture in suicide comes from a clinical psychology of bereavement. In short, Robert Neimeyer and his many collaborators have demonstrated that sudden deaths, like suicides, are shocking and compel individuals to make sense of them, to sift through available meanings as part of the bereavement process . Though not a sociologist, Neimeyer and colleagues repeatedly find that meaning-making and bereavement always occur within the confines of a collective, as they build a coherent sense of why the death happened through each member's individual meanings and more general societal ones . In the event that collective meaning-making fails or that unhealthy meanings are arrived at, bereavement can become prolonged, thereby placing the individual at a significantly higher risk of emotional distress and suicidality. A similar set of studies examine how structure and culture interact together, marrying Durkheimian insights to some of the more innovative cultural studies. For instance, research in Indigenous communities has made important connections between the social, cultural, and geographic circumscription that delimits social networks within some indigenous communities to the intergenerational negative affect experienced and passed on due to discrimination and prejudice . In one community, for instance, youth associated suicide with belongingness; that is, to die by suicide was to express one's commitment to the group's expectations and its members . In rural China, Fei also identified linkages between structure and culture: where traditional patriarchal families tightly constrained women's ability to express grievances, suicide had become means of expressing grievance, justice, and anger. Finally, in a recent publication, sociologists Muller et al. leveraged extremely unique longitudinal data linked to death records to examine how male adolescents' desired occupations translated into risk of suicide by mid-life when those occupations became unavailable due to economic declines in those occupations. The structural changes in the labor market interacted with cultural ideals for work and success, such that when worked declined, men who expected a reliable workingclass job were more likely to die by suicide than their peers. This study suggests that it is not simply occupational or education attainment that generates risk of suicide, and not simply economic societal changes; but rather, the macro-societal translates into distress through an individual's cultural values, identities, and expectations. A third set of studies revolved around an in-depth ethnographic case study of a community called "Poplar Grove, " a white, affluent, homogeneous community with an intense high-pressure culture revealed that youth and parents alike had developed suicide explanation that had expanded for whom suicide was an option . Youth believed other youth used suicide to escape the intense pressure and that the misery induced by the pressure caused suicide . Though more research is necessary on this , this study suggests that identification with perceived and socially legitimated motives for suicide may increase youth's vulnerability to suicide and may be one explanation for why suicide clusters form and persist . Further teasing out the mechanisms that translate external social environmental factors into internal psychological pain is a crucial project for the sociology of suicide. One strategy is to integrate principles drawn from sociological social psychology; a project the sociology of suicide has recently begun and which we turn to next. --- The Necessary Role of Social Psychology Although Durkheim was not and could not be a social psychologist, contemporary sociological social psychology offers key mechanisms for understanding and explaining suicide within the context of structural and cultural contexts. Durkheim recognized in Suicide that individual's membership in a specific group or category of people made them more or less vulnerable based on that collective's integrative and regulative characteristics. Contemporary accounts have extended these insights, linking them to individual feelings or beliefs about who we are and what we are supposed to be doing. However, it is the mechanism linking us to the group, or what sociological social psychologists call identities and the emotional attachment we have to our identities and to the group that help us make sense of why structures and cultures may be harmful or protective. The basic premise of a social psychological theory of suicide, then, rests on four key aspects of identity and emotion . First, persons whose identity is structurally and culturally embedded in a relationship, group, or broader social system will feel higher levels of commitment to the identity. Commitment depends on both intensive and extensive social ties that evoke the identity . Second, where commitment to an identity is high, the person will also be affectually attached to the bond itself . Third, the more committed an individual is to an identity and attached to a bond, the more influence other members have on the feelings, thoughts, and actions of the individual. Fourth, where fewer alternative identities and bonds exist, subjectively and/or objectively, cultural regulation will be at its most powerful as continued commitment and attachment are more desirable than exclusion and isolation . Below, then, we examine a little more closely what identities are and why emotions, especially social emotions, can help explain suicidality. --- Identity Identities are internalized meanings that cluster around how an individual understands themselves, as a social object, in relationship to a real person , a group , a social class , or an abstract collective , which are embedded in social structure and culture . In turn, like the example of Catholic objects imbued with collective emotions and public meaning, our identities are objects inseparable from the collectives they are anchored, which makes them as emotionally charged as the external objects themselves. They matter to us because the relationships that allow them to exist matter. And, like any object that takes on meaning in interaction, relationships are where people acquire these identities as they learn about who they are, the expectations that others have of them and that they have others, what rewards, performances, and influence they can expect to have, and so forth. Identity matters, then to suicide and mental health, because it is one prominent pathway through which the external social world comes to matter to perceptions of self. Our identity renders painful the possibility of exclusion, rejection, and isolation from cherished social groups, not simply because we feel lonely, but because a part of our self can be damaged or lost through these social experiences. And, when we assign blame to our self for the rejection by a group , emotions signaling we are "bad" or "worthless" may snowball into psychache and negative emotion feedback loops . Returning, then, to the study of Poplar Grove, youth in this community did indeed internalized a very clear, rigid, coherent sense of what was expected of the "typical" Poplar Grove youth . The small nature of the community delimited the variation in how this identity could be performed, and thus made even the counterfactual cases we spoke with painfully aware of expectations. And, because the school took on an outsized role in community life, this identity was trans-situational, defining nearly all of the relationships inside and outside of school. This had three key consequences for the suicide problem in Poplar Grove. First, youth had also internalized the cultural script of pressure leads to emotional distress, which can lead to suicide being a normative option for expressing one's identity and extinguishing the pain. Second, the community had set most of the kids up to fail, as only one kid could be captain of the football team, lead actor in the big school play, or most popular kid. Anything short of five AP classes per semester and straight A's was viewed as a failure by youth, making falling short of expectations the norm and not the exception. Third, fear of failure, imagining or anticipating failure, and actual failure all lead to the same thing: shame . Shame is a painful social emotion signaling that the person has not only not met expectations but are actually a "bad" person because of doing so; it is social in that they believe, whether true or not, that others judge them as deficient. Identities are intimately implicated in this process, as not meeting expectations generate negative affect that compels us to meet them , but because of the second consequence described above, failing was perceived as a chronic, normal state of adolescence. And thus, we must consider, in a bit more detail, the role of emotions in suicidology. --- Emotions Generally speaking, suicidology has focused on cognitive appraisals of emotions , as opposed to the affect themselves, which is very often shaped by the cultural world around us. Emotions are both the "glue" of social relationships and can signal our successful integration or fulfillment of obligations or expectations, or our failure to do so; and, as such, are a fundamental element of how Durkheim's regulation becomes internalized into psychological well-being or pain . Thus, emotions create and sustain attachments to others and our own commitment to the identity associated with the attachment. In turn, this level of integration engenders greater regulation as we are more likely to adopt the feelings, thoughts, and actions of those we are most affectually attached . On the other hand, emotions, particularly negative social emotions like embarrassment, guilt, or shame, are the signals that this connection is in danger, dissolving, or lost . The link between identity and culture points, then, to two key insights drawn from scholarship on emotions and behavior. First, when we are not performing our identities as others expect or as we expect, we feel negative social emotions like embarrassment, guilt, and shame . What makes us feel bad about ourselves, or creates the cognitive appraisals like worthlessness or hopelessness, is very much a product of the cultural milieu that provides us with expectations about who we are and why we are supposed to do. Second, depending on the structural and cultural context, these social emotions may endure over time, making it increasingly difficult to live up to expectations and overwhelming our ordinary cognitive and behavioral functions, leading us to draw from existing cultural options for dealing with those emotions. In particular, shame or the social emotions that that the self is viewed as being corrupt, polluted, deficient, and contemptuous by others -objectively or not -plays a key role . Research has demonstrated the role shame plays in a range of negative behaviors, such as domestic violence , eating disorders , and criminality . It also has some anecdotal links to suicide . The shame pathway, then, can be tied directly to our discussion of social psychology, identity, and expectations: failing to meet expectations can trigger shame. In part, this may be due to the publicly shared cultural meanings. For instance, research in cultures or subcultures with strong traditional male norms evince far more "honor" suicides as failure to meet masculine expectations are closely tied to suicide as a way of restoring honor . Sudden loss of status, in most cases, is followed by intense shame and the need to process the shame. Shame also plays a role for those in subordinate positions, whose identities are wrapped up in being powerless. In some traditionally patriarchal societies, like rural China , there may be no other culturally available recourse to processing their shame besides suicide. Indeed, as Zhang's use of strain theory and innovative methods reveal, there are severe structural constraints on access to many legitimate means to reducing anxiety and stress. Youth, too, are in a relatively powerless position coupled with being at a disadvantaged cognitive and emotional developmental state that precludes being able to see far into the future. Shame can be experienced so acutely for these kids, the availability, accessibility, and applicability of a suicide script may be the only ingredient missing for leading to suicide vis-à-vis drug or alcohol abuse. Thus, social emotions are a powerful vehicle, particularly when rooted in salient social identities in valued social environments, through which the external social world is translated to internal psychological pain. --- DISCUSSION Sociological theories of suicide, inspired by Durkheim's original work, help explain how the external social world matters to individual well-being and psychache, thereby revealing the social roots of suicide. The external social world is complex and multilayered and can be characterized by network structures and shared cultures which in turn impact individual group members through their social identities and social emotions. That the external social environment matters to human development across the life course, including to physical and mental health, and even suicide, is not necessarily new. However, as rates of suicide have climbed in the United States and around the world, the importance of understanding the social environment's roles in suicide and suicide prevention has become more prominent and even urgent . Sociology, with our long tradition of specifying how society conditions human lives, is well situated to answer this call, while also building bridges into other disciplines. --- Implications for Psychological Theories of Suicide Many psychological theories of suicide acknowledge social and environmental factors, facilitating the incorporation of sociological insights to suicide. For example, belongingness is critical to Joiner's interpersonal theory of suicide, and connectedness is a key component of Klonsky and May's three-step theory of Suicide . There are two primary ways that sociological insights should, we argue, be incorporated into major psychological theories of suicide. First, while psychological theories of suicide recognize that the external social world matters, they generally distill the social world down to an individual's perception of it . Sociological research suggests this is insufficient and that using strategies to measure the external social world independent of a person's perception or experience is important. This could be as simple as using egocentric network methodology to better measure the culture and structure of a person's proximate social environments . This approach would involve having focal research respondents report their multiplex network ties using name generators and characterizing them through theoryinformed name interpreters. An ideal research design then involves interviewing some of the nominated network ties, so that data does not rely solely on the focal respondents' perception. Second, structural-cultural insights into suicide reveal that cultural scripts for suicide that prevail in people's salient social groups may impact their capacity for suicide . While the notion of individuals' capacity for suicide already exists , recognizing that normative capacity -or how a group's beliefs about why people die by suicide, who is expected to be vulnerable to suicide, as well as when, where, and how people suicidecontributes to making suicide an accessible and applicable option for an individual. Recognizing -and measuring -this may be a useful pathway for future research to examine; particularly given research linking explicit and implicit beliefs about suicide to suicide attempts and even death . --- Implications for Suicide Prevention Recognizing the importance of the social environment is also critical to strategies for suicide prevention. Some current suicide prevention strategies recognize the potential of broader, upstream environmental interventions, such as the Centers for Disease Control and Prevention's emphasis on social connectedness in communities. The focus on building connectedness has also been leveraged to great effect in schools. Specifically, building trust between youth and adults in schools is associated with lower rates of suicidality among students ). Similarly, there are suicide prevention interventions that address cultural biases, like mental health stigma, in communities or schools . These interventions raise mental health awareness and normalize discussing mental health, which may foster help-seeking and diminish suicidality in the entire community. Additionally, suicide prevention strategies in healthcare -specifically socalled "Zero Suicide" approaches -promote changes in the social environment within healthcare organizations to improve medicine's ability to prevent suicide . Specifically, a major component in the Zero Suicide model is generating system wide cultural change that renders suicide prevention a core organizational goal of any medical setting . Finally, recent research suggests that interventions into economic safety nets are associated with suicide rates; specifically, increases in the minimum wage are associated with meaningful decreases in suicide mortality , perhaps especially when unemployment is high . This suggests that macrolevel economic policies, untheorized as suicide prevention, may actually be powerful tools for just that. While collectively these interventions show promise, limitations remain. For example, in terms of culture, much of these interventions focus narrowly on mental health stigma, despite substantial research that demonstrates a plethora of cultural beliefs that can promote vulnerability to suicide and its precursors. This may be particularly harmful when connectedness is leveraged in schools. Schools that house harmful youth cultures may find that intensifying connectedness, even when combined with positive mental health messaging, may, at worst, amplify their harmful culture or, at best, find that the unaddressed harmful culture undermines any positive cultural interventions . Similarly, with regard to organizational interventions like Zero Suicide, it is potentially not enough to encourage an organization to value suicide prevention and mental health; it is likely necessary to broaden the scope of research and understand the external pressures, obligations, or cultural directives the organization faces and examine how mental health and suicide prevention complements or competes with those other organizational directives. This critique is motivated by previous sociological research that shows that understanding how organizations balance competing goals is crucial to effective prevention . Unfortunately, when organizations face external pressures , public health safety is often deprioritized in favor of more dominant goals . --- Future Directions This last point highlights a broader limitation in suicidology that in turn points to a crucial future direction for research. Zero Suicide approaches are one of the only explicitly organizational approaches to understanding suicide or suicide prevention. In general, though we acknowledge the role of several key organizations [schools and healthcare ] in suicidology, we have largely neglected to theorize or examine empirically the role of organizations in suicide risk and prevention. This is a major limitation since suicide prevention largely takes place within formal organizations, and several formal organizations are implicated in suicide risk [e.g., occupations , military , and schools ]. It is also a missed opportunity to leverage organizational science to improve suicide prevention. Within organizational science there are substantial literatures that have identified how to build safety systems to prevent hard to predict tragedies , like suicide. An organizational approach to suicide prevention has other advantages, as it can help identify existing unused safety systems in organizations that could be leveraged for suicide prevention. For example, schools generally have existing multi-tiered systems of support -often for academic interventions or violence prevention -that potentially could be leveraged efficiently and effectively for suicide prevention . There is a second critical future direction and current substantial limitation that warrants discussion. To date, theories of suicide largely neglect how structural inequality, colonization, and intersecting systems of oppression, privilege, and power shape vulnerability to suicide. Though there have been some exiting new efforts to theorize how structural inequality and intersectionality matter to suicidology , much more work is needed. Based on broader research within the sociology of mental healthwhich does take up this issue -the patterns are likely to be complex and again not distillable to individual experiences with discrimination or prejudice . Prior research on mental health and inequality demonstrates that external social structures condition mental health above and beyond individual experiences . While it's beyond the scope of this review to propose a new theory of inequality, power, and suicide, we can point scholars to useful theories of inequality in mental and physical health to aid them as we collectively take up this critical agenda . Additionally, understanding inequality will likely have real consequences for suicide prevention. For example, though upstream suicide prevention strategies are showing great promise in schools , many schools struggle to sustain even evidencebased strategies over the long-run . This may be in part because many schools, particularly those that serve disadvantaged youth, experience intense resource scarcity . Thus, considering the complex ways that inequality shapes suicide and suicide prevention is necessary to a robust, comprehensive theory of suicide. --- CONCLUSION Sociology is best known for our Durkheimian insight into why people die by suicide -namely, that lacking meaningful social relationships that support us during difficult times and celebrate us when times are good is extremely harmful to individual well-being. However, a review of the full body of sociological scholarship, and especially the empirical and theoretical advances of the past 10 years, reveal the social roots of suicide. Incorporating sociological insights into how the external social environment can matter to suicide and suicide prevention may help us better understand the complexity of suicide and determine how to effectively intervene. --- --- 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.

The past 20 years have seen dramatic rises in suicide rates in the United States and other countries around the world. These trends have been identified as a public health crisis in urgent need of new solutions and have spurred significant research efforts to improve our understanding of suicide and strategies to prevent it. Unfortunately, despite making significant contributions to the founding of suicidology -through Emile Durkheim's classic Suicide (1897/1951) -sociology's role has been less prominent in contemporary efforts to address these tragic trends, though as we will show, sociological theories offer great promise for advancing our understanding of suicide and improving the efficacy of suicide prevention. Here, we review sociological theory and empirical research on suicide. We begin where all sociologists must: with Durkheim. However, we offer a more comprehensive understanding of Durkheim's insights into suicide than the prior reviews provided by those in other disciplines. In so doing, we reveal the nuance and richness of Durkheim's insights that have been largely lost in modern suicidology, despite being foundational to all sociological theories of suicide -even those that have moved beyond his model. We proceed to discuss broadly acknowledged limitations to Durkheim's theory of suicide and review how more recent theoretical efforts have not only addressed those concerns, but have done so by bringing a larger swatch of sociology's theoretical and empirical toolkit to bare on suicide. Specifically, we review how recent sociological theories of suicide have incorporated insights from social network theories, cultural sociology, sociology of emotions, and sociological social psychology to better theorize how the external social world matters to individual psychological pain and suffering. We conclude by making explicit bridges between sociological and psychological theories of suicide; by noting important limitations in knowledge about suicide -particularly regarding the roles of organizations, inequality, and intersectionality in suicide -that sociology is well situated to help address.

Introduction Art museums1 are continually in flux . Over the past century, art museums have slowly moved away from being spaces that contained private collections that could only be accessed by a select number of people , to now playing 'a central role in making culture accessible to the mass audience' . This emphasis is demonstrated by related research and the suggestion that social interaction, particularly for families , is a significant motivating factor for visitors going to art museums , alongside leisure, recreation, and the acquisition of knowledge . The spatial and social character of art museums has also been affected by the gradual incorporation of new and emerging technologies, such as virtual reality , which is the focus of this article. In 1995, for example, Benjamin Britton developed the VR art installation LASCAUX, which was a reimagining of the cave of Lascaux in France. Between1995 and 1997, he toured this exhibit across the US, France, Italy, England, and South Korea. The huge success of LASCAUX is not only testimony to the international power of VR , but equally demonstrates that '[despite] some initial hesitation, museology and humanities disciplines in general are catching up with this major technological breakthrough' . Today, VR is going through a renaissance . 'After many decades of incubation, fully immersive virtual reality has finally become a customer-ready technology' . This is evidenced by the mounting popularity of modern systems like the Oculus Rift and the HTC Vive, as well as Oculus's most recent standalone system, the Oculus Quest. Immersion and presence are two terms that are frequently used to describe the experience of VR . While immersion is 'simply what the technology delivers from an objective point of view ' , presence is contrastingly understood as being 'a human reaction to immersion' . In other words, immersion implicates the technological, while presence implicates the psychological. Symptomatic of this, there is mounting evidence to suggest that the ability to create virtual worlds that users 'feel as if they are part of' is predicated, in part at least, on the technology underpinning the experience. Notwithstanding the impact physical setting can, of course, have on any given experience , virtual or otherwise, VR today is significantly more advanced than it was in the 1980s and 1990s, which was marked by 'low resolution and unacceptable latency' . And the consequent potential to create more convincing experiences of 'being there' has not gone unnoticed by both scholars and industry practitioners alike. A body of work now surrounds VR and its manifold uses . Studies have explored the role VR might play in the realm of healthcare, serving as a tool to rehabilitate stroke victims , tackle phobias , and train surgeons . Research has also considered the implications of VR in education with some of the suggested effects of this technology being that it can increase the speed of learning, and deepen the retention of information . Likewise, scholarly enquiry has recently extended to the application of Mobile Virtual Reality headsets in public environment, and the suggestion this assemblage temporarily places users in a 'dislocated space', where the 'the digital no longer augments the physical, but rather the digital supersedes the physical while simultaneously being constrained by the norms and constraints of the surrounding shared space . At the same time, studies have also started to investigate the application and utility of VR in art museums as a nascent form of 'edutainment' that could provide the digital space required for new forms of sociality to arise . While these studies are undoubtedly helpful and do raise interesting questions, for the most part, such work has tended to focus on the technical character of VR and has not necessarily presented qualitative data on the lived experience of VR-based exhibitions . Given the forecasted growth in VR technology and its diffusion into other industries , a better understanding of the impact of VR in art museums becomes more critical. Our article, therefore, reports on an original research project conducted at Anise Gallery in London, UK, between June and July 2018, focusing on the multi-sensory, and VR-based, exhibition, Scents of Shad Thames2 . The research comprises 19 semi-structured interviews with participants who had just experienced the Scents of Shad Thames exhibition. 11 interviews were conducted with individual participants and the remaining 4 interviews were conducted in pairs. The study advances along three lines of enquiry. First, the research investigates whether the inclusion of VR might alter the endemic practice of people-watching. Second, the research explores whether established ways of navigating the physical setting of art museums might influence how users approach the digital space of VR. Third, the research examines whether the incorporation of VR might produce a qualitatively different experience of the art museum as a shared social space. In the following section, we outline the conceptual framework that underpins this research. This begins by detailing an antiessentialist understanding of spatiality alongside a symbolic interactionist approach to the development of social rules. We then summarise three recurring spatial and social behaviours that have emerged within the art museum: people-watching, the physical navigation of space, and the experience of this environment as a shared and social space . These behaviours inform our research questions that revolve around the possibility of VR reshaping the lived experience of art museums. We briefly describe the Anise Gallery, as well as provide a detailed description of Scents of Shad Thames, before outlining the methodological design of this research. Finally, we present and discus our findings. --- The Spatial and social Norms of Art Museum For Falk and Dierking , the experience of an art museum is affected by the physical, social and personal context. Not only does the architecture shape how visitors experience this or that exhibition, so too does the surrounding sociality, alongside the personal motivations for going . Space is not just an empty container that we simply inhabit, but instead a social product rooted in shared values and socially produced meanings ; space is modified through continued use, with ascribed meanings changing over time as a consequence of interrelations and interactions . The art museum, then, is very much a lived space in which social actions, often dependent on social norms, can alter and modify the phenomenological experience. This is not to suggest, however, that art museums are not governed by tacitly agreed rules of conduct. Established norms within these spaces make social behaviour predictable and understandable . An individual may actively decide to visit an art museum, but the behaviours enacted during this visit are strongly influenced by the social norms that predate this decision. A symbolic interactionist approach to social rules suggests that society is constructed through the continued interaction and negotiation between social agents and that 'agreement upon the relevant norms will be entwined with the activity needed to make sense of the sort of social event to which one is party' . Context is thus an important factor . Social agents apply specific context-based behaviours so that a shared understanding of a situation can be developed, and common norms adopted . Within the setting of the art museum, the following behaviours have gradually emerged, people-watching, the physical navigation of space, and the experience of this environment as a shared and social space . People-watching is a recurring motif of art museums. Visitors' attention is always partially devoted to observing others co-inhabiting their space . Visitors watch either their own social group, other social groups or art museum staff to learn how to behave appropriately and to become oriented to the implicit conventions of their environment. This process is known as modelling . Casey argues that 'the viewer is acutely aware of his place in the museum and acts appropriately. He positions himself in front of the painting in a certain way, seeing himself see the object, mindful of how he is viewed by others' . For Falk and Dierking , a notable part of this self-awareness revolves around the fear many adults have about making mistakes in front of an audience . Navigation of the art museum is another important feature of these environments. In many ways, this is indicative of the necessary ordering of objects and artefacts. Exhibits and displays are usually grouped in a decisive manner and designed to be experienced in a particular sequence that visitors are expected to follow . Likewise, adults have been socialized to not touch the items contained within these spaces . Duncan argues that art museums provide 'both stage set and script' for visitors to performatively adhere to. While the overall flow of bodies may seem chaotic, most people consistently move from one object to the next . And the majority of people are also conscious of not negatively affecting the experience of others. As Falk and Dierking point out, visitors often 'quietly wait their turn to look at exhibited objects, and respect the rights of others to have a turn' . Consequently, people are more likely to participate in interactive exhibitions if they feel their experience is not prohibiting other people from similarly taking part . Art museums are, of course, very much shared social spaces, predicated on a 'public-private duality' . '[Whether] they arrive accompanied or not, all visitors have to negotiate an often delicate balance between two irreconcilable poles: conviviality -sharing, exchanging ideas; and introspectionestablishing a personal relationship with the art works' . Being accompanied by someone else is frequently part of the experience of art museums and individuals tend to move through these spaces as a unit, sharing their thoughts and exchanging ideas to help comprehend and validate each other's reactions . Yet, the introspective element of the art museum is just as important as the social experience. The process of engaging with a work of art creates a conceptual space that can feel solitary . The practice of visiting an art museum alone can therefore make the overall experience feel more private and provide 'an opportunity for self-reflection, tranquillity and personal freedom that the elevated, austere, even magical atmosphere of a museum particularly lends itself to' . Consequently, many visitors establish verbal or physical barriers to social interaction in order to be able to seamlessly move between the public and private experience of the museum . Drawing on an understanding of spatiality as something that is socially constructed , alongside a symbolic interactionist approach to social rules it is our contention that the addition of VR in art museums has the potential to present new ways of approaching and experiencing this setting. While a body of research has gradually developed around the social norms between avatars within virtual environments little empirical research exists regarding social norms when VR is adopted in public spaces like art museums. Similarly, there are limited studies that explicate the lived experience of VR-based exhibitions . Our research, therefore, addresses these gaps in the literature. The questions guiding our study are as follows. First, does the inclusion of VR within art museums alter the practice of people-watching? Second, do established ways of navigating the physical setting of art museums influence how users approach the digital space of VR? Third, does the incorporation of VR produce a qualitatively different experience of the art museum as a shared social space? --- Method This article reports on an original research project that explores the spatial and social impact of VR in the context of art museums. Our research was conducted at Anise Gallery in London, between June and July 2018. Anise Gallery presents a variety of exhibitions and events that relate back to 'architecture, technology and the built environment' . Founded in 2012, Anise Building is situated along historic Shad Thames, occupying the ground floor of renovated spice warehouse. The multi-sensory, and VR-based, exhibition, Scents of Shad Thames, which is the focus of this article, chiefly engaged with the regeneration of London and 'how the identity of the architecture is shaped by its inhabitants and vice versa' . After putting on the supplied VR headset -which was tethered to a powerful PC -visitors were presented with a detailed virtual depiction of the London SE1 postcode area, which effectively hovered in the middle of the room they were standing in. Six selected Shad Thames buildings, including Wheat Wharf, Anchor Brewhouse, Anise Building, Cinnamon Wharf, Tea Trade Wharf, and 28 Shad Thames, then slowly floated upwards from of the map before moving towards six evenly spaced positions within the gallery. Each designated position was accordingly marked by a red building shaped block. And as each building reached its destination, the red blocks became detailed versions of the buildings listed above. Significantly, these virtual buildings corresponded to six physical tables within the gallery, each with a doll house sized crate perched on top of it, and each filled a different spice, tea or fruit. Accordingly, visitors engaged in a multisensory experience, taking in the scents historically associated with this envirnoment. It is also important to point out that this experience took place amongst various other exhibitions and was thus not cordoned off from the shared social space of Anise Gallery. In total, 19 semi-structured interviews were conducted. Participants comprised 10 females and 9 males. 11 interviews were conducted with individual participants and the remaining 4 interviews were conducted in pairs. The mix of interviews was considered appropriate because couples often wanted to be interviewed together and their individual experience was influenced by the presence of a companion in the gallery. Purposive sampling was employed throughout this project. Participants were selected if they had experienced the VR portion of the Scents of Shad Thames exhibit and were over 18 years old. In all instances, potential participants were approached within Anise Gallery, given a participant information sheet and then asked if they would like to take part in the project. Participants then completed a consent form, which detailed how their data would be used following the study. All participants were given pseudonyms. Interviews consisted of 16 questions and 45 sub-questions. Each interview was effectively broken down into three sections. The first section focused on the art gallery, with questions exploring why participants visited this gallery and how they usually felt and acted, spatially and socially speaking, when interacting in similar settings. The second section focused on the Scents of Shad Thames, with questions exploring how users described this experience, and how they approached and interacted with the exhibit itself. The third section focused on both the spatial and social effect of using VR within this context, with questions exploring how participants felt this experience differed from non-VR based exhibits, and what effect the dislocated space of VR had on the social aspect of their visit. Interview data were thematically analysed using an established series of codes that related to the key themes underpinning this study: people-watching, the physical navigation of space, and the experience of this environment as a shared and social space; with each theme containing several sub-themes or codes. A question that focused on both the spatial and social effect of using VR within the context of the art gallery, for example, elicited responses that were commensurate with the following broad themes: emerging forms of privacy; the impact of the physical setting on the digital experience; concern for how participants might appear to visitors outside of the headset; and reconciling the sensorial configuration of VR with the desire to continue socialising. This approach to data was both inductive and deductive. All data were initially coded outside of our theoretical framework, before being reinterpreted through the chief themes underpinning our study. --- People-watching In line with established understandings of art museums , people-watching remained an integral facet of our participants' experience of Anise Gallery. As Nici explains: 'I was thinking where's the VR and then I could see it in the corner. And then I saw people walking around very slowly so I thought okay there are things to take in' Here, many participants observed other visitors 'to gain information or knowledge about appropriate ways to use museums' ; or rather how to approach this VR exhibit. This point is supported by the following extracts: 'Before I did it, I saw this other woman doing it and she was very careful about what not to touch. I think she felt that she was going to touch the wall or something or fall' 'I saw some people walking into the crates. So, I was mindful of that when I was looking at the model. I was aware that there was a wooden crate in front of me' 'Well once you see somebody bump into the box, you think, I'll do better' The spatial and social norms of the art museum were therefore a significant factor in the inclusion of VR . At the same time, we would also suggest there is something noteworthy about the incorporation of this technology that differs from the people-watching norms commonly associated with more conventional exhibitions, effectively making the practice less casual and more active. Certainly, the observation of others that typically occurs in relation to non-digital displays , involves a visual connection of sorts between the actions of visitors and their physical environment. In other words, there is a tangible referent that corresponds to a certain sensibility. VR, however, is different. To account for the physical, visual and audible schism that occurs within a VR experience, Saker and Frith use the term 'dislocated space', which they define as 'concrete space being temporally superseded by … digital space' . This digital space remains concealed from those outside of the headset, just as the meanings communicated by the physical actions of users are necessarily partial. For some participants, the disruption of this practice meant that the process of people watching suddenly lacked meaning: 'When you're looking at people who are experiencing it, it doesn't mean that much to the bystander' 'You do not know what they are doing and why they are doing it. This doesn't have a meaning to you in the same way as you are looking at a picture' Interestingly, Thomas compares witnessing someone interacting with VR to looking at a picture, which is, of course, an established behaviour that facilitates a shared understanding, as well as implicates common norms . For Durkheim , norms are the essence of social order. Norms effectively provide a sense of solidarity among individuals. And this is no different in the setting of an art museum. When Thomas comments that the actions of the person he was watching lacked meaning, he is equally suggesting the solidarity of the space has been altered. Symptomatic of this, some participants found the attenuated nature of their experience evoked reflexive concerns about how they might appear to others were they to inhabit this digital space themselves. As Leah's pondering attests: 'Would I look silly? Would I trip up?' How to behave within an art museum is, of course, not something that is intuitive . Visitors with limited experience of these spaces might therefore feel more fearful of behaving in a manner that was unfitting . In the context of VR, our research found that the practice of peoplewatching not only took on a more active and contemplative hue, it also prophetically provoked anxiety in some participants about the prospect of reckoning with this seemingly normless experience. --- Navigation The Scents of Shad Thames was designed so that digital objects correlated with physical objects placed in the center of the gallery. Many participants approached this experience in a manner that closely resembled the norms ordinarily associated with the physical site of art museums . As Kat explains: 'I'm in a gallery so I have to go carefully, one, two, three, then go to the other side, one, two, three. And I think that's a way that you should be in a gallery' Importantly, this way of being is predicated on a particular sequencing of objects, which is another rule generally associated with art museums. The rules or norms for navigating art museums involves visitors sequentially moving from object to object in a methodical and purposeful fashion . As Falk and Dierking explain, 'many museums design exhibitions composed of groupings of cases or interactive displays that present a single, large, multi-faceted story or concept' . This sentiment was echoed by Peter when he explained why he chose to adopt a regimented approach to this virtual exhibit: 'To get the most out of the place' For Peter, the navigation of this space is intentionally structured to produce an optimum experience. In the majority of instances, participants described approaching the digital space of VR following a normative pattern and comprehended virtual items as being curated in a manner not dissimilar to their physical counterparts. Our research also found that many participants approached the virtual space of VR as if it were physical. These participants consequently spoke of being wary of walking into digital objects even though they knew, of course, that this was not possible: 'You know that the table is not really there. You've seen the space you could walk straight through it, but you just obey the rules of the virtual space and walk around it' 'You know it's not real but we're walking around it as if, in terms of perception for safety, if we walk into it will cause us some sort of danger' 'I found that the space was so small for me to walk because there was that map in the center and I was afraid to crash with it. It was weird because I really know that it wasn't really there' . Behaviours commonly associated with the concrete space of the art museum therefore carried over into the digital space of VR, with each participant demonstrating a marked level of 'restraint' . Here, even the idea of transgressing the virtual map was discussed as being a rebellious act: 'We can do whatever we want. And I think that's the kind of rebellious, renegade kind of idea, like I'm just going to stand in there. At the same time, this is not to suggest these perceptions of 'restraint' were necessarily experienced as being undesirable. For many participants, the sense of alignment between the physical and digital world did not restrict their experience, but instead assisted their navigation of this space: 'I felt comfortable in the space and one of the things that I liked the most is that it is the same area. So, you already saw the space, you have already seen where you are going to walk around. It is the same one, so you don't feel that you're going to fall or drop' For Beth, the relationship between the physical and the digital meant she was attuned to the experience because she already had an understanding of the concrete space of the museum. As David amusingly describes this process: 'It's like seeing a friend in a funny hat' Significantly, then, the internalised parameters of this exhibition were not cast in a deleterious light. Virtual objects acted as boundaries, performing a similar function to physical objects and the confines of the art museum itself. For many participants, going outside of these boundaries felt 'weird' specifically because the norms associated with both physical and digital aspects of the museum appeared commensurate. The practice of avoiding virtual objects adhered to the social norm of not touching or interacting with items in a museum, which adults have been 'socialized' to do . As Falk and Dierking note, people know how to act in certain spaces, and 'they do not feel coerced into this behaviour; they just do it' . Similarly, our participants reinstated established norms, even though they were not compelled to perform in a certain way. It is our assertion, however, that this process does not simply illuminate the collective desire to establish boundaries to help make sense of an exhibit per se, but rather the power of the physical setting visà-vis the VR experience. As Sherman et al concurringly suggests, '[if] the same VR system and application are placed in two different venues, such as an entertainment arcade versus the Guggenheim Museum, there will be a significant difference in the way the experience is perceived' --- A Shared Social Space The art museum has been recognized as providing both a public and a private experience for visitors . Our research found that the use of VR in this setting readily amplified both the public and private aspects of Anise Gallery: 'What I was seeing was certainly private, but how I was viewing it and how I was interacting with it, the way I was putting my hand out was public' 'It was public because there was somebody else in the room. Somebody else was watching you while you were seeing a private exhibition' While both Kat and Katherine felt that they were inhabiting an environment that was qualitatively concealed they were equally aware that their physical movements within the digital space of the virtual corresponded with a physical space that they were equivalently cordoned off from. For George, it was the acoustics of the physical setting that made him more cognisant of the space outside: 'I was aware of the presence of others, because people are being loud' For some participants it was precisely the ability of VR to place them in a space that felt distinctly removed from other people -ocularly and audibly speaking -that made their experience feel qualitatively different. Carol makes an equivalent point while discussing the difference between a gallery she had recently visited and the Scents of Shad Thames: 'When you had it all by yourself and you were the only person there, that was really cool. It was like having your own private gallery moment. I really valued that because I think it's very rare that you would get, not everywhere is necessarily busy, but it's very rare that you get gallery space to yourself to enjoy something, so you're still part of the gallery but you're having your own experience' Art museums are, of course, commonly understood as being shared spaces . It is through the sharing of space that spatial and social norms are established. In this instance, however, Carol was able to engage with the exhibit as if she were the only person there because she did not have to share the virtual space with other people. Leah makes a similar point: 'You immerse yourself in what you're looking at. Everything else was disregarded. I thought, no, no, I'm having a good time, forget about everybody else' As Elden explains, 'the meaning of the space and the space itself, is adapted and transformed as it is perceived and lived by social actors ' . For some participants, it was precisely this solipsist vantagepoint and its marked difference from the common experience of other art museums that made the lived experience feel more satisfying: 'It was private because I feel like I have the space for myself. So, I was free to go to any of the pieces of work there were. But it's only because anyone can't see what I was doing. I feel like I have more freedom without people, because you can go anywhere you want without people being in the middle of it' At the same time, Peter also touches on an interesting contextual aspect of VR. Peter intimates that part of this 'freedom' was symptomatic of people not being able to see what he was doing. The removal of people provided a mastery of space and autonomy that is rare in a crowded museum. For Debenedetti , individuals who experience the museum alone benefit from total freedom in decision-making. Once inside, they are free to go at their own pace and stay as long as they wish. This autonomy allows the singleton to set a very personal itinerary ... to preserve the intimacy of their encounter with the artworks. And this spatial and social autonomy can facilitate a more immersive experience: 'I was engrossed as soon as I started doing it. Actually, I was surprised when I took off the headset how many people there were, how busy it was, and knowing these people had been watching what I was doing and staring I was like "oh, okay". When you're actually in the environment and you know, you are smelling things, you are seeing things, you are listening to things, and you are choosing where to navigate, you do not think about external things' 'I want to be in my head and with the art. So, I'm born for VR' For both Thomas and Marcy, then, the sense of privacy VR provided notably allowed them to be more absorbed in the experience and therefore less conscious of their own reflexive role in maintaining tacitly agreed norms. This is not to suggest, however, that the carapace of VR was impenetrable, as many participants equally found the concrete space of the art museum continued to bleed into their experience in a variety of ways. As Saker and Frith put it, 'the shared norms of actual space … remain a constraint upon actions in the virtual space' . Equally, our research found that for some participants their experience of the digital space of VR was marred by an awareness that other visitors might be waiting to have their turn, which resonates with common concerns associated with this setting . As Kat and Thomas explain: 'I am in front of a big group of people. There are people watching, waiting to have a go on this thing. You can feel self-conscious' 'I think it goes back to me being always self-conscious of not being in the way of people. So, it is like in a restaurant when there is a queue and you don't feel welcome to stay because there is somebody else waiting for the table. So, in this experience, I knew somebody else was waiting, even if it was him, waiting for the headset' For Kat, her attentiveness to people waiting meant that she could never fully transcend the physical space of the gallery and inhabit the digital space of the virtual. We suggest here that the fractional immersion of VR might intensify this experience as any sense of a queue building beyond the headset is, in part at least, imaginary. For other participants, it wasn't necessarily the temporality of the VR experience that led to the shared space of the art museum seeping into the digitality of the experience, but rather a concern for how their actions might be perceived by others. Here, the prophetic anxiety provoked by the practice of watching other people engage with this exhibit, as detailed above, moved beyond mere contemplation and actually shaped the experience from within the headset: 'I didn't know if someone was laughing at me when I was touching the box and nothing was there, and I was like wearing the goggles and reading the explanation from the buildings. I don't know, I felt kind of foolish' 'Because even what I was watching the guy before me, I was just there thinking he looks so funny, but he is just getting on with it. And then when I put it on, I thought "oh, I look funny"' 'If you could see yourself doing it you would be like "oh my god, I look like an idiot"' While the private experience became more personal for some participants, then, for others the public space of the art museum intensified and affected their experience of the virtual. This is perhaps in large part due to the unease that accompanies the feeling of being watched, which resonates with Sartre's suggestion that 'the intrusive Other shatters the illusion of mastery initially enjoyed by the solitary viewer and turns him into the spectacle of another's gaze ' . For other participants, however, an awareness of being observed was intensified precisely because they had not come to the gallery alone. Of course, traditionally speaking, companions readily share their experience of art museums through dialogic interaction . Our research found that while participants still conversed with other people during this exhibition the inclusion of VR notably reshaped how this sociality was experienced: 'It is difficult to interact with someone or something when you don't see what you are thinking' 'Yeah, I agree. It's hard to interact when you're not seeing other people. So, you know they're there, but you don't know which direction to speak to them' For these participants, then, a significant difficulty encountered in VR was the inability to 'see' what their partner was thinking, precisely because the digital space of VR was necessarily dislocated from the concrete setting of the gallery. And for some participants the inability to experience this exhibit with someone else was seen as being unfavourable: 'I actually quite naturally thought we'd experience it together' 'It could have made it better to interact with someone seeing the same thing' As a corollary to this, dialogic interaction had to occur after the Scents of Shad Thames instead of during the experience. Consequently, for these participants the social bonding commonly confirmed through imminently shared 'actions, observations and reflections' became markedly fragmented. --- Conclusion This article has examined the effect VR might be having on established spatial and social norms within art museums, specifically focusing on the VR-based exhibition, Scents of Shad Thames, at Anise Gallery in London, UK, between June and July 2018. Established spatial and social norms include, people-watching, the physical navigation of space, and the experience of this environment as a shared and social space . Three main inferences can be made about the impact of VR on the experience of this environment from our data. First, the process of people watching became notably less casual and more active. Participants were aware that they were observing users perform physical actions that were necessarily dislocated from the physical environment they were inhabiting . As a corollary to this, the practice became more contemplative, with participants not only thinking about how they might react to this digital space themselves but also trying to envisage the digital environment overlaying the physical space of the gallery. Second, established norms for physically moving through the art museum influenced how participants interacted with the virtual setting of the exhibition. In part, participants did this to help contextualise and make sense of the exhibit. At the same time, we would also suggest the serious sensibility adopted by some participants was indicative of the physical space of Anise Gallery. As Sherman et al's posit, '[the] venue, or setting in which the VR system resides, can have a great impact on how an event is experienced' . Third, the public and private duality of the art museum was amplified. Through VR, participants experienced a sense of 'being there' that felt markedly more private. For some participants, this meant that they ostensibly had the exhibition to themselves, which provided a more satisfying experience. For other participants, however, this meant that they became more conscious that other visitors were waiting their turn. For Galdieri and Cazzozzino , however, this is not indicative of VR per se, but rather the consequence of a system that only enables one user at a time. In other words, this is symptomatic of Scents of Shad Thames and the decision for it not to be a multiuser experience. Here, our data support the suggestion that presence is a psychological reaction to technology . While the technology involved can create immersive environments , this does not mean all visitors will have the same experience. In a similar vein, the dislocation of participants from their physical setting impacted social connections. For some participants, this meant they were unable to share the experience with their companion. Consequently, established practices of sociality within this setting fractured with the dislocated space of VR. In sum, then, while VR has the potential to create a different kind of spatial and social relationship with art museums, for the most part, the physical setting of art museums seemingly limits the transgression of established norms. However, as visitors become more comfortable with this technology, and emerging applications surface that allows multiple users to inhabit these digital spaces at any one time, it is very possible that different spatial and social norms might still develop that are not as firmly governed by the laws of physical space. Likewise, with the mounting popularity of modern systems like the Oculus Quest, and the multitude of applications available, people are now able to experience VR-based exhibits in environments that are not art museums, which raises interesting questions about the kind of interactions these exhibits might forge given the incongruous physical settings involved. In either case, one thing seems fairly clear: VR is set to continue permeating multiple industries in the coming years . Additional research is therefore encouraged to gain a broader understanding of the spatial and social effects of this unique technology within the field of museology.

Art Museums implicate established spatial and social norms. The norms that shape these behaviours are not fixed, but rather subject to change as the sociality and physicality of these spaces continues to develop. In recent years, the re-emergence of virtual reality (VR) has led to this technology being incorporated into art museums in the form of VR based exhibits. While a growing body of research now explores the various applications, uses, and effects of VR, there is a notable dearth of studies examining the impact VR might be having on the spatial and social experience of art museums. This article, therefore, reports on an original research project designed to address these concerns. The project was conducted at Anise Gallery in London, UK, between June and July 2018, and focused on the multi-sensory, and VR-based, exhibition, Scents of Shad Thames. The research involved 19 semi-structured interviews with participants who had just experienced this exhibition. Drawing on scholarly literature that surrounds the spatial and social norms pertaining to art museums, this study advances along three lines. First, the research explores whether the inclusion of VR might alter the practice of people-watching, which is endemic in this setting. Second, the research explores whether established ways of navigating the physical setting of art museums might influence how users approach the digital space of VR. Third, the research examines whether the incorporation of VR might produce a qualitatively different experience of the art museum as a shared social space.

Introduction Lung cancer has the highest mortality rate of all cancers worldwide [1]. Socioeconomic status has been associated with lung cancer in several studies, with people from lower socioeconomic backgrounds having the highest incidence rates [2][3][4][5][6][7][8]. SES reflects one's position in societal hierarchies, and is generally assessed by the interdependent dimensions of education, occupation and income. SES is linked with health/disease through multiple interacting pathways in terms of material and social resources, physical and psycho-social stressors, and health-related behaviors [9,10]. SES is strongly associated with smoking behavior [11], the most important risk factor in the etiology of lung cancer. However, many studies on lung cancer and SES do not adequately control for smoking behavior [12], and findings about the extent to what SES is explained by smoking are not consistent [3,7,13,14]. We investigated whether SES is a risk factor for lung cancer, and to what extent the association is reduced by consideration of smoking. We operationalized SES by two different occupation-based concepts. First, we measured SES by application of the International Socio-Economic Index of occupational status [15]. ISEI was originally constructed to create an internationally comparable socio-economic index by combining data on education, income, and occupation as the three main dimensions of SES. The different ISEI scores for occupations were calculated by assuming that occupation represents an intermediate factor which converts education into income [15]. Second, we used the European Socio-economic Classification , which categorizes social positions on the basis of typical employment relations and conditions of occupations [16]. We applied these two concepts to different job periods to investigate variations of occupational SES and lung cancer associations. Additionally, we explored whether the relationships between SES and lung cancer differed by histological tumor subtype, and conducted subgroup analyses to explore effects according to study region, occupational exposures, smoking status, education, birth cohort, study control type and city size of last residence. Considering biological as well as social differences between men and women with regard to lung cancer [17], we stratified all analysis by gender. --- Materials and methods --- Data availability We analyzed data from the SYNERGY study database. Detailed information on the SYNERGY project has been published previously [18,19] and is available at the study website . Briefly, SYNERGY is an international collaboration to study the role of occupational exposures on lung cancer risk. All included studies solicited detailed information on the participants' occupational biography coded industries) and smoking history. Individual participant data from 16 studies and 22 study centers conducted between 1985 and 2010 are currently included in SYNERGY. The ethics committees of the individual studies approved the conduct of the study, as well as the Institutional Review Board of the International Agency for Research on Cancer. Study subjects or -in the case of deceased subjects-their relatives gave written informed consent to participate in the study. We included studies from Europe and North America and used data from 12 studies conducted in 18 study centers. We excluded two studies because of missing information: The MORGEN study did not contain data on the time since smoking cessation for former smokers, and the PARIS study did not have information on education and was restricted to smokers. Participants were excluded if they had no ISCO codes in their occupational history to derive occupational SES . These included, for example, housewives, participants working exclusively in the military or lifetime unemployed. Participants with missing smoking history were also excluded . Cases were histologically confirmed lung cancer cases, categorized into lung cancer subtypes , small cell lung cancer , adenocarcinoma , other/unspecified). Information was available on several further variables, which either constituted the "exposure variables" or covariates. This included gender, age, geographic area of residence, smoking history, education, and occupational history. The occupational history was used to create the "exposure variables" and to create an indicator of potential exposure to occupational carcinogens. --- Indices of socioeconomic status In order to classify the SES of study participants, we used two indices that can be assigned by the participant's occupation, namely, the ISEI [15] and the ESeC [16]. The ISEI is a continuous independent research institute of the Ruhr-Universita ¨t Bochum. The authors are independent from the German Social Accident Insurance in study design, access to the collected data, responsibility for data analysis and interpretation, and the right to publish. The views expressed in this paper are those of the authors and not necessarily those of the sponsor. This does not alter our adherence to PLOS ONE policies on sharing data and materials. status score for occupations, derived by Ganzeboom and co-workers based on age, education and income. The minimum score was 10 , the maximum 90 . We used each participant's job history in conjunction with the ISEI score for the occupations to assign an ISEI score to each job. We categorized subjects into categories in two ways: first by dividing the entire ISEI range into four equal sub-ranges and second by calculating frequency distribution quartiles based on the gender-specific distribution of scores among control subjects. The ESeC is a derivative of the Erikson-Goldthorpe-Portocarero scheme [20]. In contrast to the continuous ISEI scale, ESeC defines discrete categories of social positions: Occupations are classified according to their typical employment relations and conditions referring to the labor market and work situation [21]. We applied the ESeC with 3 classes , which shows a hierarchical order unlike the original scale of 9 classes . The condensed version is recommended by the ESeC-authors when additional information about employment status and size of organization is missing [21]. For the assignment of the indicators we utilized instruments available on the authors' websites [21,22]. We assigned scores based on each participant's longest, first and last held job period and additionally, the lowest and highest score ever reached . Jobless periods due to unemployment were assessed separately. We categorized the maximum duration of unemployed periods and, for comparison, the sum of unemployed years for each participant . We further categorized participants in those who ever or never worked in blue collar jobs by the first digit of ISCO codes . Education was categorized as follows: no formal/some primary education , primary/some secondary education , secondary education/some college , university. --- Covariates The smoking history was parametrized by means of multiple variables: smoking status , years since quitting smoking, and pack-years ). Non-smokers were defined as participants who smoked less than one pack-year. Smokers were considered former smokers if they had quit smoking at least 2 years before the interview/diagnosis; otherwise they were considered current smokers [23]. Former smokers were subdivided into categories of 2-5, 6-10, 11-15, 16-25, 26-35 and more than 35 years since quitting smoking. To indicate occupational exposures to lung carcinogens, we used a classification of occupations developed by Ahrens and Merletti [24] on the basis of occupational categories and industrial sectors . The list of occupations with potential carcinogenic risk is known as 'list A' and includes, among others, jobs in metal production and processing, construction, mining, the chemical industry, asbestos production [24,25]. Participants were classified as ever or never having worked in a 'list A' job. We combined countries to the following study regions: Northern/Central Europe , Eastern Europe , Southern Europe , and Canada. We differentiated whether controls were recruited population-based or in hospitals. We categorized birth cohorts and city size of last residence . --- Statistical analysis We estimated odds ratios with 95% confidence intervals by unconditional logistic regression models, and used the longest held job for the main analyses. Categories with the highest SES were set as reference. We adjusted for log and study center in model 1 and added smoking variables in model 2. We stratified analyses by gender, restricted in some cases to men because of insufficient numbers in women. We calculated tests for trend for all analyses. To quantify the difference of ORs between the two models, we applied / Ã 100) [13,26]. We additionally adjusted models for educational level as a second SES indicator and 'list A' to study the impact on the association of occupational SES and lung cancer. To investigate whether the SES-lung cancer associations differed by histologic type, we conducted separate analyses in the main histological subtypes of lung cancer . Subgroup or sensitivity analyses were conducted to elucidate possible effects by education, study region, city size of last residence, birth cohort, employment in 'list A' job, employed in a blue collar job, smoking status, and type of control recruitment. We calculated correlations between the selected job periods and correlations with education by Spearman's rank correlation coefficient for ISEI and by Crame ´r's V for ESeC. We used random-effect meta-regression models to examine heterogeneity between study centers. The LUCA study was not included in the meta-analysis because adjustment for smoking was not possible due to missing cases in the reference category . All statistical analyses were carried out with SAS, version 9.3 except for meta-analyses, which were performed using Comprehensive Meta-Analysis Version 2.2.027 software . --- Results --- Characteristics of the study population Altogether, 17,021 cases of lung cancer and 20,885 controls were included in the final analysis. The characteristics of the study participants are shown in Table 1. Approximately 80% of cases and controls were male. Lung cancer cases less frequently held jobs in the highest occupational categories, had lower education, were more frequently smokers at time of interview, had smoked more pack-years and slightly more often experienced unemployment than controls. Fractions of participants with higher occupational SES , higher education, and non-smokers were lower among men. The maximum duration of periods of unemployment was higher for women than for men. When combining the upper categories of ISEI to high SES and the lower categories to low SES, current smokers represented 47% of men and 36% of women with low SES compared to 34% of men and 31% of women with high SES. Non-smokers accounted for 12% of men with high SES and 20% of men with low SES. In women, the proportion of non-smokers was equal for low and high SES . The distribution of SES among the controls varied by study center in particular with a higher proportion of lower SES in CAPUA and higher SES in TORONTO . of selected job periods of ISEI and ESeC, with corresponding tests for trend . --- Associations between SES and lung cancer The average reduction due to adjustment for smoking habits in men was 50% for ISEI and 26% for ESeC, and in women 34% for ISEI and 9% for ESeC. Unemployment with a maximum duration of >5-10 years and >10 years was associated with an increased risk of lung cancer for men . Similar results were observed for cumulative unemployment of 5-10 years and > 10 years . The results for either ISEI categorization, based on the score-range or the gender-specific control distribution , showed similar ORs. We also observed similar associations between SES and lung cancer for the longest and last job periods and the highest ever reached ISEI on the one hand, and for the first job and the lowest ever reached ISEI on the other hand. The job periods within these two groups were highly correlated . Additional adjustment for education further reduced risk estimates on average by approximately 50% whereas adjustment for 'list A' resulted in a slight reduction . Occupational SES correlated moderately with education . When stratifying the data by histological tumor subtype , we observed increased ORs for SQCC and SCLC and slightly reduced risks for the lower SES-categories for ADC. In women, adjustment for smoking behavior increased ORs for SQCC and SCLC in the lower SES categories. --- Subgroup and sensitivity analyses, meta-analysis Table 5 shows results for the subgroup analyses: The effect estimates remained unchanged for participants who never or ever worked in a 'list A' job and for male non-smokers of the lowest SES category. ORs were comparatively higher for population than hospital controls; lower for participants most recently residing in an urban area, and also for men who never held a bluecollar job. When exploring last residence in urban area for the younger half of the study population ORs increased marginally for women . Stratification by study region revealed higher ORs in Northern/Central Europe and lower ORs in the other regions with a negative association for women in Eastern Europe. In comparison to the score-based categorization of ISEI, applying gender-specific ISEI-quartiles attenuated associations for women except for Canada, and increased ORs in men for Southern Europe. ORs increased in the birth cohort of 1930-1939 for men and, especially in the middle SES categories, in the birth cohort >1939 for women . The lung cancer risk of the lower SES-groups decreased when stratifying for education, especially in the strata of higher education . Meta-analyses showed slightly lower overall ORs than the corresponding pooled ORs. The stronger the association of lung cancer and SES, the higher were the proportions of heterogeneity with above 60% for at least the lowest vs. highest SES-categories. --- Discussion In this study we confirmed a social gradient for lung cancer, with greater risk associated with lower occupational SES that persisted after adjustment for smoking habits and was higher among men. Smoking habits reduced only up to half of the lung cancer risk of lower SES. Additional adjustment for education further attenuated the ORs. Despite regional differences, lung cancer risks were still elevated especially for the lowest SES categories with exception of women in Eastern Europe. Unemployment was not associated with lung cancer except for subjects who experienced unemployed periods >5 years, and this finding was restricted to men. Strengths of this study are primarily based on the large international SYNERGY database with participants' detailed occupational and smoking histories. Smoking information was nearly complete, which allowed for a detailed control of smoking behavior, as recommended in the literature [14]. The ISCO-coded job biographies permitted the assignment of international validated SES indicators to nearly the entire dataset . Limitations include the validity of the SES indicators: ISEI was developed based on data restricted to men. ESeC was developed for comparisons of European countries. Additionally, ISEI and ESeC are occupational indicators restricted to gainfully employed subjects. Even though we analyzed the influence of being unemployed due to loss of job or periods of illness, we could have missed possible influences of activities outside of the workforce, such as housework, part-time work, retirement, which could have underestimated socioeconomic differences [27]. This concerns not only non-occupationally active periods, but also participants without any gainful employment in their job history who were excluded from the analysis. Unfortunately, for lifetime housewives we did not have information on the husband's occupation for derivation of the SES. We also could have missed effects of early retirement as a hidden form of unemployment. Even though our classification of education was based on an international classification, it generally remains problematic to capture the country-specific implications of time spent in the educational system and corresponding educational attainment. Another limitation concerns residual effects of smoking behavior due to misclassification: Stratification by histological subtypes revealed higher SES risks for the smoking-associated subtypes and reduced SES risks for ADC, which is the histological subtype of lung cancer showing the weakest association with smoking [19]. Furthermore, regional differences as well as elevated risks in the younger female birth cohort in our study correspond to the international patterns of the international 'smoking epidemic' observed with regard to SES and lung cancer [6]. The 'smoking epidemic' describes the historical prevalence of smoking that differed by countries/regions , gender, and SES [28]. We identified elevated risks for male non-smokers, which could be due to our definition of non-smokers that also includes occasional smokers. Measuring smoking in pack-years as cumulative lifetime dose may underestimate the role of smoking duration in relation to smoking intensity [29]. Despite evidence for the accuracy of self-reported smoking habits across various occupations and industries [30], recall bias and differential misclassification of smoking cannot be ruled out. Given the several indications and possibilities for residual effects of smoking, we assume that we rather overestimated the effects of SES on lung cancer. Third, the possibility of selection bias was implied in our analysis because the association between lung cancer and SES was stronger among population than hospital controls. In population-based studies subjects of lower SES tend to show lower participation [31], and case-control studies on lung cancer and SES with population-based controls revealed higher ORs for low SES [12]. SES-related non-response bias, i.e. less participation of cases with high SES and of controls with low SES, was observed in one study which was also included in SYNERGY [32]. However, in our study hospital-based recruitment was mainly done in study centers from Eastern Europe making it difficult to distinguish between region-specific and recruitmentbased effects. Further limitations include that we did not have information on other risk factors for lung cancer, e.g. environmental tobacco smoke [33] or residential air pollution [34]. We analyzed the city size of the last residence as a proxy for air pollution, but in contrast to the assumption of increased associations in more urban areas, we found risk estimates to be reduced. This also included the subgroup of participants < 63 years of age, indicating the absence of a 'mobility' effect among senior citizens. Potential confounders of the association between smoking and lung cancer, which we did not include could have also affected our results in terms of mediator-outcome confounding [35]. An important fraction of lung cancer has been attributed to occupational carcinogens [36], but their role in explaining the association of SES and lung cancer has not been fully disentangled yet [4,37]. We considered occupational risk factors by adjustment for 'list A' occupations and, alternatively, by excluding participants never working in a 'list A' job and did not identify strong differences in the association between SES and lung cancer between these subgroups . However, 'list A' only lists jobs with a possible exposure to occupational carcinogens and does not include information about exposure probability, intensity, or duration. Blue-collar jobs may include occupational exposures which are not included in 'list A'. In contrast to subgroup analyses by 'list A' occupation, we found slightly higher risk estimates for low SES among ever blue collar as compared to workers never employed in a blue collar job. However, blue collar workers also include participants who were not exposed to occupational carcinogens. Finally, the applied concept of SES reflects a variety of health-related circumstances and behaviors, but disregards inconsistencies as well as changes of status. Indeed, we recently analyzed social mobility based on occupational prestige in SYNERGY and observed slightly increased associations between lung cancer and downward prestige trajectories over the work life [38]. Here, we measured SES on the individual level with historical information on occupation and additionally education, but extended concepts of SES should involve the entire life course [39], and include income/wealth and area-based measures [40]. We found that adjustment for smoking reduced estimates for the association between SES and lung cancer by up to 50%. This is similar to the findings of Scottish [3], Dutch [41], and European studies [13], and the results for men in a study from Eastern Europe and the UK [7]. In contrast, in a Canadian study the association between SES and lung cancer disappeared after fully adjusting for smoking habits [14]. In our study, the remaining risk estimates were comparatively higher than in most studies on occupational SES and lung cancer, but similar after adjustment for education [12]. However, we focused on the results without education to avoid over adjustment as education is an indicator of SES in early life that remains stable and determines the following SES indicators such as occupation and income [42]. The extent of reduction of ORs due to adjustment for smoking was distinctly lower when we applied ESeC as compared to ISEI. This could point to the different underlying concepts of SES, implying different exposures and pathways to lung cancer. Additionally, ESeC-especially in the condensed version we applied-as well as ISEI categorize ISCO-codes which comprise a hierarchy of occupational skill levels. Applying three ESeC categories may therefore have led to dilution of effects in comparison to the four ISEI categories. A subsequent possible attenuation between SES categories may also have attenuated the effects of smoking in the ESEC categories. Our analysis of occupational SES was primarily based on the participants' longest held job, which might reflect durations of possible exposures. As the longest job was highly correlated with the last job, and associations with lung cancer were even slightly elevated-in contrast to the first job-, the last job might be an appropriate choice in similar studies lacking complete occupational histories. The lung cancer risk we found for unemployed men was nearly equal to a large study in five Nordic populations [43], which did not control for smoking behaviors. The observed gender differences in the association of unemployment and lung cancer point to different careers patterns of men and women. Our data confirmed the trend of an increased proportion of ADC at the expense of SQCC and SCLC, when comparing diagnosis before and since the year 2000 , and our analysis of histological lung cancer subtypes supported previous findings, which showed that lung cancer risks for low SES were lower for ADC than for SQCC [6] or SQCC and SCLC [8]. Socioeconomic inequalities in cancer incidence are greatest for lung cancer [8] and our study shows that these inequalities were not explained by smoking behavior. To explain the observed excess risk of lower SES groups, approximately 60% of female non-smokers of the two lower ISEI categories would have had to be misclassified as current smokers with corresponding pack-years. However, assuming 90% of misclassification for men, an OR of approximately 1.5 would have remained for low SES. When we additionally classified former as current smokers, still an OR of 1.2 persisted for low SES. This confirms the need to explore the pathways from SES to lung cancer. First, the effect of exposures to occupational carcinogens via job based SES on lung cancer needs to be further studied. Despite minor effects when considering 'list A' jobs in this study, occupational SES directly reflects occupational hazards. Most occupations, such as workers in asbestos production or truck drivers, for which elevated lung cancer risks were demonstrated, were assigned to low SES. As these occupations were traditionally held by men, they may account for the higher ORs for men in this study. This is supported by the reduced ORs for men who never worked in blue-collar jobs. Further, ETS is also a work-related risk factor for lung cancer [44] and could be particularly linked to occupational SES, as smoking prevalence is higher in lower SES groups. Other possible, more speculative pathways can be derived from the association of SES and health in general, because occupational and other SES indicators, mainly education and income/wealth, are interdependent. As shown e.g. for education [45], faster biological aging may be associated with low SES. --- Conclusion Our study showed a persistent SES gradient for lung cancer, even after adjusting for smoking behavior and education. There was some evidence for residual effects of smoking due to misclassification, and at least a part of the regional variance of the association of SES and lung cancer may be explained by these residual effects. Still, the strong associations we found in this study in particular for men emphasize the continuing need for the exploration of the pathways from SES to lung cancer. Clarifying these pathways could then contribute to further understanding of lung cancer etiology and shape prevention approaches. --- S2 Fig. Forest plot of odds ratios by study center. --- Data are from the SYNERGY study, which comprises several single studies, whose authors may be contacted at [email protected]. Others would be able to access these data in the same manner as the authors. The authors did not have any special access privileges that others would not have. --- Supporting information S1

An association between low socioeconomic status (SES) and lung cancer has been observed in several studies, but often without adequate control for smoking behavior. We studied the association between lung cancer and occupationally derived SES, using data from the international pooled SYNERGY study.Twelve case-control studies from Europe and Canada were included in the analysis. Based on occupational histories of study participants we measured SES using the International Socio-Economic Index of Occupational Status (ISEI) and the European Socio-economic Classification (ESeC). We divided the ISEI range into categories, using various criteria. Stratifying by gender, we calculated odds ratios (OR) and 95% confidence intervals (CI) by unconditional logistic regression, adjusting for age, study, and smoking behavior. We conducted analyses by histological subtypes of lung cancer and subgroup analyses by study region, birth cohort, education and occupational exposure to known lung carcinogens.The analysis dataset included 17,021 cases and 20,885 controls. There was a strong elevated OR between lung cancer and low SES, which was attenuated substantially after adjustment for smoking, however a social gradient persisted. SES differences in lung cancer risk were higher among men (lowest vs. highest SES category: ISEI OR 1.84 (95% CI 1.61-2.09); ESeC OR 1.53 (95% CI 1.44-1.63)), than among women (lowest vs. highest SES category: ISEI OR 1.54 (95% CI 1.20-1.98); ESeC OR 1.34 (95% CI 1.19-1.52)).SES remained a risk factor for lung cancer after adjustment for smoking behavior.

Introduction Leprosy is a chronic, disabling infectious disease caused by Mycobacterium leprae, an intracellular bacillus that attacks Schwann cells causing their destruction and leading to severe neuropathies that cause deformities and physical disabilities [1][2][3][4]. Leprosy, an endemic member of the neglected diseases group [5], mainly affects developing countries, constituting a serious global public health problem. In 2015, the global leprosy detection rate was 3.2 cases per 100,000 inhabitants, with 16 countries accounting for 92% of all cases of the disease. India, Brazil and Indonesia reported more than 10,000 cases, comprising 81% of new cases of the disease during the period [6]. Brazil is the country with the second highest number of cases , with a detection rate of 14.06 cases per 100,000 inhabitants in 2015 [7]. The World Health Organization has launched the Global Leprosy Strategy 2016-2020, which aims "Accelerating towards a leprosy-free world", which aims to detect leprosy early and provide immediate treatment to prevent disability and reduce transmission of the disease in the community [8]. This strategy aims to reduce the prevalence of leprosy by improving the capacity of health services to diagnose cases in the early stages of the disease, to provide timely treatment aimed at a cure, and to eliminate sources of infection [9]. One of the main obstacles in the elimination of leprosy concerns the social inequalities faced by people in developing countries, caused by poor housing conditions, low levels of education, low incomes, deficits in health services, and migration to urban centers [10,11]. Endemic areas with unhealthy housing conditions lacking basic sanitation, human agglomeration, the sharing of collective spaces, especially dormitories with shared beds and sleeping nets, are risk factors for the transmissibility and development of the disease [12,13]. So, understanding the social determinants that influence the risk of leprosy transmission is of fundamental importance for the development of actions and strategies that aim to accelerate the process of elimination of the disease in Brazil [14]. For the WHO [14] social determinants of health include individual behaviors and living and working conditions, as well as the relationship between the economic, cultural and social structure. Chaptini & Marshman, based on this premise, grouped social determinants into three dimensions, these being socioeconomic, environmental, and gender determinants [10]. A review of the literature, using the descriptors "social determinants" and "leprosy", found few studies with this focus in Brazil. Among the studies that proposed to test the relationship between social determinants and leprosy risk [15][16][17][18], none were found that considered spatial dependence or incorporated it into their analysis. By definition, spatial dependence is understood to be the tendency that the value of a variable associated with a particular location resembles the value of its neighboring samples more than the rest of the locations of the sample set; it also becomes the concept of neighborhood, which is formulated on the basis of the property of proximity, in which the events that are closer together tend to be more similar to each other than distant events [19]. Thus, a case of leprosy in a given region tends to influence the epidemiological situation of its neighbors, therefore not being a random event. The quantification and appropriate use of methodologies to deal with this characteristic are also considered to be a knowledge gap, since the majority of traditional epidemiological studies tend to ignore this effect. There is also a lack of studies in border areas of Brazil, where these diseases tend to be more critical or more difficult to control due to the fragility of the services in monitoring the dynamics or migration of populations between the different countries [20]. Therefore, this study aimed to analyze social determinants and their relationship to the risk of contracting illness due to leprosy, as well as the temporal trends of leprosy cases in a tri-border region in Latin America. --- Methods --- Study design This was an ecological study [13]. --- Scenario The municipality of Foz do Iguac ¸u, is on the triple border between Brazil, Paraguay, and Argentina , situated in the extreme west of Parana ´, it has a total area of 618,353 km 2 , being the seventh most populous city in the state. It is organized in 327 census tracts, 320 in the urban area and seven in the rural area. The population of the municipality of Foz do Iguac ¸u is approximately 263,915 inhabitants, with 99.2% of the population residing in the urban area [21]. Table 1 presents the main socioeconomic information regarding the municipality. It reveals important social indicators, such as an equality of income distribution of 0.545, which represents an unequal income distribution, and a human development index of 0.751, close to the national average. A critical social condition observed was that only 75.3% of the households had an adequate sewage system [21], which constitutes a risk factor for leprosy [10]. In the context of healthcare, Foz do Iguac ¸u has 388 health facilities; 61 public, 321 private, and six non-profit institutions. The Municipality has 30 Primary Health Units , 25 specialized clinics/outpatient clinics, two emergency care units, three psychosocial care centers, 46 diagnostic and therapeutic support units, 10 mobile units, 44 polyclinics, four hospitals, two Health Surveillance Units and one central laboratory [22]. --- Study population This comprised cases of leprosy that were diagnosed between 2003 and 2015 and residents of the urban area of the city. --- Study variables and data source The study variables selected were diagnosis date, date of birth, sex, race/color, level of schooling, address, new case , pregnancy, clinical form, operational classification , number of skin injuries, and disability grade at diagnosis and at cure . These variables were obtained through the Disease Notification Information System . Regarding the social determinants, data from the Demographic Census of Brazilian Institute of Geography and Statistics were used. These included general characteristics of the household and people, personal income, household income, color/race, age, gender, and level of schooling. Following the theoretical framework of Chaptini and Marshman [10], the social determinants were grouped into three dimensions: a) Socioeconomic determinants: proportion of households with monthly income per capita, proportion of literate people and proportion of people according to race/color. b) Environmental determinants: proportion of households according to the number of inhabitants and housing and basic sanitation conditions. c) Gender determinant: proportion of households without a female resident. --- Data collection Data were gathered from the Health Surveillance Department of the Health Bureau in Foz do Iguac ¸u, Brazil. --- Data analysis Exploratory data analysis was performed using Statistica version 12.0, Software, calculating absolute and relative frequencies for the categorical variables, such as sex, schooling, race, operational classification, clinical form and degree of disability at diagnosis. The continuous variables age and number of lesions were categorized according literature [4,8,9]. The number of cutaneous lesions were classified following the criteria defined by WHO [8], in other words up to five lesions the authors considered paucibacillary and more than five lesions were classified as multibacillary. The incidence rate was estimated considering the number of cases in the population in the middle of year. This was multiplied by the time of observation [7]. The size of population was obtained from IBGE -https://www.ibge.gov.br/. The technique of georeferencing new cases of leprosy by geographic coordinates, according to the address of the residence, was applied. In order to obtain geographic information latitude and longitude, Google Earth Pro open access software was used considering the UTM projections for South America . After conversion of the database to CSV format and its configuration, TerraView version 4.2.2 software was used for the geocoding itself, which corresponds to the linear interpolation of the complete address, to a point in the corresponding segment of the street. Cases with incomplete addresses were excluded from the study. Thematic maps were produced through the digital mesh in the Shapefile extension of the census tracts provided by IBGE. ArcGis version 10.5 software was used for this analysis. The relative risk , or risk of illness due to leprosy, for each census tract was obtained through the Spatial Scan Statistic [23], using the discrete Poisson model, with age and sex as covariables through SatScan version 9.3 software. In this analysis it was defined a window of 50% of the population at risk with the statistical significance tested for 999 interactions. The risk was estimated within a grouping divided by the estimated risk outside and within a given area, calculated based on the expression [24]: RR ¼ c=E½c ðC À cÞ=ðE½C À E½c ¼ c=E½c ðC À cÞ=ðC À E½cð1Þ Where, c is the number of cases observed within the cluster and C is the total number of cases in the data set E[C] = C, with the analysis being conditioned to a total number of observed cases. The thematic maps were developed in ArcGis version 10.5 software. Investigation of the spatial dependence of the social determinants and of the risk of illness due to leprosy was carried out by calculating the Global Moran I Index, using GeoDa 1.6 software. This analysis shows the degree of spatial autocorrelation, where the values vary from -1 to 1; positive values indicate direct correlation and negative values inverse correlation, i.e., dissimilarity among neighbors; the value zero indicates that there is no spatial dependence. The Global Moran I Index was obtained using the formula [25]: I ¼ P n i¼1 P n j¼1 w ij ðz i À " zÞðz j À " zÞ P n i¼1 ðz i À " zÞ 2ð2Þ With n representing the number of areas, z i the value of the variable considered in the area i, z the mean value of the variable in the study region, z j the value of the variable considered in the area z, and w ij the neighborhood matrix . In this type of matrix, units with common boundaries or vertices are defined as neighbors, the neighboring unit being defined as w ij = 1, while elements that have no neighborhood relation are defined as w ij = 0 [25]. To test the association between social determinants and the risk of illness due to leprosy, the authors used bivariate Global Moran followed by multivariate analysis for the variable statistically significant in the first phase. Calculations of the bivariate Moran index were performed using the GeoDa 1.6 software and the permutation of the neighboring attributes approach was used in evaluating significance, as described by Anselin [26]. In the multivariate analysis, the Ordinary Least Squares regression was considered initially, in an attempt to explain the global relations between social determinants and risk. The diagnoses of an OLS model were followed by examination of multicollinearity and the residuals, quantifying the Variance Inflation Factor values, and in cases where VIF was higher than 10, this indicated multicollinearity. The GWR local model was used to analyze the relationship between social determinants and risk of illness due to leprosy for each area . This technique consisted of a localized multivariate regression that allows to make an estimate of the parameters of a regression and then how they have changed and ranged locally [27]. The adaptive kernel was chosen in this phase. The adjusted coefficient of determination and Akaike Information Criterion obtained for OLS and corrected Akaike Information Criterion calculated for GWR were used to compare the OLS and GWR models according to the procedure carried out by Lin and Wen . All these analyses were performed using ESRI ArcGis 10.5.1 and R 3.4.2 software. For the determination of the temporal trend, the annual incidence coefficient of leprosy was considered as the predictive variable and the time the outcome variable . The annual incidence coefficient was converted into a logarithm, which provides the reduction of the heterogeneity of the residual variance of the linear regression analysis [28]. Temporal trend analysis was performed through the Prais-Winsten regression [28], based on the coefficients of annual detection of new cases of leprosy in the period studied. The confidence interval of 95% was used, which results in the Annual Rate of Increase [29]. The rate trend was classified as increasing, stable or decreasing. In a rate with a positive value, the time series was considered to be increasing. A negative rate was considered to be decreasing and when there was no significant difference between the value and zero it was considered to be stationary. These analyses were performed using Stata version 13 software. For all tests in the study, Type I error was set at 5% statistical significance. --- Ethical aspects --- Results --- Epidemiological profile of cases A total of 840 new cases of leprosy were identified, of which 427 were females, 685 were white race/color, 624 were people between 15 and 59 years of age, and 504 were those with incomplete high school education . Regarding the operational classification in Table 2, 648 cases were multibacillary and 192 were paucibacillary. The predominant clinical form was dimorphic , followed by the lepromatous form . With regard to the number of cutaneous lesions, 441 had five or fewer lesions and 342 had more than five cutaneous lesions, with a zero degree of disability . Of the 840 cases included in the study, 765 cases were geocoded . Of the excluded cases, 50 did not have a full address and 25 cases were not located through the Google Earth Pro open access software. Table 4 shows the results obtained from multivariate analysis applying Weighted OLS Regression, where the variables proportion of households with monthly nominal household income per capita greater than 1 minimum wage and proportion of people of brown race were statistically significant associated with risk of illness due to leprosy. --- Social determinants and their relationship with the risk of leprosy Fig 2 shows the maps obtained through GWR analysis. From this figure it is possible to observe the distribution of R 2 and β of the variables. The results of GWR are really clear in the map and confirmed the GWR was more suitable than the OLS and2C. The spatial variations in parameter estimates for variables proportion of households with monthly nominal household income per capita greater than 1 minimum wage and proportion of people of brown race are shown respectively in Fig 2E and2F. Fig 2E demonstrates that negative coefficients values ranged from -0.12 to 0.23, from which it is evident that the risk is lower or nonexistent in areas where the income is higher than 1 minimum wage. Turning to Fig 2F, the parameters were from -0.04 to 0.35, which means there is a positive value in association of brown race to risk of illness due to leprosy. --- Temporal trend Table 5 presents the results of the temporal trend, as revealed through the Prais-Winsten regression, in the annual detection rate of new cases of leprosy. --- Discussion The aim of this study was to analyze social determinants and their relationship with the risk of illness due to leprosy, as well as the temporal trends of leprosy cases in a region of the tri-border region in Latin America. Income and brown race were found to be determinants associated with the risk of leprosy. There was also a decreasing trend in the detection of leprosy in the region of the study, of the order of 4.3% in the period investigated, which was found to be significant, using the Prais-Winsten regression. Initial exploratory analysis of the data revealed that there was a predominance of patients of white race/color, in the age group of 15 to 59 years and in those with incomplete elementary education. These results are similar to those found in other Brazilian studies [20,[30][31][32][33][34][35]. Regarding age, the majority of the patients were in the economically-active age group, which is of concern, since the non-diagnosis of the disease can lead to the development of lesions and further evolve into disabilities and reactional states, which may impact on the family, and wider local, economy [9]. Patients under 15 years of age were also observed in the patient population ; this is a serious finding that indicates delayed leprosy diagnosis and a possible failure in primary care [5,8,12,36,37]. It is interesting to compare this result with Table 4, where the leprosy detection rate was identified as decreasing in the region since 2003. Therefore it is likely that cases of disease are underreported, or not found, by the health care system, which might be due to a weakness of Primary Health Care in active case-finding and decentralization to general practitioners without support from a reference center for the diagnosis and treatment of leprosy [38]. This epidemiological situation is evidence that Brazil is really far from achieving the goal of WHO's "Global Strategy for Leprosy 2016-2020", which aims for zero children diagnosed with leprosy and visible deformities [8]. Also according to the results, most of the patients had incomplete elementary education, which is an important characteristic also found by Chaptini and Marshman [10,18,36] whose studies reported that low levels of literacy are associated with higher rates of leprosy. This relationship is due to a reduction in seeking and understanding clinical information regarding the disease and an association with low income. Santos, Castro and Falqueto [39] also found that people with low levels of education have greater difficulty in accessing health services and understanding health promotion actions and disease prevention measures. Regarding the operational classification of leprosy, the multibacillary form, the most infectious form of the disease, was observed predominantly. This may indicate late diagnosis of the disease in the study region, contributing to the leprosy transmission chain and to the increase in the degree of physical disabilities [2,9]. A prevalence of the multibacillary classification has also been reported in other studies [30,34,35,40,41]. Regarding the association between social determinants and risk of illness due to leprosy, it was possible to detect a negative association with the proportion of households with monthly nominal household income per capita greater than 1 minimum wage. This finding has been confirmed by other studies [24,[42][43][44][45][46]. One important issue that has been shown by the literature is that income is a social protection factor in the development of neglected diseases that are related to poverty, such as leprosy [47]. It can clearly be observed in Fig 2C, where the relation between higher income and lower risk of leprosy is expressed, represented in the dark red color in the mapping. Chaptini and Marshman [10] stated that leprosy is associated with poverty due to a multiplicity of factors, including the effect of lower income and unemployment which reduces access to health services to purchase necessary medicines, poorer housing, and overcrowding of homes. Another aspect that is linked to low income is access to adequate food, since poor nutrition is related to an increased risk of developing leprosy [12,48]. Through multivariate analysis it was found that areas with a predominance of people with brown race/color presented an increased risk for leprosy, which demonstrates the relationship of this disease with this social determinant. Castro et al. [49] also found a correlation between race/color and the risk of the occurrence of leprosy. Biologically there is no evidence that color/ race is a risk for the development of leprosy , however in Brazil these characteristics are more related to social inequality. Due to reasons of history, people of brown race/color have less access to elementary school, high school and higher education; they have lower income when compared to white people and also fewer opportunities to participate in formal work, and to live in houses with basic sanitation [50]. However, there is a strong politics of "quota" in Brazil to improve the access of these people in an attempt to give underprivileged people better chances of getting free higher education and thus access to better opportunities. The politics of quota in Brazil addressed to color/race have changed in recent years, and class criteria have become more acceptable than race for reducing Brazil's social and racial inequalities, however there is much discussion about the appropriate policy solutions for the problem [51]. This study may contribute with this debate once it is evidenced that areas with a higher proportion of brown race/color patients have a higher risk of leprosy. It is important to highlight that the characteristics of the space where these people are living is what is associated to the leprosy risk, which means that is not necessarily these individuals are getting sick or have been affected by the disease. The association was found in the ecological level, a phenomenon that is classically known as ecological fallacy. The results also show a decrease of 4% per year in the rate of detection of new cases of leprosy, which is similar throughout the entire state of Parana ´ [22,52]. This result leads to the question of whether the apparent decrease reflects the actual epidemiological situation of leprosy or possible underreporting. Considering the social inequalities that increase the risk of leprosy found in the study, the second hypothesis is more likely to be confirmed. Because it is a border region, where there is a major flow of people between the three countries, the control of communicable diseases is highly complex. It is difficult to precisely determine the areas of risk due to the very dynamism of this population [20,54,55]. However, the study advances knowledge by highlighting the social determinants of the risk of leprosy, serving as a reference for future studies that include spatial dependence in their structure, notably those that test relationships between social determinants and diseases of poverty, such as leprosy. The elimination of leprosy goes beyond merely knowing what the social determinants of leprosy are, and medical technology for diagnosis and treatment will only have an impact if there are significant advances in social areas. Although there have been improvements in social conditions in Brazil, due to large investment in government programs, this has not yet been enough to overcome the sanitary conditions where leprosy is spread. Control of the disease includes the dimensions of sanitation, living conditions and income, as well as the inequalities that still exist among people of the brown race/color and lower income, as, according to the findings of this study, they are at higher risk of illness due to leprosy. Inclusion policies can be a measure to reduce the differences in risk. Latin America achieved a detection rate of about 2.7 new cases per 100,000 inhabitants, taking second position in the world in terms of cases numbers. Brazil was included in the priority list for leprosy elimination, even though the number of cases detected has substantially decreased during the last decade. It is important to emphasize that in Latin America, Brazil presents the largest number of cases followed by Paraguay and Argentina [56]. An issue is the migration between these countries likely influences in leprosy incidence and prevalence in the region studied, however it was not possible to estimate its impact due to data from Paraguay and Argentina were not available. The study has shown that leprosy is spatially determined, this being relevant for studies that aim to understand the dynamics of the disease using more robust methodologies, mainly through GWR, which resulted in better models than traditional methods . The study advances knowledge about the distribution of the disease in the tri-border region of Brazil, Paraguay and Argentina, presenting evidence for improvement in health policies for control of the disease. Castro et al. [49] reinforce the associative nature of the disease and social inequality, demonstrating that leprosy is not limited to illness process individual, but essentially from social factors accumulated in the life course of a population, since intrauterine growth until later adult life, which is really complex to analyze in a single study. Other issue is that leprosy is not limited to social stressor factors within a single generation but should intertwine biological and social transmission of risk across generations, which also requires more complex and dynamics approach. For future study, it would be interesting to develop a prospective study and also including data from Paraguay and Argentina through a data collect in loco. The inclusion of data from both countries was not possible because all their registers were still handwritten, they do not use computerized systems for monitoring leprosy cases. Based on the findings of the study, it can be concluded that the leprosy elimination strategy must transcend the technological apparatus of the health sector, which is often focused on diagnostic technology and treatment, and must include essential aspects of human development and welfare. The study contains limitations related to the use of the secondary data obtained from SINAN, which may present data instability and incompleteness due to failures to complete the data in the notification form. Another limitation was the difficulty in selecting the variables related to social determinants, these were based on the literature [10,15,43,45,53] but the number of studies using such an approach for leprosy is still limited. The methodology also addresses the issue of areas identified as protection actually being areas of underreporting. --- --- Formal analysis:

Brazil is the only country in Latin America that has adopted a national health system. This causes differences in access to health among Latin American countries and induces noticeable migration to Brazilian regions to seek healthcare. This phenomenon has led to difficulties in the control and elimination of diseases related to poverty, such as leprosy. The aim of this study was to evaluate social determinants and their relationship with the risk of leprosy, as well as to examine the temporal trend of its occurrence in a Brazilian municipality located on the tri-border area between Brazil, Paraguay and Argentina.This ecological study investigated newly-diagnosed cases of leprosy between 2003 and 2015. Exploratory analysis of the data was performed through descriptive statistics. For spatial analysis, geocoding of the data was performed using spatial scan statistic techniques to obtain the Relative Risk (RR) for each census tract, with their respective 95% confidence intervals calculated. The Bivariate Moran I test, Ordinary Least Squares (OLS) and Geographically Weighted Regression (GWR) models were applied to analyze the spatial relationships of social determinants and leprosy risk. The temporal trend of the annual coefficient of new cases was obtained through the Prais-Winsten regression. A standard error of 5% was considered statistically significant (p < 0.05).

reorienting health services and developing personal skills . While there is recognition of, and some research on health inequities as they relate to Black communities globally, often addressing these unique inequities gets subsumed into a general focus on 'marginalized' or 'vulnerable' populations. . It is comforting for some of us to assume that, as a field, we are doing 'enough' because we hold good intentions. But the reality, as demonstrated by almost any measure of health or wellbeing-e.g. life expectancy ; disease prevalence ; maternal mortality rates -is that, however well-intended we are in our approach, health promotion is not affecting meaningful change for Black communities quickly enough. Clearly, it is not sufficient to hold an unspoken intention-to research and publish about health inequities or to craft piecemeal interventions to patch up the divide experienced in Black communities-without centering race as a primary and legitimate determining factor. If we are to make any significant difference, we need to bring to the fore the knowledge, methods and strategies we have honed over the last 34 years. We must move to intentionally dismantle oppressive structures and address the historic injustices of colonialism and slavery, the resulting intergenerational trauma, as well as racist laws and policies, institutional practices rooted in racial biases, and the economic deprivation that continues to harm Black communities. This perspective article has been written through collaborative, intergenerational and transnational discussions among a group of members of the International Union of Health Promotion and Education. We situate ourselves variously-as academics, front-line workers, researchers, students and activists; as new and established members of the formal health promotion field; as differently abled, gendered, aged individuals of various racial identities-who share a commitment to questioning how the field of health promotion has been blind or complicit in perpetuating anti-Black racism, scholarship and practice and challenging ourselves to remain silent no longer. We offer our thoughts with the intention that they are part of an essential conversation within a field such as ours, which is designed to value and uphold human dignity, equality and action. The Black Lives Matter movement emerged out of a historical and urgent need to reflect on the realities of Black communities and their own understandings of how both structural and interpersonal racism experienced daily and for the past 400 years under the many cloaks of colonial rule is unique to Africans, the African diaspora and people of African descent . Through this article, we outline the intersection of social determinants of health and anti-Black racism. We describe how in the first 8 months of 2020 Black communities around the globe have been disproportionately affected by COVID-19, while also having to respond to new instances of police brutality. We assert that the time has come for health promotion to stop neutralizing the specific needs of Black communities into unspoken 'good intentions'. Instead, we offer some concrete ways for the field to become outspoken, intentional and honest in acknowledging what it will take to radically shift how we promote health and wellbeing for Black people. --- Black Lives and the disproportionate death toll due to COVID-19 The impact of the COVID-19 pandemic on the health and economic wellbeing of people around the world has been devastating and even more so for Black communities. The pandemic has exposed inequities that have existed for hundreds of years between the descendants of enslaved Africans and those of European descent. In all contexts tracking such data, Black communities experience disproportionate numbers of infections, more severe complications and higher death rates than other racial or ethnic groups. For example, analysis of federal data in Brazil reported that, of those hospitalized with COVID-19, one in three Black Brazilians had died compared with one in 4.4 White Brazilians . This pattern holds to a greater or lesser extent in other countries with significant Black populations. In the USA, a recent report highlights discrepancies indicating that Black people are 2.3 times more likely to die of COVID-19 than White and Asian people. The ageadjusted rate is 3.7, meaning that younger Black people are dying much more frequently than White and Asian people of the same age . This may reflect the concept of 'weathering', which was originally observed among young African-American women who experience early health deterioration as a consequence of the cumulative experience of ongoing social, economic and/or political exclusion and which may take the greatest toll among Black people in racially charged contexts, who must engage in high-effort coping mechanisms . Canada has not historically collected or reported race-based health data . This practice is being challenged in the context of the COVID-19 pandemic, as various sources have called attention to the disproportionate impact the pandemic has had on Black and Brown communities . In response, jurisdictions have begun to collect race-based data in relation to COVID-19 . A press release from Toronto Public Health of a small sample of respondents for example found that '83% of people with reported COVID-19 infection identified with a racialized group' . Research suggests that Personal Support Workers, one of the groups at highest risk for COVID-19 infection in Ontario, are disproportionately visible minorities and immigrants . Looking at proportions-Black people make up 12.4% of the US population and 23.3% of the COVID-19 deaths . In South Africa, which has the fifth highest number of cases worldwide, COVID-19 is hitting the Black townships harder than areas predominantly occupied by White people . For example, in mid-June nearly 12% of total infections were found in Khayelitsha, a large and predominantly Black township in Cape Town, in spite of it having just 6% of the province's population. In contrast, Stellenbosch , which makes up about 4% of its population had just 1% of total cases . These patterns hold even in country contexts with relatively homogenous populations such as Norway, where the Somali community, especially in Oslo, has been disproportionately affected by COVID-19, accounting for 6% of all cases-or more than 10 times that of the general population . --- Higher risk of infection and complications While the Black community must contend with disproportionate death rates due to COVID-19, the risk of infection is also greater due to increased exposure to the virus. The world over, Black people are overrepresented in jobs that expose them to greater risk of person-toperson transmission on the frontlines-such as health care and personal support work, taxi driving, housekeeping, grocery store clerks, factory and farm workers . It is also likely that Black people in low-income jobs and the large numbers working in the informal sector [still comprising more than 60% of the world's employed population and 86% in Africa ], lack access to paid leave and are unable to stay-at-home when ill or after exposure to an infected person. Workers across the spectrum have been impacted by COVID-19. In Trinidad and Tobago for example, nurses and sanitation workers have been protesting for increased wages, including hazard pay and better working conditions, as a result of the increased risks posed in their day-to-day work due to COVID-19 . African countries with predominantly Black societies have not been spared from racist and capital-driven measures that have weaponized the fears of a community spread outbreak. As the pandemic took its toll on Zambia's economy, reports began to emerge about Chinese business owners who allegedly did not allow local workers to return home to their families and kept them on-site during the pandemic to increase the factory's productivity. Zambian workers were made to sleep in small containers with up to six people on mattresses on the floor. In response to this gross example of human rights violations, factory owners claimed that workers were not being held against their will but that these measures protected them from the pandemic . Housing also factors into increased risk of exposure. Housing in Black communities is a complex modern and historical issue, with wide-reaching health consequences. For instance, in South Africa, township layouts, which are a legacy of the apartheid regime, limit safe housing options within Black communities . Poor Black families are squeezed into overcrowded one-bedroom shacks, where adherence to strict lockdown measures such as social distancing and regular hand washing is nearly impossible . This pattern holds true in other African countries with similar histories of legislated segregation , including the former Portuguese colonies of Angola and Mozambique. Similarly, in the USA, redlining policies, beginning in the 1930s, created unequal geographies through racially discriminatory lending practices, segregation and underresourced tenement housing in predominantly Black neighborhoods and cities. These environments, by design, are more likely to provide unsafe housing , typically defined through higher prevalence rates of crime and homelessness, which are often a direct result of systemic racism and intergenerational poverty-further exacerbated by a lack of dedicated resources for community development . These housing environments and community spaces frequented by Black communities are neglected by the state because of their socioeconomic status and are hotbeds for COVID-19 transmission . Even beyond the compromised conditions of Black neighborhoods, the location of these housing settlements also increases the likelihood that Black people have to rely on public transportation to get to work and commute to city centers, further increasing their risk of exposure . Black people are not only more vulnerable to infection but are also more likely to have complications once infected with COVID-19, due to a lack of adequate care and pre-existing conditions. Those individuals with underlying medical conditions are generally at an increased risk of being hospitalized due to COVID-19 . However, Black individuals often face these medical conditions at a much higher rate than White individuals . In the USA, for example, Black Americans are 60% more likely to be diagnosed with diabetes than non-Hispanic White people ; are more likely to suffer with hypertension with a much earlier onset of symptoms than other racial groups ; and more likely to live with asthma and other respiratory diseases . These complications are reflected in the overrepresentation of deaths due to COVID-19, worsened by the fact that Black Americans face significant barriers to accessing adequate care. Contributing socio-economic factors include a lack of health insurance among Black populations and the biased attitudes of healthcare workers, which routinely leads to symptoms and pain being ignored or undertreated . Many Black Americans have understandably lost trust in the healthcare system because of these negative experiences and developed help-seeking avoidance behaviors, further compromising their care . Health communication messages have also failed to meet the needs of Black communities . For example, the Somali population in Norway, a small proportion of the population, has been disproportionately affected by COVID-19 . They are at greater risk of exposure due to their work as taxi drivers or as health and service industry workers, and their living situations . Still, despite this disproportionate risk, most official information about COVID-19 was presented in Norwegian and English . Galvanized by these gaps in the state response, members of the Somali community mobilized key gatekeepers to initiate a collaboration with the public health institute to develop an inclusion strategy, comprising targeted outreach and the production of knowledge products in Somali . This community response revealed critical oversights in the Norwegian health system and strengthens the humanitarian call for equitable, flexible and inclusive strategies to be applied in emergency responses to ensure equal access for all. In the context of COVID-19, specific guidance has been developed around the dissemination of health information for migrants and refugees in the appropriate languages, given the higher healthrelated risks and vulnerabilities of these communities . Despite this, the challenge still remains in health and other field to disseminate, track, monitor and ensure compliance for protocols that are handed down by development agencies. Ground-up approaches that allow people to become aware of and challenge their own biases, like the example in Norway, build a strong foundation toward lasting change. --- Police brutality, uprisings and COVID-19 On 25 May 2020 in Minneapolis, Minnesota, George Floyd was killed while in police custody, when an officer kneeled on his neck for nearly nine minutes, cutting off the air to his lungs . The horror of this incident is indicative of the rampant brutality experienced by Black people at the hands of police in the USA dating back to the era of slavery . The murder of George Floyd, which was witnessed firsthand and shared globally via social media, prompted uprisings in major cities across the country and solidarity marches worldwide. In the USA, thousands gathered across the nation, braving the risk of contracting the virus, to protest the brutality that countless Black people continue to face. The response from law enforcement highlighted the inhumane treatment Black people and their allies endure when fighting for racial justice-from the coordinated physical attacks on protestors with tear gas and pepper spray, to the destruction of water and first-aid supplies . In the midst of a viral pandemic that directly threatens human respiratory functions, this reaction from the state ultimately jeopardized the capacity of medical facilities that were already overwhelmed. With the unrest sparked by various direct actions of the BLM movement, Black people throughout the diaspora have experienced increased levels of violence, death and maiming at the hands of police. Examples from the Caribbean include the June 27th police shooting of three Black men, Joel Jacobs, Israel Clinton and Noel Diamond, who could be seen with their hands in the air surrendering to the police at the time of their shooting, which was captured by a neighbor on his mobile camera and also picked up by nearby CCTV cameras . These extrajudicial murders triggered widespread protests throughout Trinidad and Tobago that lasted three days. During this time protestors were beaten, tear gassed and, to the country's horror, a 30-year-old pregnant woman, Ornella Greaves, was killed by a stray bullet when an officer allegedly discharged his gun into a crowd of peaceful protestors who were gathered in their own community. And if this story wasn't tragic enough, 7 years earlier in 2013, Ornelle's brother Christopher Greaves was also killed by police while walking home from the corner store . In South Africa, people also took to the streets in the spirit of the BLM movement and in protest of police brutality in townships. The violence with which COVID-19 lockdown regulations were enforced brought the culture of abuse and non-accountability of the South African Police Service into sharp focus. In the first three days of lockdown, three people were killed by police, and within a month, the UN Human Rights Office had received reports of excessive use of force by security officers , including the use of bullets, tear gas, water bombs and whips to enforce social distancing, particularly in poor and informal settlements. Due to mounting pressure from civil society, several court orders were issued to compel authorities to prevent police and army brutality during the enforcement of lockdown measures . Since April, the Canadian Civil Liberties Association has been tracking and visualizing the 'massive and extraordinary expansions of police power in response to the COVID-19 Pandemic and the unequal patterns of enforcement that may arise as a result' through their Policing the Pandemic Mapping Project . The stated aim is to 'highlight COVID-19 related patterns of police intervention to help understand who is being targeted, what justifications are being used by police, and how marginalized people are being impacted' . In Trinidad and Tobago, police arrested scores of lower-income Black people during lockdown, citing the breach of COVID-19 regulations . The intentions of police and circumstances around these arrests have however been scrutinized, given the state and police narrative that gang leaders were attempting to destabilize the country, versus an organic uprising of citizens affirming their right to peace and freedom from violence. This type of discursive reframing, much like the performative paternalism of the Chinese business owners in Zambia and the more visceral brutality of the South African police, all demonstrate how the pandemic has been racialized and politicized, resulting in detrimental effects for Black people's health and safety globally. The overlapping and intersecting realities of Black communities are laying bare the harsh realities that Black people have endured since the dawn of the Trans-Atlantic slave trade, when they were stripped of their right to be considered human. How can we in good conscience assert that this history has no impact on the social determinants of health? Anti-Black racism-historical and current, structural and interpersonal-have unique and devastating consequences on the health and wellbeing of Black people today. The reckoning of historic violence, economic deprivation, under and precarious employment, underinvestment in education, inadequate access to food, environmental injustice-all now manifest in COVID-19 infection, complications and death-are juxtaposed with the profound present-day violence, inequity, and injustice faced by Black communities. The life and death threat posed by police violence has motivated people out into the streets to fight for their lives and to rise up in opposition to generations of structural violence. The response has been more police brutality, violence, and exposure to COVID-19. So the question is, fellow health promoters: what are we going to do about it? And of course I am afraid, because the transformation of silence into language and action is an act of selfrevelation, and that always seems fraught with danger.-Audre Lorde --- Dismantling racism within our profession The structural racism described in this editorial dates back to the slave trade and colonization of Africa and involved not only the theft and forced displacement of people, their land and wealth but also the trojan horses of 'education', 'health services', and religious indoctrination. The 'education' and 'services' ushered in by missionaries and other colonial actors very rarely served African people , but instead were instrumental in stripping away Indigenous identities and knowledge and privileging Western ways of knowing , which has had devastating consequences, for instance, in HIV prevention work and countless other examples. Indeed, chattel slavery was a system of extreme violence, where the Black body was considered property and animalistic, meant only to reproduce for the sake of creating more workers and forced to work under the most dehumanizing conditions. Once African countries achieved independence from this iteration of colonialism-new forms of domination were devised and mainstreamed through neoliberal financial dependence as implemented through economic institutions like the World Bank and International Monetary Fund. These policies included predatory trade agreements, and structural adjustment programs that mandate unfavorable currency manipulation, vulnerable cash crop economies, and the gutting of social and health infrastructure . The first WHO Conference for Health Promotion, organized in Ottawa in 1986, coincided with an expansion of privately provided social and health services in the wake of neoliberal trade and structural adjustment policies that emerged around that time and remain largely intact. As governments were forced to scale back their services, non-governmental organizations swooped in to fill in the gaps-at times, employing health promoters as a part of those efforts. As health promoters, it is important that we understand how the field of health promotion was ushered in along this timeline. Our field is indelibly part of a colonial history and therefore demands that we be transparent, reflective, and explicit in our attempts to actively dismantle these systems . As a profession, we must share our commitment to promote equity and dismantle racism and engage collaboratively far and wide. To echo the title of this article, it is high time for health promotion to become authentic, and outspoken in acknowledging what it will take to radically shift how we advocate, promote and demand inclusion in our field that prioritizes the health and wellbeing of Black people. To that end, we offer the following suggestions to fight anti-Black racism on at least three planes: health promotion policy work to dismantle structural racism; health promotion interventions to fight anti-Black interpersonal racism; and critical self-reflection on anti-Black racism as a core tenet of our professional preparation and ongoing development. --- Dismantling structural racism We have covered many aspects of present-day structural violence and racism as illustrated through the current COVID-19 pandemic. We have described ways in which the historic and modern deprivation of all the prerequisites for health including peace, shelter, education, food, income, a stable ecosystem, sustainable resources, social justice and equity impact Black communities . The precise policies, processes and practices that dictate this deprivation vary from context to context, but the universality of anti-Black racism and its effects are undeniable. Health promotion, the action arm of public health, has honed powerful tools for promoting health through policy change at state and institutional levels and we need to work collectively to harness the full potential of these tools to identify and uproot policies and practices that harm Black communities. Inevitably, what can be unearthed by this process, is a fundamental recognition that the injustices faced by Black communities, influences every aspect of their health and well-being. And so armed with this information, we must act. In nearly every country where Black people exist, there are organizations already working to bring about positive change . With most of the heavy lifting already being done by communities of color, we can find the alignment of our goals and seek partnerships that leverage our skills, resources, and platforms to meaningfully collaborate with existing initiatives, led by Black organizers. Dismantling structural racism requires, among other things, linking health promotion research to decolonization efforts and equality struggles . The research we do influences intersectoral structures and the social determinants of health, through our technical advice to policy makers, contributing to professional preparation, and guiding practice. It is incredibly important that health promotion research asks some fundamental questions: Whose worldview is prioritized in our research? Who funds the research and what does that buy them? Whose questions are being asked? What methods are being used? Who gets to analyze the data ? What knowledge are we creating and who is it for? Are there Critical and/or Indigenous methodologies that might be more appropriate to apply ? Is this particular research addressing questions that will illuminate dimensions of intersectionality of how this issue is experienced? Who is conducting this research and who will benefit from the findings ? We also need to fundamentally question the status quo and what we are willing to accept as 'normal'. During the pandemic, there have been lots of discussions of a 'new normal' or returning to the 'normal' way of doing things. COVID-19 has presented the world with a new opportunity to rethink the utility of the 'ways things were' and who the beneficiaries of these handed down systems really are. Significant harm has resulted, particularly harm to Black communities, in the reinscription of the status quo, as evidenced by premature reopening strategies that have caused disproportionate depredation. Relevant questions to ask during this pandemic, but also more broadly are: Is 'normal' just? Who does 'normal' serve? And at whose expense? As we look at the actions of governments and companies-who seems to be expendable in a pandemic? --- Fighting against anti-Black interpersonal racism The first step to becoming an ally to Black people and practicing anti-racism requires admitting that racist structures exist, are reproduced, and enforced through conscious and/or unconscious interpersonal racial bias. We all live in a world that was built on the subjugation of Black people, regardless of our location . The next step requires the intentional processes of uncovering, naming and preventing the violence that flows from this socialization. Health promotion is not new to such work. There are many examples of antiviolence work in our field [e.g. our use of critical pedagogy in gender-based violence programs ]. It is important for us to recognize that as much as any individual behavior change is part of a health promotion agenda anti-racist education must also be considered a fundamental objective of health promotion practice. Employing relevant strategies such as the settings approach is particularly relevant for teaching anti-racism in schools, universities, workplaces, prisons, etc. We need to build on and develop theories and models that support this work. Some promising examples that health promotion can adopt include Critical race theory , theories of Whiteness and decolonial approaches . We can also learn from the groundbreaking work of Aboriginal and Indigenous scholars who have been promoting health in diverse contexts and challenging harmful Western approaches of exclusion. --- Professional preparation and development to address anti-Black racism in our profession and within ourselves Lastly, it is incredibly important to recognize, as described above, that our field has roots in colonial dynamics and that those dynamics continue in our day-today interactions as professionals. Funding for research and programs is largely provided by former colonizing countries, whose medical and health fraternities may not have always had the best intentions or practices around race research. Eurocentric education and professional training reinforces the notion that Northern/Western episteme and social ideals are superior to other ways of knowing . Not only do funds and ideas from predominantly White countries dominate the field of health promotion, White people dominate our ranks. The lack of representation of Black health promoters in the field reifies racist power differentials and is a fundamental hindrance to appropriately serving Black communities. How can we meaningfully change the dynamics of White dominance if the majority of health promoters are White? This raises important questions for our approach to professional preparation and outreach: How can we recruit and retain Black health promoters? How can we encourage Black students to see themselves in these spaces? Can universities design recruitment campaigns or summer pipeline programs for stand-out Black talent in scientific and other fields? What outreach can we do to engage high school students or to host 'visit the university' days that target schools and communities that are generally overlooked? Can we provide compensation or recognition for Black professionals who mentor Black students and young professionals? What alumni networks exist and how can early career professionals find Black colleagues with whom to connect? We must also look deeply at our health promotion degree programs and curricula. Especially those curricula that claim to explore global contexts, while exclusively focusing on the exploitation and deficits of nations within Africa. These courses often fail to apply an equally rigorous critical lens to nations within the global North, specifically nations with the economic ability to provide comprehensive social services and continue to not do so. We can no longer run the risk of reinforcing racialized power structures that fuel antiblackness in our class or staff rooms. As knowledge producers who also have the power to influence the thought and action of future generations, there are a number of questions we need to ask ourselves, including: who is teaching our classes? What are the topics we privilege over others? Are we examining the histories of the institutions we occupy to better understand the sources of inequity? What authors are we reading and requiring others to read-and where are they from? How are we demanding diversity in our faculties? Who is represented in our leadership within the field? What critical and selfreflective practices are being taught and at which levels? Indeed, self-reflective praxis is important for all of us, but it is particularly crucial for White health promoters, who must be at the forefront of challenging racism in the field. We must be continuously engaged in ongoing self-reflective practices. We must constantly seek to learn about the neglected, buried and ugly histories that have been minimized and erased, largely to enable our retention of positions of power and authority. We must question our motivations, learn about and reflect on our own positionality within the discourses and materiality of White privilege, White fragility, White saviorism, White silence, White exceptionalism, White apathy, anti-Blackness and unconscious bias, which extends beyond the singular category of race . We must also cultivate unflinching honesty and humility and hold ourselves accountable first. These efforts must be ever-present not only in our thinking but in our daily practices as well. Every project, initiative, course we plan must be interrogated according to these critical ideas. We are ethically compelled, as White professionals, to question the hegemony within which our work is situated. Returning to the question of 'normal'-what are the 'normal', taken-for-granted assumptions embedded in our health promotion programs? Is it normal to ignore questions of race? How have 'normal' health promotion programs worked for Black communities so far? Are we willing to do what is required to achieve equity for Black communities? Are we willing to dismantle systems we personally and collectively benefit from? It is important to end by saying that while this commentary has laid out only a fraction of the ways in which Black communities and people have been deprived of fundamental access to health and wellbeing historically, recently, and profoundly in the current moment defined by COVID-19, the intention of this commentary is not to frame this work as some kind of new rescue effort but to affirm the humanity of Black people as a radical and necessary act of solidarity. As health promoters, we must acknowledge that modern society's dominant and postcolonial ways of working, educating, policing, and caring are harmful to Black communities and it is our responsibility to work toward correcting those wrongs. Dismantling these structures requires the re-envisioning and repurposing of existing health promotion tools and approaches but most importantly it requires imagination. We need new ideas, new goals, new knowledge, new models and new leadership-leadership that is not only different from but that directly challenges neoliberal and postcolonial frameworks and ways of thinking. It is time to turn toward the unrelenting strength, resilience, courage and leadership of our Black colleagues in health promotion and related movements and become outspoken that Black Lives Matter in our words and in our deeds. Anything short of that is complicity. Your silence will not protect you.-Audre Lorde

Racism is a public health crisis. Black communities (including Africans, the African diaspora and people of African descent) experience worse health outcomes as demonstrated by almost any measure of health and wellbeing-e.g. life expectancy; disease prevalence; maternal mortality rates. While health promotion has its foundation in promoting equity and social justice, it is clear that however wellintended, we are not affecting meaningful change for Black communities quickly enough. Through this article, we outline the intersection of social determinants of health and anti-Black racism. We describe how in the first 8 months of 2020 Black communities around the globe have been disproportionately affected by COVID-19, while also having to respond to new instances of police brutality. We assert that the time has come for health promotion to stop neutralizing the specific needs of Black communities into unspoken 'good intentions'. Instead, we offer some concrete ways for the field to become outspoken, intentional and honest in acknowledging what it will take to radically shift how we promote health and wellbeing for Black people.

Introduction The one-child policy, i.e., of having only one child per couple, has been the essential family policy in China since its adoption in 1979, and it has effectively controlled the trend of rapid population growth in China. The policy has also significantly influenced China's economic and social transformation through the nation's progress of urbanization and modernization. However, the opposite side of this policy has also been apparent since the beginning of the 21st century [1] and has given rise to the problems of special families under the one-child policy caused by the death or disability of families' only children. In current policy and regulations, the National Population and Family Planning Commission of China defines special families under the one-child policy, or, in another phrase, "the family planning policy", as those families whose only child is disabled or deceased and who have not given birth to or adopted another child. This study used the concept of "special family" according to the official definition raised by China's Health and Welfare Commission, rather than another popular definition--the "Shidu family", an expression used in academic areas to refer to a family that has lost its only child--because the concept of special family includes and is broader than the "lost-single-child family", and not only the "Shidu family" but also the families in which the only child has a severe disability are targeted as recipients of some special welfare policies, and the concept of the special family serves better if it involves those who have unique experiences and interpretations about the implementation of these welfare policies, while we also used the concept of "Shidu family" sometimes in the article according to the context. Some scholars estimate that the number of special families will reach 4.5 million by 2050 [2], while the issue of how to provide them with the necessary support services and solve their difficulties caused by the loss or disability of their only child has gradually become an unavoidable welfare concern for the authorities and the society as a whole. In 2021, the Chinese government introduced a policy allowing couples to have three children, this may solve the problem of low fertility rate to a certain extent, and eliminate the possibility of special families in the future; however, we believe that the three-child policy will not diminish the importance of special family research, but make it more realistic and urgent, because the special families are now at risk of being ignored and forgotten by the society. In 2006, the State Council of the Central Committee of the Communist Party of China issued a decision on "Comprehensively Strengthening Population and Family Planning Work to Solve Population Problems", which explicitly proposed to actively explore the establishment of a support system for families with disabled and deceased only children. On the basis of that decision, in 2007, the National Population and Family Planning Commission and the Ministry of Finance issued the "Pilot Program of National Support System for Families Whose Only Child Died or is Disabled", officially implementing a pilot program that established a support system for special families in some provinces and soon spread throughout the country. Since then, the social support system for special families has been continuously improved, with the contents of the social support being developed from an initial, limited financial compensation and material support to the inclusion of psychological support and spiritual consolation, and with its provision of social support gradually being institutionalized and systematized [3,4]. However, a large concern remains that the welfare demands of special families are not being fulfilled effectively--specifically, that the standards and level of support are a little low, the supply of social support services is insufficient to cover all of the special families in need, and the capacity of social support services is not able to satisfy the demand that special families have for professional social support services [5]. In response, the primary policy recommendations have called for increasing the investment of support funds, raising the standards and level of support, and increasing the number of personnel and the professional capacity of the social support services [6]. The existing research focused largely on the issue of special families from a macrosocial level, with fewer studies having been concerned with the experiences and interpretations of those families. Our research is a qualitative study that we conducted in Jinan city, Shandong Province, in which we attempted to explore and analyze the subjective welfare experiences of the parents in special families. The research question is: what is their welfare experience and how do they interpret it, being the recipient of the special family support policies which were developed specially for them? According to our findings, the special families' support system meets their needs for social support in many ways, but there are still a large number of cases that do not receive the social support they deserve. Some special families refuse to accept certain support provided by the government and society because they feel the services are presented negatively toward their special status, while other families insist on special support arrangements because of their special status, and some of them even insist on special support activities that go beyond what the government and society can provide. From our qualitative analysis of our in-depth interviews, the participants' experiences and interpretations of achieving related welfare policies are shown to be complicated, and we have attempted to explore and understand the complications. On one hand, these special families need to be provided with "special" services, but on the other hand, they are still ordinary older people who also need "general" and "de-specialized" services. Thus, the dynamic between the specialization and de-specialization dimensions of the services approach poses a serious challenge to the authorities' efforts to provide welfare services for special families. --- Literature Review Although other countries also have special hardship families formed by older parents who have suffered the death or disability of their children, and there are some researches on families who lost a child which mostly focused on issues such as poor physical and mental health [7], low economic well-being level [8], and high subsequent fertility [9], special families under China's one-child policy are a unique social problem because the Chinese government has strictly implemented the family planning policy of only one child per couple for more than 30 years. The general two-child policy implemented in 2015 signaled the end of the "one-child policy" in China, which made many one-child families, especially the special families, feel that they had become victims fooled by the one-child policy and were about to be forgotten by the welfare system. This led to some collective rights protection actions of special families with the aim of obtaining welfare support from the government, which aroused the high attention and concern of government and society [10]. In addition, in just the past 20 years, the number of special families has considerably increased, which has attracted attention from academicians, who have mainly focused on the welfare demands of special families and related policies as coping strategies. --- The Welfare Demands of Special Families under the One-Child Policy From the existing knowledge, the analysis of the welfare demands of special families has been presented as a process of development from one simple dimension to a situation with multiple dimensions, and from focusing on material needs to attending to spiritual needs. Researchers gradually have clarified that the needs of special families are multifaceted, multidimensional, and differentiated, rather than being just a simple need for homogeneous economic or material support. Specifically, studies have found that the older parents who had lost their only children asked not only for financial compensation, but also for spiritual comfort [11,12]. The death of an only child brings unbearable psychological trauma to the family, creating needs both for medical services and for spiritual consolation [12]. The social support needs of the special families comprise not only financial compensation and material support, but also social support and old-age care, and the social support needs of the special older adults are different [13]. The vulnerability of the special older adults is manifested in various respects, and their demands are all-round [14]. The social support needs of the special older adults are affected by their gender, age, household registration, marital status, class structure, health status, reasons for loss of independence, and the presence or absence of their third generation [15,16]. Scholars paid attention to the differences in the social support needs of special families in urban versus rural areas and determined that their demands for older individuals' care services tended to increase from rural areas to general urban areas and then, to more developed cities [17]. Rural special families have the highest level of demand for health care services, followed by material and economic needs, followed by spiritual and recreational needs, and then by needs for life care services [18]. Current research analyzed the welfare demands in special families, in terms of targeted needs, such as the need for psychological guidance due to grief over the loss of their child [11,12,19] and the need to strengthen their social capital in response to the breakdown of their social networks [20]. The existing studies about the demands of the special older individuals emphasized their targeted needs as older people without children, and paid less attention to their general welfare needs. Thus, the research findings based on the literature and second-hand studies, with rare empirical studies, raise the urgency for further exploration of the experiences of these special families. --- The Welfare Policies for Special Families under the One-Child Policy In conjunction with the evolution of the welfare demands of special families under the one-child policy, the related welfare policies for special families have also undergone a transformation, from focusing only on financial compensation and material support to providing the families with diversified, stratified, and differentiated types of social support. Some studies focused on how to provide the necessary economic support and social security for special families under the one-child policy [21], while other scholars gradually stressed the importance of establishing a more comprehensive social support system for these special families and have emphasized the need for an economic relief mechanism for them [12,22]. Scholars highlighted that these special families are an unintended consequence of one-child related policies and should receive comprehensive welfare support in response to their needs [19]. In order to provide for the special families' diverse welfare demands, some scholars suggested the establishment of a "five-in-one" mechanism for social support, involving a collaboration of the government, the family, the market, society, and the community [17]. Another group of scholars emphasized that the government should take the lead in establishing a "caring community" model that can build a more compassionate and caring social environment by integrating and directing various public and social resources [23]. Other scholars expressed the concern that interventions should instead focus on age-related issues [24]. It was demonstrated that existing support policies improved the living conditions of special families to some extent [25], while it was also suggested that special support policies overemphasize the special nature of loss of independence, which may have a labeling effect and even lead to extreme cases of over-reliance or non-use of the policies by special families [26]. --- The Narratives of Special Families under the One-Child Policy In addition to investigating the welfare demands of special families and the related policies that exist for them, we reviewed studies on the subjective experiences of special families that focused primarily on those families' identity and marginalization. The sudden withdrawal of their family role at the loss of their only child made the parents unable to orient themselves and created temporary barriers to emergency interactions [27]. Furthermore, the changes in the parents' family structure caused their long-established family identity to disappear and their self-cognition to appear to be chaotic [28]. The Shidu status is a new starting point for special families in their effort to rebuild themselves, and it can become their main identity [29]. Studies highlighted that some special families are more resistant to their special Shidu identities, however, reporting that those respondents give avoidance-type responses to collective emotional compensation for special family groups and to status-targeted benefits such as early retirement [30][31][32]. In the studies of marginalization of special families, some scholars suggested that their marginalization includes both structural and psychological dimensions [29], and that their individual marginalization follows a psychological-structural path in which the special families gradually form a psychologically marginalized identity that changes their thoughts and actions in social interactions, and ultimately affects their interactions with the outside world and individuals, leading to structural marginalization [31]. From the perspective of psychological marginalization, they are desperate and self-isolated due to the influence of traditional cultural concepts such as 'many children, many fortunes' and 'four generations in one family', and they actively separate themselves from their normal families [33,34]. They may also reject sympathy and compassion from others, manifesting themselves as being "disassociated" from friends and community interactions [27]. From the perspective of structural marginalization, special families gradually dissolve their original social interaction system due to the denial of self-worth [27,31]; they gradually develop inward-looking interactions, form in-groups, and consciously alienate themselves from the general group [27,32]. The dramatic decrease in social interaction due to low emotional energy and the accumulation of a large amount of negative emotions contribute to their pathological reputation, which easily invites social rejection [31]. Scholars also noted that the mindset of avoiding bad luck led to the exclusion of special families by normal families, reinforcing the structural marginalization of this group [35]. Some scholars are concerned that the existing studies, with their "suffering narratives" and "social redemption narratives" as the main narrative approach, have labeled special families as stigmatized, thus strengthening their marginalization at the social structure level [36]. --- Research Methods This study adopted an interpretivist paradigm and qualitative research method to explore the welfare experiences of the parents who lost their only child, and we attempted to present the subjective narratives and the inductive qualitative themes of welfare experiences of those special families. This study belongs to a program of "promoting the service system for special families under the One-child Policy" delegated by the National Health Commission. The research team conducted their empirical qualitative study in Huaiyin district, Jinan city, performing 33 in-depth interviews in January and February 2022. Jinan city was chosen as the study site mainly because it is a pilot site of special families welfare policy, which was awarded "the national Warm Heart House Demonstration Site" by the authority; moreover, it is also the location of the study team which provided convenience for conducting the study. --- Data Collection In this study, 33 interviewees were selected by following the purposive sampling method, as we sought to gain a comprehensive and deep understanding of the welfare experiences of special families under the one-child policy. Sixteen participants from special families were selected by following the sampling criteria including: being a citizen living in the Huaiyin district, they have lost their only child. The semi-structured in-depth interviews were conducted, and the questions were centered around the following subjects: the special families' welfare demands for community support services; their status of satisfaction with whether their demands were met, and their further requirements for support services; the current situation, experiences, difficulties, and challenges faced by communities in carrying out support services for these special families. Additionally, the other 17 were providers of related welfare services who could elaborate the welfare experiences of the special families and enrich the understandings of the researchers. They were selected by following the purposive sampling of recruiting the direct welfare providers under the current welfare system, including managers of a planning family support committee for the streets and community, full-time or part-time social workers in the community, related staff members of the District Civil Affairs Bureau, and the person in charge of special family support from the National Health Commission. The semi-structured in-depth interviews were also conducted, the questions focused on the opinions on the current welfare system and policy suggestions for improving support services in the future. The codes of the interviewees and related specific information are shown in Table 1. The study took the appropriate ethical considerations, including requiring all participants to complete an informed consent form, following the "no-harm principle", and keeping all of the participants anonymous. All of the interviews were audio-recorded and were fully transcribed; each interview lasted 1-2 h. --- Data Analysis The qualitative data analysis software NVivo 12 was used for our qualitative interview analysis, and among other things, it provided inductive themes that emerged from the data analysis. The NVivo 12 software has powerful coding, summarization, and visualization functions that are especially suitable when the number of textual materials is large, and that helped to improve the accuracy and efficiency of the analysis and ensured the objectivity and truthfulness of the research findings [37,38]. On the basis of the purpose of the research, which attempted to analyze the welfare experiences of the special families, the grounded theory qualitative method was adopted as the data analysis method. We conducted the grounded theory analysis following the guidelines laid by Corbin and Strauss [39] which included three stages: open coding, axial coding, and selective coding. Specifically, the coding process of this study was divided into the following three steps. First, at the open coding stage, materials from the 33 interviews were imported into the software, the original materials were carefully read and open coded, and the interview materials were open coded as comprehensively as possible, forming 59 free nodes and 1043 information reference points related to the welfare experiences that the special families reported with regard to family planning. Next, at the axial coding stage, the open codes were summarized and the logical relationships were extracted to form summarized categories, such as specific welfare demands, reliance on the government, excluding themselves from normal life, feeling afraid of "specialized treatment", and so on. Then, at the last selective coding stage, we conducted a selective and theoretical analysis for the existing categories, and we attempted to present a deep interpretation and reflection of the interviews and build inductive theoretical themes: the deployment of the specialization and de-specialization dimensions of welfare experiences. The dynamics between those two dimensions were also discussed. --- Findings --- Specialization of Welfare Experiences: A Dimension Characterized as Being Identity-Oriented, Targeted, and Comprehensive There is an old saying in China that refers to "yang er fang lao ". The death of the only child, who was supposed to grow up and assume many family functions, such as supporting the parents when they are old, carrying on the family line, and providing financial support for the family, has an irreversibly negative impact on the structure and function of the family. Compared with ordinary families, the special families in our study tended to need special care in order to maintain a normal life, and we defined those needs as the "specialization" of welfare experiences, which were shown to have the characteristics of being identity-oriented, targeted, and comprehensive. China's one-child policy is said to be "the boldest and largest experiment in population control in the history of the world" [40], meaning that the special status of special families was created in a specific social context. In the context of social identity, an individual's psychological status is closely related to his or her family members' status [41], and special families tend to define their self-perceptions and welfare needs on the basis of their special Shidu identity status. The worry of our Shidu families was that "if something else happens, there is nothing we can do, we have [no] child to rely on. To put it in a bad way, we lost our family, [next] is to rely on the Communist Party, I hope [it will] care more about our care. Auntie Lu called two days ago and said that I still have a big brother to take care of me, but for her, there is really nothing she can do if something happens. This is the biggest worry". . In the temporal dimension, the welfare experiences of special families exist in different forms that have different impacts on their identity construction at different life stages, with traces of transformation and continuity constructed by society during their life course [42]. Taking the loss of their only children as the starting point, the special families continued to play socially prescribed roles and practices in their subsequent life course, which can be divided into the early stage, which is characterized by self-isolation and sensitivity , the middle stage, which is characterized by despondency and requests for changes , and the late stage, which is characterized by a reestablishment of order . The early stage of Shidu: We also have this, [which] is a kind of severe depression, people simply do not want to see people, so we can only [go] through his relatives and friends to understand his recent situation. The middle stage of Shidu: It's definitely not the best, but it's definitely necessary at first. You first have to have a process that makes them want to get out of the house, and after they get out of the house, you can only integrate [them] into normal activities if they see themselves as normal people. The late stage of Shidu: After the first phase of the project, the older individuals saw you with tears in their eyes, but now they are laughing and happy to see you coming from far away. In the past, they used to watch the performance on stage, but now they are performing on stage for you. Before, he was waiting for others to help him, but now he comes out to help others. The trauma of losing an only child or having a disabled only child cannot be healed, and the welfare experiences of special families permeate all aspects of their lives and remain with them forever. In sorting through the data, we found that the welfare demands of special families related mainly to economic support, psychological counseling, medical care, ageing care, and daily life care. The welfare needs of special families were twofold: the comprehensive needs of ordinary older families, and also their own unique targeted welfare needs. According to the data, special families varied in age, health status, and economic level, and they were influenced by multiple factors and had different combinations of needs. Differentiated needs are the source of precise welfare experiences, whereas an adequate connection between the supply and demand sides is a necessary guarantee for subjective welfare experiences [4], thus requiring comprehensive and also targeted services from the government and social organizations. Old age . . . I think you are able to do with what you have, we can move, he can send us food at noon, it is okay. The other [issue] is the future needs, the main thing is that we don't know how to pay for doctor's appointments, and we don't know how to escort people. Although I have this child, he is still small. The last time I almost called 120 at night, I didn't dare to call 120, I had money, but no one taught me how to pay the bill. The second year after the death of the child, this thing hit really big, at that time [I] was as numb as if this thing [was] not true, but after six months and a year, our pain came, just think [ing] of the child [and] those things, feel [ing] so regretful, if anyone can forget? --- De-Specialization of Welfare Experiences: A Dimension Characterized as Being Identity-Denied, Excluded, and Hidden As we mentioned earlier, the loss of a family's original identity as parents brought deep and intense grief that lasted for a long time, and that persistent grief had a serious and negative impact on the physical and mental health of family members. From our analysis of the interviews, we found that the loss of the only child forced special families to face the deconstruction of their original identity and the reconstruction of their existing Shidu special identity. As we discussed above, the special families asked for special and targeted welfare when they were in their special-identity-oriented stage, and it was not expected that they could reconstruct and accept that identity peacefully. Identity-denied, excluded, and hidden characteristics also emerged from the subjective welfare experiences, and we defined this dimension as "de-specialization". There are significant differences among the acceptable ways and outcomes of identity reconstruction in grief [19], and different special families have contrasting attitudes toward and perceptions of reconstructed special identities. The special families are regarded as similar to the "five-guarantee households" and the disabled groups that are included within the government's list of the most vulnerable groups in need of policy support at present in China. The label of Shidu has rashly classified families from different backgrounds into one special category, and the identity of a disadvantaged group is not something that all special families can openly accept. After my husband died, I was left on my own at home. My brother would sometimes come over to take care of me. I can't walk very well, and I don't participate [in] many activities, so I find it quite troublesome. I have not been in contact with people [for special families] before, but I may have had two phone calls with Xiao Chen . The traditional Chinese Confucian culture suggests that "there are three forms of unfilial conduct, of which the worst is to have no descendants", and in ancient times, the premature death of a child was also associated with some taboos and curses in feudal thought [43]. That stigmatized ideology has been preserved in the course of social development, and in community life, that cultural tradition influences the interpersonal interactions of special families with their friends and neighbors. A social worker who provided services for special families mentioned: "He was in city L before, but now he has moved to city H. The reason he moved here is that he wanted to leave his familiar community, and after he came here, no one knew he is a person who lost his only child". During the interviews, some special families were satisfied with their current life and the benefits they had received and did not ask for additional special attention. Differences in the levels of needs were present within the overall group of special families. On one hand, a low level of needs existed objectively for some and was influenced by the economic status, life background, and family structure of some of the special families. On the other hand, a high level of needs of special families can be hidden, with the causes of a "hidden high level of needs" being explained from the individual's and society's perspectives. From the individual's perspective, special families tend to have low self-esteem, and the welfare system provides support for their retirement and livelihood, while on the spiritual level, it also carries the burden of worrying about bothering others. From society's perspective, the culture of the community and the services used to help special families are closely related to the welfare experiences of special families. The emphasis on special status that staff and providers sometimes promulgate raises the risk of creating a secondary victimization for special families, whose dissatisfaction with the welfare that is provided can gradually cause them to tend to hide their true needs and marginalize themselves. Some years ago, some social organizations signed contracts with community committees to organize regular activities for us to participate in, such as spring excursions, crafts, etc., and also organized us to visit their agencies. Those activities were so targeted that only our special families participated, and gradually no one want [ed] to go. --- The Dynamics between Specialization and De-Specialization in Different Special Families, Family Members, and Family Life-Course Periods In the process of providing support and assistance to special families, an overemphasis on the unique dimension of being special families as the whole unit, and a failure to account for the differences among different special families and different family members, can play the role of "labeling" and can lead to the two extremes, in which the families are either overly dependent on welfare and or have become welfare resistant [26]. By examining the interview data, a dynamic between the specialization and de-specialization dimensions of welfare experiences, as presented above, occurred within different special families, and also across the life courses of the family members. --- The Complexity of Welfare Demands among Different Special Families and Family Members All special families experienced the trauma of their only child's death and suffered the ensuing irreversible shocks and traumas in their family structure and functioning. However, as we have noted, significant differences also existed among their welfare experiences, according to our study, and the welfare demands of special families can be divided into two groups: one group of families that are welfare dependent and another group that are welfare resistant. The welfare-dependent special families regarded being special families as the outcome of a collectively created traumatic event during the implementation of the one-child policy, and they believed that the support they lacked due to the loss of their only child should have been compensated by the government and related services. In recent times, due to the general aging of the population, China is no longer limiting the nation's population growth through the fertility policy, and the one-child policy is gradually receding from the historical stage. In the interim, special families continue to make their voices heard through petitions and other means, hoping to receive attention and obtain a version of targeted welfare. We also found that, nationwide, these people were actually a hidden or potential group of petitioners, who felt they responded to the national one-child policy by themselves: "Now, I am not given any other protection policy because of the loss of independence. What about my next generation, the third generation? What will happen to them if they are gone?" Therefore, they have created a kind of petitioning group. In contrast, welfare-resistant special families believed that their needs for welfare services were similar to those of the general older adults, with needs in common that included medical care, ageing care, caregiving, and psychological comfort. Thus, different special families have different welfare needs, just as other ordinary older people do. Special families are ordinary families who have had the misfortune of losing their only children, and they may need some special support to help them through a tough time, but ultimately, they need to go on with their normal daily life. In the process of implementing the special family welfare policies, an overemphasis on their specialness as a group not only ignores the differences among them, but it also makes it harder for them to live their life as ordinary older people, and that can be the cause of their resistance. Community workers considered the concerns of welfare-resistant special families in their services and in planning activities for them, and the workers optimized their support methods without treating special families differently. I never said this is an event for these families specially, I always said that our residents are invited to come. I never tell them that it's for "Shidu" families, but [that] it's a warm heart house, or from our family planning association, or a little something from our street, so they won't be uncomfortable coming. --- Life Course Approach to the Special Families We found that special families often gradually shifted from the specialization-characterized welfare attitudes to de-specialization characterized welfare attitudes, over their life course, with some special families gradually accepting their specialness as a Shidu family and, at the same time, rebuilding their identity as normal older people. In the early stage of Shidu, special families tended to be more inclined to spend time together in similar groups, while in the middle and late stages of Shidu, their social support networks expanded and their identity diversified, so that they were not only special families, they also were part of many ordinary older age groups. This change revealed from their welfare experiences was not only a continuation of the special family's life course, but also a reflection of their improved mental and physical health and the effectiveness of the welfare provisions they were receiving. In fact, at the beginning, these families are inferior, [and] it is certainly more appropriate for such people to be together. They are in the same situation, they can share the pain, there is also more common language, they also huddle for warmth. In the healing process, the real wound healed, they can gradually get over this hurdle. --- As discussed earlier, the specialization and de-specialization dimensions of welfare experiences can coexist within the special families. Special families wanted to both make a difference in and gain a positive influence from normal social networks, while they also regarded themselves as essentially different from other older adults and feared that their negative emotions might become a burden to others. Trying to join new groups is an important step in expanding their social support networks, but the social stigma of being a special Shidu family is invisibly internalized [44]. One particular family member said: "It works to go out all day long, people are not in the same situation as me, [and] I don't bring my kind of unhappiness to that group of people, who can influence me--I'm happy when people are happy, but when I'm not happy, I go home. If I'm not happy, I'm not going to influence people". --- Discussion and Implications This study adopted a qualitative research method to analyze the welfare experiences of special families in Jinan city, Shandong Province. The findings of the study were based on a generalized analysis of interviews with special families and related service providers, which connected the specialization dimension of welfare experiences with identity-oriented, targeted, and comprehensive characteristics, and the de-specialization dimension of welfare experiences with identity-denied, excluded, and hidden characteristics. The study also revealed the complexity of welfare experiences and the dynamics between those two dimensions in different special families, different family members, and different life-course periods. Some special families and some members of special families preferred more specialized welfare benefits which were provided only for them, some special families and some members of special families preferred more de-specialized welfare benefits which were provided for them not just because they were special families but because they were people who needed the benefits, and some special families and some members of special families wanted to receive specialized welfare benefits while refusing to be labeled or stigmatized. There were some special families or members of special families who gradually shifted from the specialization-characterized welfare attitudes to de-specialization characterized welfare attitudes over their life course. The following discussion and implications from the study's findings are categorized into the theoretical and practical domains. In the theoretical domain, the positive social welfare theory from Giddens [45] suggests that "no responsibility means no rights", and holds that responsibility and rights are inseparable. In our study, as beneficiaries of welfare benefits, special families were labeled as "lost and alone", which lead to a distinctive focus of attention on them from the outside world. That excessive attention was a manifestation and unintended consequence of the alienation of these special families' social rights, and it was different from the responsibilities that they were supposed to assume and their own initiative that they were expected to bring into play. Although the different tendencies that accompany specialization and de-specialization are subjective interpretations at the individual level, the stigmatization brought about by the label of "lost only child" is a construct of the mainstream social discourse and cultural environment. Furthermore, the stigmatizing consequences of another label, "dispossessed", are constructs of the dominant social discourse and cultural environment, and have proven to be a transmissible social process [46]. The dynamics between the specialization and de-specialization dimensions and their effects on special families is a social problem arising from the operation of social mechanisms and welfare policies, rather than a personal problem of special families. This study attempted to understand the different characteristics of the two dimensions, and to elucidate the inherent tension between the assignment of a specialized status and that of a de-specialized status, which is better understood as a complex social problem of social welfare policies for disadvantaged groups, involving the concepts of rights, labeling, and stigmatization, than as simply negative consequences for special families themselves. Moreover, the stigmatization of this group should be eliminated on the macro social level. In the process of realizing their rights, special families are forced to bear the consequences of being stigmatized, and this additional result of alienation is neither within the scope of responsibility they should bear, nor should it exist. After the loss of their only child, some nuclear families face a crisis of disintegration, while relatives and friends gradually reduce their frequency of communication, social support networks weaken, and social stigmatization leads to both a real life crisis and a psychological crisis. In the practical domain, the policy makers and service providers should realize that while supporting special families, the implementation of social welfare policies may also label and stigmatize special families and become structural and cultural factors that cause the marginalization and isolation of special families from society. The government should optimize the welfare services designated for the special families, including offering them additional professional service providers, more elaborate service content and processes, and more refined support services. First, we should optimize social welfare policies for special families by taking into account their specialized and de-specialized welfare needs. The policy for special families should fully consider their specialized and de-specialized needs, and the tension between the two dimensions. It should not only provide policies and services to meet the specialized needs of special families, but also provide support and services to meet their de-specialized needs, so as to help them overcome the pain of the death or severe disability of their only child and return to normal life. Second, the process of providing social welfare services to special families should be improved to meet their differentiated specialized and de-specialized welfare needs. Service providers for special families should have a full understanding of, and be sensitive to, their specialized and de-specialized needs, and the tension between the two dimensions. In the process of providing services, they should provide targeted specialized or de-specialized services for different special families and family members, special families in different living conditions and different life course to meet their differentiated specialized or de-specialized needs. Third, a related policy package should be provided to the special families as instrumental guidance, including special social welfare policies available only to them and general social welfare policies for their age groups, and follow their autonomies to choose to receive specialized and de-specialized social welfare services when they need them. This research had certain limitations. First, the study was conducted in Jinan only, and mainly in urban communities, so it cannot reflect the nationwide situations of community welfare experiences of special families, and their needs and interpretations, especially in rural communities. Second, the issue of special families under the one-child policy is a unique problem, to a degree, and the findings presented were based on an understanding of both the specialization and de-specialization dimensions of the families' subjective experiences. Less consideration was given to theories about losing one's child in normal situations, and that perspective should be included in further studies. --- Data Availability Statement: Not applicable. --- Funding: The authors disclose receipt of the following financial support for the research, authorship, and/or publication of this article. This study was funded by the National Health Commission of the People's Republic of China, Grant SK200356. --- Institutional Review Board Statement: All procedures in the study involving human participants were carried out in accordance with the ethical standards of the institution and the national research committee, the 1964 Helsinki Declaration and its later amendments, or comparable ethical standards. Informed Consent Statement: Informed consent was obtained from all individual participants in the study. ---

The one-child policy, i.e., of having only one child per couple, was adopted as the essential family policy in China from 1979, and since the beginning of the 21st century, it has given rise to problems of special families under the one-child policy caused by the death or disability of only children. The existing research focused on the issue of special families from a macro-social level and analyzed the welfare demands and welfare policies of those families, whereas less research has been concerned with the families' individual experiences and interpretations. This study adopted a qualitative research method and conducted in-depth interviews with 33 participants to analyze the welfare experiences of special families in Jinan city, Shandong Province. The findings of the study were based on generalized analyses of the interviews, including the "specialization" dimension of welfare experiences with identityoriented, targeted, and comprehensive characteristics, the "de-specialization" dimension of welfare experiences with identity-denied, excluded, and hidden characteristics. The dynamics between the two dimensions among different special families, different family members, and different periods in the families' lives were also examined. We present a discussion of the study's findings and their implications, categorized into the theoretical and practical domains.

I. INTRODUCTION "The coronavirus was made in military labs", "Do not go to Chinese stores or Chinese restaurants because many products come from the Wuhan area and therefore may contain the virus and you might get infected", "Hospitals are full of infected people, but health officers don't say it openly to avoid panic", "Coronavirus is a big pharma commercial operation designed to sell the related vaccine". These are just few examples of fake news that have been circulating on social media since the health emergency was launched at the beginning of January 2020. Fake news has always existed and will always exist because people love such stories. Umberto Eco, the famous Italian semiotician who wrote "The name of the rose", once said that traditionally the creators of fake news never hurt anyone because they were listened to by just few friends, but he blamed social media to amplify their voice by giving them the opportunity to make proselytes on a global scale. That's why we have many people who believe in a flat earth and in airplanes that release chemicals to infect people, in theories that state that big pharmas create diseases in labs to sell vaccines, in aliens ruling the world, and so on. From a cultural phenomenon, fake news have become a dangerous threat to our society when the focus moved to public and personal health issues [1]- [3]. Healthcare institutions have been trying to warn the public for years that social media can give rise to misinformation, that on social media there are people without any competence who disclose and create false information [4]- [6]. However, if you look for disease information on any social media platform, you will find both interesting discussions and exchanges of views [7], [8] and a different parallel world, having less and less trust in institutions and relying in improvised gurus who use social media to amplify their thoughts [9]- [11]. You will likely find conversations about useless treatments, non-existent diseases, denial of official medicine, apotheosis of alternative medicines, cancers that can be treated with a good mood, diabetes that heals by drinking a glass of hot water at wake-up time [12]. Some of such remedies might rise smiles, but the issue is serious because personal health choices might affect in a negative way the welfare of the whole society [13], with the Covid-19 pandemic being just the latest example. Indeed, the World Wide Web provides an abundant source of medical information and this information has the potential to increase the anxieties of people who have little or no medical training [14]. For example, the increasing number of social media posts that talk about measles vaccinations is decreasing the measles vaccination coverage [15] and vaccine-skeptical websites create communities of people that disseminate misinformation [16]. In various fields, social media are exploited to improve the real-world scenario. Indeed, business intelligence analyses try to identify influencers [17], people's sentiment [18], users' behavior [19], [20], to promote TV programs [21], [22] and even mathematical models to predict the future [23]. In the health sector, although social media are changing the way individuals transmit and receive health related information [24], the voice of the health authorities seems to be dominated by the conversations produced by ordinary citizens. In this context, motivated by the high number of news circulating on the Web and in social applications about the Covid-19 pandemic, we have tried to understand how fake news affects the knowledge that people have about this virus. The research question we asked ourselves is "How much do fake news affect Covid-19 perception?" To address it, we focused on news related to the origin, spread, treatment and effects of the Covid-19 coronavirus. We selected 40 different news, true, allegedly true or blatantly false taken form both authoritative health sources and unknown sources . To understand how people perceive every single piece of news, we designed a 7-point psychometric Likert scale and we asked 293 volunteers of different ages and with different educational backgrounds to express their opinion among: Strongly Agree, Agree, Somewhat Agree, Neither Agree nor Disagree, Somewhat Disagree, Disagree, Strongly Disagree. We defined the Awareness Index to weight every single opinion and provides a degree of the knowledge that participants have on each individual news. Results show that fake-news do affect the knowledge related to Covid-19. For instance, people believe that the virus is due to the Chinese culinary tradition. The study also highlighted a communication problem of health institutions: news related to medical aspects do not reach people, a clear evidence of the wrong or incomplete communication of health authorities. The remainder of this paper is organized as follows. Section II describes the research question and the proposed Awareness Index; Section III shows and analyzes the obtained results. Main findings and conclusions are drawn in Section IV. --- II. THE INVESTIGATION The main Research Question that we address in this study is "How much do fake news affect Covid-19 perception?". We considered news related to four topics concerning coronavirus: • Origin: the origin of the coronavirus; • Propagation: the diffusion of the coronavirus; • Treatments: the treatments against the coronavirus; • Effects: the effects of the coronavirus on the human body. We browsed the Web and collected 10 different news related to each topic: some news were taken from Healthcare authorities and some others were taken from social networks and/or social applications. The selection was made in order to have some true statements and some false statements. As for the latter, we established they were false by searching among authoritative sources: either the news was explicitly denied , or was given no mention at all. It is interesting to note that sometimes news concerning on going studies or theories yet not proven by the scientific community have been divulged even by newspapers as if true . We collected the 40 news that are listed in Table II. We investigate users' perception using a 7-point Likert scale. For each statement, the user is asked if she Strongly Agree, Agree, Somewhat Agree, Neither Agree nor Disagree, Somewhat Disagree, Disagree, Strongly Disagree. We consider the 7-point scale because it is considered the most accurate of the Likert psychometric scales and it gives a better reflection of the respondent's true evaluation [25]. To understand the degree of knowledge of the participants on each individual statement, we introduce the following Awareness Index: AIndex = N i=1 Likert/N (1 ) where news is the considered news, N is the number of participants, U ser i is the i -th participants, Likert returns a value ranging from zero to six. A zero score means the participants has a completely right perception of the news, i.e., if the news is false, the strongly disagree option returns zero, and if the news is true, the strongly agree option returns zero. Conversely, a six score means the participant has a completely wrong perception of the topic reported in the news. Table I shows the values returned by the Likert function. Therefore, the more the value of the Awareness Index is closer to zero, the higher is the participants right overall perception towards the considered news. --- III. PERCEPTION ANALYSIS The invitation to fill the questionnaire has been posted to our Universities forums and to many different Whatsapp Groups. It has been posted on April 11 2020 and it stayed on-line up to April 26 . We had 293 people who answered the questionnaire: 57% female and 43% male. The age group of the participants is the following: 16..20 , 21..30 , 31..40 , 41..50 , 51..60 . No Knowledge 25% --- News Perception Fig. 2. News Perception according to the Awareness Index. Figure 1 shows the Awareness index for each individual news. To deepen the analysis, we consider the following categorization: • Full-knowledge. News with A-Index smaller than 1; • Somewhat knowledge. News with A-index between 1 and 2; • No-Knowledge. News with A-index larger than 2. We divided the 40 news according to such categories: • Full-knowledge news: the category contains 19 news: #1, #4, #5, #6, #8, #10 ; #11, #14, #15, #17, #18 ; #21, #24, #25, #26, #29 ; #31, #37, #38 . • Somewhat knowledge news: the category contains 11 news: #2 ; #12, #16, #19 ; #22, #23, #27, #30 ; #32, #34, #35 . • No-knowledge news: the category contains 9 news: #3, #7, #9 ; #13, #20 ; #28 ; #33, #36, #39, #40 . Figure 2 shows the cardinalities of the three categories. It is to note that, in general, participants showed no really good perception, as only half of the news were correctly perceived: a random choice is expected to give the same result. However, only one fourth was really badly perceived, and there are only two news scored more than three, meaning that no news misled the large majority of participants. In the following, we analyse the news in each category, with a deeper interest in the most critical no-knowledge one. Indeed, understanding why participants ended up with a wrong perception about certain news, might help delivering a better communication to people in the future. Full-Knowledge news: the news with lowest AIndex (below Somewhat Knowledge news: this category contains some of the most controversial news, in particular those that seem to be supported by scientific evidence. The real issue is that scientific knowledge about the virus is continuously evolving, many studies are set up to test different hypothesis on virus origin, propagation and treatments, and very often news report ongoing researches as if they already were confirmed results. Once such studies are concluded, even if they prove their initial thesis to be wrong, people already got a distorted perception. For example, News #30 , appeared on Italian newspapers since the end of April 2020 citing a scientific study conducted at the University of Torino [26]. However, a deeper investigation, conducted at the time we prepared the questionnaire, revealed that the document was a preliminary study that did not undergo any review process yet. At the time we are writing this paper , the peer-review version of that article has been published [27], however the evidence that the news is true is still hypothetical. Indeed, the abstract of the paper states that "higher vitamin D 3 doses might be useful. Randomized controlled trials and large population studies should be conducted to evaluate these recommendations". Another example concerns News #16 stating that scientific studies proved that Covid-19 survives on surfaces. At the time we posted the questionnaire it was only known that the SARS virus did survive on surfaces. However, given that the Covid-19 virus belongs to the same virus family of the SARS virus, hypothesis were that also Covid-19 survives on surfaces. At the time we are writing this paper, scientific results did show that Covid-19 might survive on surfaces , but participants were convinced the news was true even before any scientific solid result was delivered. No-Knowledge news: the news in this category mainly fall into those concerning the origin of the virus and effects of the virus. Participants do not have a clear idea about the origin of the virus: News #3 , News #7 . Likely, the perception has been influenced by an Italian politician who blamed the Chinese culinary tradition of eating alive mice to be the origin of the coronavirus Covid-19 spread [29], [30]. This news likely affected the knowledge related to the first Covid-19 infection. Although science has no explanation, so far, of how the first Covid-19 case happened, participants believe the opposite. Figure 3 shows details of News #9. There is no substantial difference between men and women, whereas there is a difference with the educational background. Surprisingly, participants with a Ph.D. believe in a correlation between the origin of the virus and the Chinese culinary tradition. Participants do not know how the transmission of the virus works: News #13 . So far, as we already mentioned, there are scientific evidences that Covid-19 might survive on surfaces, but no studies refers to clothing, shoes or hair. Likely, the underlying idea is that studies talk about surfaces in general and therefore, participants are confused and derive a wrong knowledge on the topic. Figure 4 shows details of News #13. There is no substantial difference between men and women, whereas the educational background produces different perceptions. Participants with a Ph.D. have a correct knowledge. In general, participants know how to deal with the virus, with the exception of News #28 . So far, science stated that "there is no evidence that walking surfaces are involved in transmission of the virus. Moreover, spraying hypochlorite could increase the amount of harmful substances in the environment. Street cleaning with conventional soaps/detergents is still advised." [31]. Likely, people confuse a simple suggestion with a necessity. Figure 5 shows details of News #28. There is little difference between men and women, and there are large differences among people with different educational background. Although participants with a Ph.D. have a higher knowledge than people with other educational background, the AIndex is remarkable even for people with a Ph.D. Participants have little knowledge about Covid-19 effects: News #33 , News #36 , News #39 , and News #40 . The poor knowledge towards Covid-19 effects concerns very detailed medical aspects. This shows a weakness in health communication. Indeed, very detailed medical aspects hardly find space in the big newspaper headlines or in the main TV news reports. Therefore, people without medical background are often unaware of the consequences of the virus. What really surprises is News #33. Indeed, during the lockdown, almost daily Italian media have repeated over and over again that the majority of patients that contracted Covid-19 did not need hospital treatment, and that a period of isolation at home was sufficient. The lack of knowledge about this news could be due to a lack of trust in the institutions and their communication. Figure 6 shows details of News #33. There is little difference between men and women , and there are no substantial difference among people with different educational background. Also in this case, surprisingly people with a Ph.D. do not believe that most of the people do not need to be treated in hospitals. --- IV. MAIN FINDINGS AND CONCLUSIONS Confusion and ignorance are surely normal, but fake news contributed to confuse peoples' perception on what is true and what is not with respect to Covid-19. In this regard, the study highlighted that men and women behave similarly, whereas it surprised that people with the highest educational qualifications have less knowledge in many of the considered news. The study also highlighted a communication problem of our institutions: news related to medical aspects do not reach people, a clear evidence of the wrong/incomplete communication of either health authorities or politicians. In conclusion, the obtained results showed that, in half of the cases, people have not been influenced. However, some news has managed to breach the general confusion surrounding the virus. A possible approach to fight this misinformation is to increase the presence of health authorities in social channels.

Covid-19 is a virus developed to rule the world" is just one of the many fake-news published on the Web. In this pandemic period, the Web is flooded with real news, allegedly true or blatantly false. To understand how fake news is affecting the Covid-19 perception, we selected 40 news (either true or fake) related to the origin, diffusion, treatment and effects of Covid-19 and we asked 293 volunteers to express their opinion on the truthfulness of the news. Then, we propose an Awareness index to compute knowledge degree of the volunteers. The results highlight a large ignorance on medical news, ignorance that goes beyond educational background. The study highlights the need for Health Institution to enter social media platforms in order to clearly explain what is true and what is false on Covid-19.

ancestry and sociocultural influences . One of the methodologies used for the characterization of the human brain has been morphological neuroimaging analysis, which consists of the implementation of computational analysis methods of brain magnetic resonance imaging , aimed to identify the structural characteristics of the brain, highlighting analysis of volume and area, such as cortical surface area and cortical thickness . Brain volumetry is a measure that includes surface area and cortical thickness ; the former being a parameter of cortical folding and gyrification and the latter a parameter of density and dendritic arborization . Neuroimaging studies have been implemented to identify brain morphometric differences due to educational level , socioeconomic status , gender, and age . Nevertheless, few neuroimaging studies are designed to explore brain morphometric differences related to racial identity. In this sense, it has been reported that African American persons diagnosed with hypertension and cognitive impairment, commonly referred to as a decline in memory and cognition performance, have lower insular thickness compared to White persons with the same diagnosis . Moreover, Isamah et al. implemented a volumetric analysis using magnetic resonance imaging in neurotypical White and African American persons. After controlling for variables such as age, sex, years of education, and total brain volume, they reported that African American participants had a greater brain volume of the left orbitofrontal cortex than White participants. These authors agree that morphometric studies in populations with diverse racial identifications will reduce the under-representation of ethnic minorities as well as the comprehension of the influence of these variables on the differentiation in specific brain structures and the prevalence of neuropsychiatric diseases among different populations. Racial identity has generally been used as a demographic variable and not as a variable of interest in neuroimaging research, which contributes to generalization bias of brain findings based on persons with high educational and socioeconomic status belonging to majority racial groups . Furthermore, studies including minority racial groups are mainly implemented in small samples and in populations with cognitive impairment . Thus, our study aimed to identify morphological brain variability among distinct racial identities in a representative sample of neurotypical young adults. We analyzed brain morphometric data from the Human Connectome Project . Our selection criteria indicate that White and African American racial identities were the most representative samples in the HCP database. In this regard, we expect to identify differences in brain morphometry between people identified as Whites or African Americans. --- Methods In order to access participants' racial identity information, all authors accepted the terms of data used to access restricted data of the HCP database. After the request was accepted by the WU-Minn HCP Consortium, the database from 1,206 participants was downloaded from the ConnectomeDB, a web-based user interface from the HCP . Apart from racial identity information, restricted data included demographic, clinical, psychiatric, and morphometric brain information for each participant. Data were filtered to exclude participants with psychiatric symptoms, substance use and abuse disorders, endocrine disorders, irregular menstrual cycles, neurological abnormalities, and technical issues in the acquisition or preprocessing of their structural brain images. In the filtered database, participants identified as Hispanics were discarded due to unbalanced sample representation between the selected racial identity groups . Beyond this classification, ethnic identity was not considered for further analysis. Racial identity categories referred to Whites and African Americans were taken from the HCP demographic data based on the NIH Toolbox and U.S. Census classification. 1 Three hundred thirty-eight participants identified as Whites [M age = 29.12, SD = ±3.60 , M education = 15.15, SD = ±1.69] and 56 subjects identified as African Americans [M age = 29.25, SD = ±3.62, M education = 14.41, SD = ±1.90] satisfied the inclusion criteria from the filtering process of the original HCP database. Although age [t = -0.2533, p = 0.800] was not significantly different, years of education between groups resulted in significant differences [t = 2.760, p = 0.007]. Moreover, three participants identified as Whites were excluded from the permutation analysis because of missing education and economic income information . Summary statistics of FreeSurfer morphometric measures from the HCP database previously processed by HCP investigators were analyzed . These preprocessing methods consist of a PreFreeSurfer pipeline, which was implemented to preprocess highresolution T1w and T2w brain images for each participant to produce an undistorted "native" structural volume space. The pipeline aligned the T1w and T2w brain images, executed a B1 correction for each volume, and co-registered the participant's undistorted structural volume space to MNI space. Subsequently, a Freesurfer pipeline was executed to divide the native volume into cortical and subcortical parcels, reconstruct white and pial cortical surfaces, and perform the standardized FreeSurfer's folding-based surface mapping to their surface atlas . Volumetric, cortical thickness, and surface area brain measures were grouped by participants' racial identity-Whites or African Americans. Before applying statistical analysis, volumetric results for each participant were standardized by dividing the raw volumetric scores by intracranial volume . Due to unbalanced samples between groups, ANCOVA permutation analyses adjusting for age, sex, education, and economic income were implemented to identify differences between groups for each brain morphometric measure. The estimation of value of ps was based on the criterion in which iteration stopped when the estimated standard error of the estimated proportion of the value of p was less than one-thousandth of the estimated value of p . A maximum of 5,000 iterations were selected for the analysis. Adjustment of value of ps for multiple comparisons were implemented by the family-wise error rate method S4). ANCOVA permutation analyses corrected for multiple comparisons on the same morphological parameters described above were implemented for this subsample. The ordering of the database, data filtering, and statistical analysis was carried out in the programming language R version 3.6.3 mounted on the RStudio software version 1.2.5033. ANCOVA permutation analysis was implemented by the aovp function of the lmperm package in R . The pipeline used for the statistical analysis can be consulted at https://github.com/Danielatilano/HCP_structural_analysis.git. --- Results --- Brain volumetry differences between groups Volume comparisons resulted in significant differences in cortical and subcortical brain structures . Volumetric measures were obtained from a volume-based stream where MRI volumes are labeled to classify subcortical and cortical tissues based on subject-independent probabilistic atlas and subjectspecific measured values of voxels. Anatomical visualization of brain regions with significant statistical volumetric differences is represented in Figure 2. Similar results were found on the paired subsample volumetric measures; nevertheless, after multiple comparisons correction , the bilateral and total cortical white matter, the left cerebellar white matter, the bilateral thalamus, and the anterior section of the corpus callosum maintain significant differences . --- Differences in cortical thickness between groups Cortical thickness results indicated significant differences in frontal, temporal, parietal, and occipital brain regions . Cortical thickness measures were obtained from the mean distance between the white and the pial surfaces of the cortex. Anatomical visualization of brain regions with significant statistical cortical thickness differences is represented in Figure 4. Similar results were found on the paired subsample cortical thickness measures; nevertheless, after multiple comparisons correction , the right banks of the superior temporal sulcus, left cuneus cortex, the right middle temporal gyrus, the right supramarginal gyrus, and the right lateral occipital cortex maintain significant differences . --- Differences in cortical surface areas between groups Cortical surface results indicated significant differences in frontal, temporal, parietal, and occipital brain regions . Cortical surface measures were obtained from the sum of areas of triangles from the tessellation of the brain surface. Anatomical visualization of brain regions with significant statistical cortical surface area differences is represented in Figure 6. Similar results were found on the paired subsample cortical surface area measures; nevertheless, none of the brain regions presents significant differences after applying multiple comparisons correction . --- Discussion Social, educational, and economic inequalities have impacted the health and human rights of ethnic and racial minorities, causing their under-representation in biomedical studies, leading to bias in the effectiveness of clinical treatments and misconceptions of genetic and environmental diseases affecting these groups . According to some estimates, reducing such disparities would have saved the United States more than $ 1.2 billion in direct and indirect medical costs . Even though the White non-Hispanic population has been steadily declining in recent years, African Americans and Hispanic/Latinos only represent 5 and 1% of participation in human research, while Whites represent over 70% . In this regard, racial/ethnic identity is essential to contextualize neurophysiological and neuroimaging results on structural inequities in society . In neuroimaging research, this under-representation bias may be responsible for the reproducibility, generalizability, external validity, and inference crisis in brain research, which exacerbates the disparities and inequalities of minorities in neuroscience . Data sharing and open access to multimodal brain imaging in consortium repositories have been proposed as research opportunities to diminish racial disparities and methodological bias . Consequently, some advantages of using the HCP database are its public accessibility, a large ethnic/racially diverse sample, preprocessing methods, high-resolution structural brain imaging, and demographic and clinical information of participants . Based on the HCP database, our results indicate volumetric brain differences in white matter structures, subcortical regions, plexus choroids, and total subcortical grey matter between participants identified as African Americans and Whites. Differences in subcortical brain volumetric regions were identified in the bilateral caudate, left thalamus, right globus pallidus, and right ventral diencephalon. Moreover, differences were identified in other brain structures, such as the optic chiasm, the white matter of the right cerebellum, and the corpus callosum in their anterior and posterior portions. In contrast with Isamah et al. 's study, where differences in bilateral amygdala and total cerebral volume between persons identified as African Americans and White were found, we identified volumetric differences in the bilateral caudate and total cortical white matter. Differences in regional brain volumes in cortical and subcortical structures, such as the bilateral caudate, have been identified between White and Chinese populations . Moreover, brain differences in total cortical gray matter volume, total cortical white matter volume, total gray matter volume, estimated intracranial volume, and cortical regional volumes have been reported between Indian and White persons . Furthermore, our results indicate surface area differences in frontal, parietal, temporal, occipital, and frontal brain regions between African American and White racial identities. Specifically, cortical thickness differences were identified in the bilateral cuneus cortex, left fusiform gyrus, bilateral Permutational ANCOVA brain volumetric results between Whites and African Americans with significant differences after applying multiple comparison correction test . Brain volumetry is standardized as the ratio of cubic millimeters/intracranial volume . CC anterior: anterior subregion of corpus callosum. CC posterior: posterior subregion of corpus callosum. WM, white matter. GM, gray matter. Asterisks indicate significant results at a value of p of <0.001 and a value of p of <0.05. --- FIGURE 2 Brain regions representing volumetric differences between Whites and African Americans. CC, corpus callosum; WM, white matter. GM, gray matter. Brain images were created with BrainPainter software . 10.3389/fnint.2023.1027382 Frontiers in Integrative Neuroscience 06 frontiersin.org Jha et al. identified cortical thickness differences in the bilateral postcentral gyrus, superior parietal lobules, precuneus, supramarginal gyrus, right precentral gyrus, insula, inferior parietal lobule, supplementary motor area, and rolandic operculum in a large cohort of neonates of African American and White mothers. Furthermore, in middle-aged cognitively impaired hypertensive persons, differences in insular cortical thickness were identified between African Americans and White people . Similar to our results, Kang et al. identified differences in surface area and cortical thickness in frontal, parietal, temporal, and occipital subregions; however, these results were based on an analysis of brain surface morphometry between older Chinese and White adults. The U.S. Census has created racial categories that include White and African American people, allowing the self-identification of individuals in groups that represent their community and cultural background . In a segregated society, racial identity has emerged as the sense of collective identity based on a perceived common heritage with a racial group , promoting wellbeing and protection against racism in African Americans . In this sense, Afro-American identity is constituted by an African conscience that establishes behaviors, spirituality, and ancestral knowledge affecting self-concept, selfesteem, and self-image. Moreover, racism and oppression, rooted in a historical background of environmental and interpersonal adversity, have caused a mental and physical pathologization of their identity . In contrast, White American identity is rooted in social and economic privileges that establish racial attitudes, beliefs, behaviors, and experiences in a racially hierarchical society . From this perspective, racial identity is defined and addressed as a social construct from which racial groups are socially created to attach differences between groups . In this sense, the descriptive results in our sample related to years of education indicate that participants identified as African American reported less years of education than White participants; moreover, Whites tend to report higher economic income than African Americans. These results may reflect the inequalities in education and socioeconomic status between White and African American people. Low socioeconomic status has been associated with reduced cortical gray matter thickness in middle-aged persons . In addition, diverse studies have indicated that socioeconomic status and parental education strongly influence cerebral cortical thickness, surface area, and volume during childhood , particularly average cortical thickness in neonates of African American mothers . Although our analysis was adjusted for economic income and education, these are only dimensions of socioeconomic status that also imply prenatal and postnatal factors such as biological risks , psychosocial stress, variability in cognitive and linguistic stimulation, and parenting practices during childhood . Our results referred to differences in volume, cortical thickness, and surface area in diverse brain regions between distinct racial identities may be due to these prenatal and postnatal factors anchored in racial inequalities. In this FIGURE 3 Permutational ANCOVA cortical thickness results between Whites and African Americans with significant differences after applying multiple comparison correction tests . Asterisks indicate significant results at a value of p of <0.001 and a value of p of <0.05. --- FIGURE 4 Brain regions representing cortical thickness differences between Whites and African Americans. Brain images were created with BrainPainter software . Atilano-Barbosa and Barrios 10.3389/fnint.2023.1027382 Frontiers in Integrative Neuroscience 08 frontiersin.org regard, it has been reported that African Americans, compared to the White population, have a higher risk of developing Alzheimer's disease due to exposure to air pollutants , access to healthcare and educational disparities . Moreover, racism and discrimination have been related to higher levels of blood pressure , preterm infant birth , and stressful life experiences . Furthermore, the recent study by Fani et al. identified that racial discrimination experiences of Afro-American women were associated with functional activation of the middle occipital gyrus, ventromedial frontal cortex, middle and superior temporal gyrus, and cerebellum. Assari and Mincy have reported that racism may impact the volume brain growth of African American children. Accordingly, with these studies, the morphological variability identified between White and African American identities in our study may also be related to racism and oppression, mostly affecting the African American community, due to historical racial segregation . In this regard, acknowledging inequalities in education , health , justice , and socioeconomic status between Whites and African American people is fundamental to acknowledge that racial identity implies social and environmental factors that can impact in human development and brain morphology . Most studies in human cognitive neuroscience come from majority identities, such as the White population, in contrast to Hispanics, Asians, and African Americans, who have been markedly underrepresented . In this sense, our results suggest brain morphological variability between overrepresented and underrepresented samples, supporting the . Accordingly, it is important to consider the human brain as a multilevel ecological system that regards social and biological factors from which it is necessary to develop crosscultural sampling methods and multidisciplinary collaboration to improve the generalizability of neuroscience studies and the comprehension of individual differences in the human brain . Neuroimaging research groups have developed structural MRI brain atlas and templates based on specific populations due to differences in brain morphology while contrasting with WEIRD samples . Permutational ANCOVA cortical surface area results between Whites and African Americans with significant differences after applying the multiple comparison correction test . Asterisks indicate significant results at a value of p of <0.001 and a value of p of <0.05. --- FIGURE 6 Brain regions representing cortical surface area differences between Whites and African Americans. Brain images were created with BrainPainter software . Atilano-Barbosa and Barrios The African Ancestry Neuroscience Research program has emerged as an initiative to reduce health disparities in the African American community and to promote focused brain research in this population to treat brain disorders by developing personalized therapies and treatments . The evidence of morphological brain variability in our study could contribute to understanding brain disorders and psychological factors affecting African Americans and the prospect of developing brain templates for this population. Although our study was based on a large sample from the HCP database, some limitations must be considered. First, the sample is unbalanced due to the overrepresentation of persons identified as Whites compared to persons identified as African Americans according to the original HCP database. 2 Even though the HCP project is focused on neurotypical young adults, this database includes participants with heavy consumption of tobacco, alcohol, and recreational drugs . Moreover, we identified participants with psychiatric symptoms, endocrine disorders, irregular menstrual cycles, and neurological abnormalities, as well as technical issues in the acquisition and preprocessing of their structural brain images. In this sense, we consider implementing exclusion criteria to discard these confounding variables that could affect morphological brain results in large neuroimaging data . Nevertheless, these considerations maintain the imbalance of our sample between Whites and African Americans persons, which reduces the possibility to apply parametric statistical analysis . In this regard, we implement a method of sub-selection of persons identified as White and African American paired in age, sex, economic income, and education to overcome the confounding bias. Finally, racial identity was defined from the self-identification of participants. However, genetic ancestry information could have contributed to a more careful characterization of the sample from which specific genetic sequences and gene/environmental interactions could be analyzed to further interpret brain morphological results . --- Conclusion The human brain is constituted in a unique genetic, social, and experiential domain that is embedded in global hardships such as poverty and discrimination . In this regard, morphological brain differences in persons identified as African Americans and Whites may be embedded in historical inequalities, oppression, and racism in American society that may impact brain structure. In this study, white matter, forebrain, midbrain, and hindbrain structures display morphological variability between racial groups which could be relevant for understanding neurological or psychiatric disorders differentially affecting these 2 https://wiki.humanconnectome.org/display/PublicData/Summary+Demo graphic+Data+for+HCP+Young+Adult populations. Due to the recurrent misrepresentation of ethnic and racial minorities in neuroimaging research, their inclusion in further studies is fundamental for the comprehension of human brain morphometric variability. --- Data availability statement Publicly available datasets were analyzed in this study. This data can be found at: https://wiki.humanconnectome.org/display/ PublicData/HCP+Wiki+-+Public+Data. --- Ethics statement The studies involving human participants were reviewed and approved by Data Restricted Access by the Washington University -University of Minnesota Consortium of the Human Connectome Project . 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/fnint.2023.1027382/ full#supplementary-material

In a segregated society, marked by a historical background of inequalities, there is a consistent under-representation of ethnic and racial minorities in biomedical research, causing disparities in understanding genetic and acquired diseases as well as in the effectiveness of clinical treatments affecting different groups. The repeated inclusion of small and non-representative samples of the population in neuroimaging research has led to generalization bias in the morphological characterization of the human brain. A few brain morphometric studies between Whites and African Americans have reported differences in orbitofrontal volumetry and insula cortical thickness. Nevertheless, these studies are mostly conducted in small samples and populations with cognitive impairment. For this reason, this study aimed to identify brain morphological variability due to racial identity in representative samples. We hypothesized that, in neurotypical young adults, there are differences in brain morphometry between participants with distinct racial identities. We analyzed the Human Connectome Project (HCP) database to test this hypothesis. Brain volumetry, cortical thickness, and cortical surface area measures of participants identified as Whites (n = 338) or African Americans (n = 56) were analyzed. Non-parametrical permutation analysis of covariance between these racial identity groups adjusting for age, sex, education, and economic income was implemented. Results indicated volumetric differences in choroid plexus, supratentorial, white matter, and subcortical brain structures. Moreover, differences in cortical thickness and surface area in frontal, parietal, temporal, and occipital brain regions were identified between groups. In this regard, the inclusion of sub-representative minorities in neuroimaging research, such as African American persons, is fundamental for the comprehension of human brain morphometric diversity and to design personalized clinical brain treatments for this population.

Introduction This article will draw on the work of Pierre Bourdieu to suggest a new theory of purity and impurity. These are powerful themes, which have been acknowledged to be 'under theorised' . The central claim that will be put forward is that purity and impurity should be interpreted as an assessment of the correspondence of a phenomenon with its essence. A pure person or thing is, in short, self-identical. The Zinc in front of you is pure if it is completely comprised of Zinc, as one of the basic Mendeleevian elements. The citizen in front of you is racially pure if you believe that he instantiates solely the nation, as a biological and cultural essence. To jump for pure joy is to jump in a way that manifests no other animating principle, that corresponds simply and truly to the emotion 'joy' that stands behind and within its instantiations. What makes purity and impurity such significant discourses, Bourdieu's writings show, is that they can be used to performatively construct the essence against which they merely appear to compare phenomena or forms of subjectivity. To call you a 'dirty slut', for example, is not simply a categorisation of you as an individual, but also imputes an essence to femininity from which your behaviour indicates you have departed. Perhaps the main theory of purity and impurity currently available to scholars in the social sciences and humanities for considering the topic is Mary Douglas' Purity and Danger. Douglas claims that classifications of purity and impurity are activated by an innate ordering mechanism in the human mind and help to preserve the social order of society as a whole, by marking that which is 'anomalous' to this order as impure. Douglas suggested that themes of impurity become attached, as a 'spontaneous by-product' of an instinctual human drive for order, to that which is 'rejected from our normal scheme of classification' . She presents a functionalist account, that this process operates for the overall 'advantage of society at large' . Yet there have been repeated calls for further theoretical reflections on the topic of purity and impurity beyond Purity and Danger, including from Douglas herself. Since the publication of Purity and Danger, a variety of scholars across the social sciences have criticised Douglas' theory for making unmediated explanatory links between categorical systems and the social structure of society as a totality. Valeri posits, for example, that bypassing the realm of social practices leads Purity and Danger 'to speak of "the system", of "form", or of "order" as if they were one monolithic thing'. In fact, 'there are many coexisting orders of classification; what is residual to one may be central to another'. Each of these orders of classification is marked by a dynamic interplay of relations of power. Purity and impurity are not the result of a single homeostatic process . Hetherington also criticises Douglas' theory in ways that that agree with Valeri. He argues that the anomaly theory of purity and impurity offers a flawed analysis of processes of categorisation. This has been largely because Douglas presumes that the classificatory system is 'a stable and representable thing', and prior to the anomalies that it designates. As a result, she 'misses the ongoing way in which order is made as uncertain process ' . Douglas herself agreed with her critics that the two central issues with the anomaly theory which cause its flawed account of power and categorisation were the assumption that society is a unitary whole, and the assumption that there exists a unitary 'ordering' mechanism within the human mind. In an informal seminar in 1997, Douglas retracted both assumptions. Furthermore, as a logical consequence of these retractions, Douglas further admitted that there is no universal, 'intrinsic value to purity' for either the individual or for society: 'the only thing universalistic about purity is the tendency to use it as a weapon or tool' . Douglas herself concluded that there exists a pressing need for further social theoretical reflections on the topic. The subsequent year, Douglas publically retracted the claim that there is a universal human or societal urge for a complete and ordered cognitive system. She asserted that the most common response to social or cognitive anomalies tends to be simply to 'tolerate ambiguity' . Taking a more historicist approach than Douglas to the topic, in Moral Purity and Persecution in History , Barrington Moore Jr. insists upon the foundational importance of themes of purity and impurity to the form taken by Western civilisation. He suggests that, from their biblical origins, these concepts have indelibly shaped Western society, contributing to intolerance and extremism through the absolutist world-view that they mandate. He expresses exasperation that, given their pressing importance, purity and impurity have received so little attention from social scientists and social theorists. He contends that the Western notions of purity and impurity have served as an 'indispensible' antecedent to moral catastrophes such as the Wars of Religion in the sixteenth century and the rise of Nazism and Stalinism in the twentieth century . Whilst we do not wish to follow Moore in suggesting that discourses of purity and impurity have endured over millennia at the beating heart of a Western civilisation, we do think that they have great importance for many modern Western societies. In this article we will present a new interpretation of Bourdieu's writings, excavating an account of purity and impurity. Other theorists could have been chosen. Elsewhere one of us have examined the work of Kristeva, Girard and Agamben, considering their contributions to the social theory of purity and impurity. Other contemporary thinkers have also looked for resources for theorising the topic of purity and impurity. Karakayali , for example, recommends Simmel's work on the figure of the 'stranger' as apposite to the study of representations of dirt and impurity. We have chosen to engage in an interpretation of Bourdieu here because of what we see as his particular acuity in examining the role played by discourses of purity and impurity in the differentiation of both fields of practice and forms of subjectivity. He does not examine the stark dichotomy between 'us' and 'strangers', as do Kristeva, Girard, Agamben and Karakayali. Instead, he explores the incremental classification of subjects, embedded within a plurality of fields at any given time, as pure or impure. Prior commentators on Bourdieu have failed to draw out his insights into this topic, however; as Fowler rightly states, these insights often remain at the level of 'hints' in Bourdieu's texts. Yet, as we aim to demonstrate, interpretation could productively bring this theme into focal attention. --- Purity, impurity and practice In 'The Genesis and Structure of the Religious Field', Bourdieu presents a general account of the origin of purity and impurity discourses. He hypothesises that appeal to a realm that transcends ordinary life initially occurred through 'magic' before the advent of institutionally organised religion. He follows Weber in defining magic as the direct attempt to master nature through symbolic practices for the purposes of immediate utility : 399-400, 403). He suggests that in early agricultural societies, everyday life was firmly embedded in nature, with its cyclical patterns, spatial dispersion and immediate needs. This worked against the production of abstract rationalised thought, and hence against a separate cultural field for organised religion. Drawing on Marx and Engel's description of structural differentiation in The German Ideology , Bourdieu proposes that the emergence of the city and with it an early form of the division of labour meant that 'consciousness is in a position to emancipate itself from the world and to proceed to the formation of 'pure ' theory, theology, philosophy, morality, etc.' . Following this line of reasoning, Bourdieu theorises that appeals to moral purity are facilitated by the material and practical differentiation of social fields. Yet following Weber : 536-40), Bourdieu treats the relations between religious professionals, between professionals and the laity, and between the laity, as crucial for the production of discourses of moral purity and impurity. Whereas functionalist perspectives treat purity and impurity discourses as a unmediated reflection of society as ordered or in chaos, Bourdieu argues strongly that this position misses out the role played by social practices within relatively differentiated fields as a crucial intermediary factor . According to Bourdieu, the 'systematisation and moralisation of religious practices and representations' to form organised religion as an arena of social practice, was associated with a 'transfer of the notion of purity from the magic order to the moral order' . Whereas magic had operated through the unsystematic enactment of rites and prohibitions which appealed to purity and impurity, the relatively autonomous religious field became structured by a continual discursive appeal to the essence of the world. Modelling his account on Chapter 7 of Weber's Ancient Judaism, which describes the 'monopoly' the priesthood possess on access to salvation, Bourdieu argues that the priesthood have a monopoly on the legitimate appeal to God. This monopoly operates through theological discourses appealing to purity and impurity as symbols of sin and redemption, which grant the priests a continuous authority over the laity : Religion contributes to the imposition of the principles of structuration of the perception and thinking of the world, and of the social world in particular, insofar as it imposes a system of practices and representations whose structure, objectively founded on the principle of political division, presents itself as the natural-supernatural structure of the cosmos . The religious field comes to define itself through the 'repression of worldly interests ', such that these factors continue to operate, but in a masked form that is understood to express a 'purer meaning' . Bourdieu's analysis of the religious field suggests that the particular appeals that facilitate purity and impurity discourses are those that impute an 'essence' as the ground for reality. Comparing the artistic and religious fields, Bourdieu, for instance suggests that both are places in which competition occurs over the control and definition of phenomena taken to express 'the origin, the spirit, the authentic essence' . Bourdieu is attuned to the fact that purity may mean different things in different contexts: there are a variety of essences against which to compare phenomena . However, Bourdieu's writings also indicate certain commonalities between regions of practice where essence is invoked. We need to be careful in suggesting such an interpretation; Hacking has pointed out that the term 'essence' has become a 'slur word', making its meaning unclear. The essences that we wish to identify as particularly associated with the operation of purity and impurity discourses do not contain any heterogeneous, foreign or inferior elements; all of their elements are 'the same' in some relevant sense. Furthermore, the essences in question are understood to be situated at the conceptual or ontological ground of their instantiations. Besides his consideration of the religious field, where this account first appears, this is especially visible in his analysis of the field of cultural production and consumption. Written in 1971, the same year as 'Genesis and Structure of the Religious Field', Bourdieu's essay 'The Market of Symbolic Goods' applies this analysis of the relatively differentiated religious field to cultural practices more generally. Like the religious field, spheres of cultural practice are separated in Western societies as relatively autonomous realms through the development of a 'public of potential consumers' and 'an ever-growing, ever more diversified corps of producers', as well as 'the multiplication and diversification of agencies of consecration' . In particular, Bourdieu documents the discourses of purity and impurity that arose as the result of the emergence of modern high culture, the domain of elite cultural production and consumption. As techniques of mass production proliferated, in 'reaction' to 'the pressures of an anonymous market', from around 1830 the Romantic movement in Europe proposed an 'ideology of free, disinterested "creation"'. The artist came to be constructed as a pure 'creative genius', as were those dominant consumers 'whose understanding of works of art presupposes' a similar elevated and purified form of existence . Cultural goods that position themselves against the popular, 'are 'pure' because they demand of the receiver a specifically aesthetic disposition' . They are rare precisely to the degree that the disposition and knowledge to decipher their meaning is rare, and valuable because they serve to consecrate more broadly the social and political power possessed by those with this disposition. The classificatory social and cultural structures of the field of practice thereby serve as a 'system of conservation and cultural consecration' separating different consuming publics from one another in lifestyle, and seemingly in their very nature . Thus even sexuality -which Bourdieu sees as generally coded as impure in the first instance -is an effective site for demonstrating elevation and refinement of taste for those whose cultural capital permits such displays, for example in the form of an elevated 'pure love' or a 'purely' aesthetic relation with an artwork with a sexual theme . The products of the field of cultural production can be mobilised as symbolic capital: cultural resources that serve to tacitly naturalise relations of cultural and material stratification in society . The strong association between symbolic capital and purity in Western societies is caused by the fact that a common way in which these cultural resources naturalise relations of power is through making forms of subjectivity appear to be no more than the expression of or 'phototropic' orientation towards essence -such as adoration in the religious field or in love relations, or knowledge in the educational and academic fields. Though Bourdieu's main focus is social class, he also explores purity and impurity discourses in relation to gender. Generalising from his fieldwork in Algeria in the 1950s to contemporary Western societies, Bourdieu suggests that the 'demeanour which is imposed on Kabyle women [in Algeria] is the limiting case of what is still imposed on women, even today, as much in the United States as in Europe'. Bourdieu suggests that women are situated as a 'source from which impurity and dishonour threaten to enter', with the purity of cultivated feminine nature always at risk of being breached by either an inner tendency or an external vulnerability to impurity. The construction of the feminine as either pure or impure legitimates masculine possession, protection and control of women to ensure that impurity does not enter; masculinity is situated as relatively pure -and this relative purity serves as a tacit norm against which the purity or impurity of women are compared . Yet alongside such generalisation, in Bourdieu, is recognition of the way that discourses of purity and impurity have been marked by the particular history of Western societies. Bourdieu suggests that purity and impurity should be understood as 'concepts originally developed in the theological tradition' of Christianity, which have been strategically 'imported' into the operation of other Western cultural fields in ideals of beauty, truth, morality, the natural, and so on . These re-worked theological notions serve a concrete need in the present, 'converting the theology of grace into a worldly, "society" ideology of good grace' and thus pinning moral responsibility for social success or failure on to the individual subject . --- Bourdieu's advance Douglas theorised purity as an expression and affirmation of social order, and impurity as an expression of 'matter out of place'. Bourdieu, however, argues explicitly against the 'reductionist theory' which treats 'discourse' as 'the direct reflection of social structures' contained in society, conceptualised as a single whole . On our interpretation of Bourdieu, purity and impurity discourses are not the reflection of social structures: mediating between the discourses and embedded social forms are the strategic discursive practices of different relational subjects, acting within fields in which competing visions of 'the origin, the spirit, the authentic essence' are precisely a stake in this interaction . In the cultural field, purity and impurity discourses are strategically mobilised in the affirmation of the specificity of the cultural field as a domain oriented towards essence, in the course of strategic social practice. The cultural field itself is characterised as pure, compared to the economic field which is characterised as impure; those areas of culture which are most oriented towards 'the origin, the spirit, the authentic essence' are constructed as pure than others; and those regions of the field most proximate to this essence are seen as purer than those that are considered as worldly, and distant from this originary essence. Purity and impurity hierarchies may therefore come into conflict, simply bypass, or nest within one another, depending upon the various ways in which essence is appealed to within social fields. For instance, discourses invoking purity and impurity may be deployed by prophets and other ideologues, for example, against the compromises that consecrated authorities must make. They may also be deployed by these consecrated authorities themselves, in order to designate as defiled those activities which do not correspond to what is thereby imputed as the proper form of human life . In the political sphere, for example, Bourdieu describes a dichotomy and structural conflict between ideologues and pragmatists, 'those people, that is, who thus advocate a return to basics, to a restoration of the original purity; and, on the other hand, those people who are inclined to seek a strengthening of the party'. Or again, Bourdieu describes the opposition between two poles of academia. The applied faculties such as law and medicine affirm the public utility of the academic field. Those faculties dedicated to 'pure, rational knowledge' such as mathematics, or philosophy, affirm the autonomy of the academic field as oriented towards the disinterested search for the truth of existence. Bourdieu's theory has been criticised for over-emphasising the division between cultural producers and cultural consumers, and thus for focusing too strongly on intrainstitutional dynamics . Whilst this issue does mark Bourdieu's texts on the religious, cultural and political fields, it is questionable whether it concretely impacts upon the interpretation we present of Bourdieu's account of purity and impurity classifications. On this reading, purity and impurity discourses are both facilitated by and contribute to the emergence of institutionalised fields able to facilitate sustained discourses on the truth of reality, and on the extent to which particular phenomena or forms of subjectivity are in correspondence with this ideal. This reading of Bourdieu circumvents a certain strand in his thought which makes discourses of purity and impurity the necessary effect of the structural differentiation of relatively institutionalised social practice. The citation from Marx and Engel's The German Ideology, and certain other passages in 'The Genesis and Structure of the Religious Field ' , imply a structurally determinist account. In such an account, discourses such as purity and impurity would be no more than reflections of the relative autonomy of a field of social practice. Yet Bourdieu can also be read as indicating that classifications occur within various, sometimes competing, discourses, rather than representing the super-structural reflection of material conditions. In a later interview, Bourdieu affirms that this was his intended point in 'The Genesis and Structure of the Religious Field', and that it represents the most important lesson that he feels that he learnt from Weber. On such a reading of Bourdieu, the material division of labour represented by the relative autonomy of fields does not 'spontaneously' or mechanically produce purity and impurity discourses. For instance, the field of cultural production, the scientific field, and the field of national sovereignty do not naturally and inevitably produce corresponding discourses of purity and impurity. The reason why they are often associated with such discourses is that at stake in each of these relatively autonomous fields is an appeal to an essence; in these cases beauty, truth and the nation as a timeless truth of collective identity. Crucial for purity and impurity discourses is less the autonomy of the field than the appeal to an essence that is often associated with established, autonomous fields which have the authority to shape forms of subjectivity that are willing to make and recognise claims about the truth of existence. Thus in The Rules of Art, Bourdieu argues: If there is a truth, it is that truth is at stake in the struggle; and even though the divergent or antagonistic classifications or judgements made by agents engaged in the artistic field are indisputably determined or oriented by specific dispositions and interests linked to their positions in the field and to points of view, they are nevertheless formulated in the name of a pretention to universality, to absolute judgement, which is the very negation of the relativity of points of view. 'Essentialist thought' is at work in all social universes and especially in the fields of cultural production -the religious field, the scientific field, the literary field, the artistic field, the legal field, etc. -where games which have the universal at stake are played out. But it is quite clear in that case that 'essences' are norms . --- Purity and power A concern that has been raised regarding Bourdieu's work generally relates to his view of the relationship between structure and agency. There has been a sustained debate in the literature on Bourdieu, regarding whether he believes that social structures are constraining or whether agency is possible . Whilst in some texts, Bourdieu does indeed seem to argue against the possibility of meaningful agency for most subjects , elsewhere he is insistent that practices are not simply the result of 'rules' which subjects follow or express, but the result of strategic agency by social and institutional actors . On the topic of purity and impurity discourses, Bourdieu's argument is quite clear: they tend to support incumbent relations of material and cultural power, but this does not preclude their successful use by actors for achieving greater agency or substantial social change . Bourdieu suggests that 'the more people are... rich in specific capital, the more likely they are to resist' cooption by right-wing interests . Representations of purity tend to support the hierarchies of the social field and society at large by serving as symbolic capital, naturalising forms of subjectivity so that they appear to be no more than expressions of an imputed essence. Yet they can also serve as a legitimation strategy through which particular actors can offer social critique. It is possible, Bourdieu theorises, that 'in the name of the values of purity, freedom, truth... I can enter the political field', campaigning for 'reason, truth, virtue' . Bourdieu suggests that the political potency of such representations of purity in the public domain is, in part, because the economic and political fields are socially constructed in contemporary Western societies as impure: 'The order of the polis and politics' is always seen as opposed to 'the "free" and "pure"' . Those rich in cultural rather than material capital are oriented towards 'cultivated cults of the natural, the pure and the authentic' . Bourdieu calls these subjects 'the dominated of the dominant', since they are less powerful than the economically and politically dominant, whose forms of subjectivity are legitimated as dominant through their consumption of the goods of the fields of cultural production. Bourdieu thus writes of the 'boundary which distinguishes the pure reign of art and culture from the lower region of the social and of politics, a distinction which is the very source of the effects of symbolic domination exerted by or in the name of culture'. Dominant actors strive for and exercise their power, but occlude and legitimate these actions through the mobilisation of what Bourdieu refers to as 'ideals', produced and sustained in the first instance by other, 'purer' fields . Such interaction between the purity and impurity discourses of different fields is missed by Douglas, and represents a crucial advance made by Bourdieu's account. Bourdieu proposes that dominant subjects gain legitimacy as social elites by discursively constructing themselves as representatives and guardians of these pure and transcendent ideals: 'I am nothing but the delegate of God or the People, but that in whose name I speak is everything, and on this account I am everything' . Discourses invoking purity can thereby be used to legitimate programmes of action, and to impose a particular vision of the world on the national citizenry via 'the statist capital granting power over the different species of capital and over their reproduction' . The relative and vicarious proximity of the dominant to essence is 'unmarked' and so appears to subjects as a 'universal' potential, a 'neutral', 'natural' norm andideal . This is perhaps what Douglas was identifying in suggesting that purity functions as a symbol of society as a whole; however, Bourdieu identifies the intersecting power-relations upon which such a function depends. Such a process can be seen in Bourdieu's discussion in The State Nobility of 'pure activities, that is, activities that have been purified of all profane, pragmatic, and profitable purpose' . These pure activities are available in elite educational institutions, and demonstrate 'the qualities of disinterestedness and endurance'. These qualities show that a subject is able to exercise 'control over nature, that is to say, over those who cannot control their nature': the ability to rule oneself serves as a mandate to rule others : 235). Mastery of the cultural realm as the domain of 'the pure' makes social power and authoritative forms of knowledge the seemingly natural possession of the dominant elite, granting them a tacit and relative purity . Bourdieu theorises that dominant subjects in contemporary Western societies are represented as possessing a tacit and relative purity. By contrast, an impure subject-position is relatively attributed across human society, depending on properties like class and gender and the way in which these are enacted with propriety or impropriety in the domain of consumption. Impure forms of subjectivity may seem to represent no more than that which is extraneous to a fullyhuman, 'proper' subject-position. However, Bourdieu's account suggests that such a figure may in fact be part of the means through which this subject-position is socially and materially constructed and made to seem 'proper' and pre-given. Bourdieu's account of themes of purity and impurity shows the important role they play in the processes through which contemporary Western societies continually construct 'a whole relationship to animal nature, to primary needs and the populace who indulge in them without restraint', such as women and the working-classes . Though appearing as the expression of their 'impure nature', the actions and capacities of the dominated are in fact, Bourdieu insists, shaped by their materially subordinate position. This position practically necessitates association with matters classified by the dominant culture as impure, such as physical labour, dirt, and lasciviousness . Bourdieu writes of the dominant: They reduce the strength which the dominated ascribe to themselves to brute strength, passion and instinct, a blind, unpredictable force of nature, the unreasoning violence of desire, and they attribute to themselves spiritual and intellectual strength, a self-control that predisposes them to control others . On our reading, Bourdieu's writings suggest that one of the fundamental properties of discourses of purity and impurity is that they hide their relationship with dynamic relations of power and materiality . What is the mechanism through which this takes place? Though Bourdieu does not draw this out explicitly, we wish to suggest that when phenomena or forms of subjectivity are taken to be unadulterated, simple expressions of their essence -that is, as pure -they tend to be granted a moral and epistemological privilege. With no heterogeneous processes impacting upon the existence of the phenomena or form of subjectivity, it appears that only a universal and immutable essence stands as the ground of a phenomenon or a particular form of subjectivity. Bourdieu proposes that those who are dominant within a field, and within society more generally, can hide the historicity and the power-relations upon which their position in society depends by presenting themselves as no more than delegates of an ideal of essence. This legitimates their place in society, and situates their attributes and ethos as the universal norm. Rather than expressing and re-affirming society as a collective entity, the mobilisation of purity and impurity discourses is structured by hierarchies both within and beyond particular cultural fields as differentiated realms of social life: 'the most "pure" products, "pure" art or "pure" science' serve what would otherwise be considered the 'totally "impure" social functions', most notably 'distinction and social discrimination or, more subtly, the function of a disavowal of the social world which is inscribed' . --- Douglas proposes representations of purity and impurity as an unmediated expression of the maintenance or breach of any social or cognitive categories that reflect the general social structure. Designations of purity and impurity regarding the boundaries of the body, Douglas therefore argues, always mirror the internal and external social and symbolic boundaries of society as a pre-given 'bounded system' . Examining virginity, Douglas : 200-1) argued that it will inevitably be coded as pure since 'the yearning for rigidity is in us all'. She explains that by this she means that through the figure of the virgin, we can 'enjoy purity at second hand', experiencing this state which we all inherently desire vicariously in the untainted body of the virgin. The more general fact that themes of impurity seem to 'cluster' particularly around both sexuality and femininity is explained by Douglas : 194) as simply natural to human societies, as she does not believe that it can be adequately accounted for using her theory. By contrast, on the account presented here, purity and impurity should be regarded as discourses that measure phenomena vis-à-vis an imputed essence, considered to be the selfidentical ground for true phenomena in the world. Impurity is invoked to classify those elements that particular discursive actors are unwilling or unable to place as expressions of the particular construction of essence. Representations of purity and impurity can thus serve to bridge the gap between ostensibly natural and discrete phenomena or forms of subjectivity, whilst facilitating mechanisms of social stratification and of social and self-regulation. Thus femininity and sexuality should not be regarded as natural and inevitable sites of purity and impurity classifications; a social theory that suggests such risks affirming rather than providing tools for the analysis of misogyny, as Buckley and Gottlieb have claimed. Rather, we wish to contend, female sexuality is often described using purity/impurity classifications because it is a key biopolitical object, able to be regulated and normalised through appeal to an ideal of human essence. For instance, the division between pure and impure forms of female sexual subjectivity is an apparatus which serves to manage and mask the contingencies that attend the material and discursive construction of racial identities as essences. In order to appear to be an essence underpinning particular human instances, 'race' is indelibly dependent upon the exigencies of the reproductive choices of its members who therefore must be subject to self-regulation and social policing. 'Race' appears to be a transhistorical essence that is merely instantiated by each generation of the nation; however, in fact the race is indelibly dependent upon the exigencies of the reproductive choices of its members who therefore must be subject to self-regulation and social policing. --- Concluding reflections As Douglas predicts in Purity and Danger, pure phenomena can sometimes appear to symbolise the order and benevolence of society, and the impure its disequilibrium or disruption by anomaly. Bourdieu himself maintains that there is a tendency in contemporary Western societies to exclude 'all misalliances and all unnatural unions -i.e. all unions contrary to the common classification, to the diacrisis which is the basis of collective and individual identity'. Yet not all exclusions are classified as impure, and not all that is within social or cognitive bounds is classified as pure. Our reading of Bourdieu suggests a new theory of purity and impurity: purity should be conceptualised as the adjudication of phenomena in terms of their relative identity with or divergence from their imputed essence. Bourdieu may focus on the economic processes that are euphemised in the cultural activities of social actors, but his point regarding the role played by purity and impurity can be extended to a variety of cases. It can be suggested, based on his account, that discourses of purity and impurity are mobilised to manage the practical tensions associated with the construction of particular subjects or phenomena as relatively corresponding to an imputed ideal, standing outside of history or relations of power. Through the discursive adjudication of phenomena or subjects as relatively impure or pure, an image is constructed of what these phenomena or subjects are in their truth. Impurity does not spontaneously characterise any anomaly, as Douglas predicts, but is invoked to classify those elements that particular discursive actors are unwilling or unable to place as expressions of an imputed essence. In contrast to Douglas who sees power as only ever operating to ensure the homeostatic stability of the social and cognitive structure of the community as a whole, our reading of Bourdieu has suggested that pure processes, things, or people appear to be simple expressions of essence, with no dependence on anything outside of themselves. They are devoid of and prior to complexity and the dynamics that organise social and material inequalities. This makes purity and impurity a discursive, material and affective resource peculiarly adapted to facilitating social consensus, and compelling a shared practical demand to protect or attain purity through the deployment of mechanisms of social exclusion, and social and self-regulation. Purity and impurity, therefore, help manage and mask the tensions that would otherwise become apparent when an essence is imputed at the ground of particular phenomena or forms of subjectivity, as the impure is held epistemologically and morally responsible for the distance of reality from its essential truth. It is this operation which facilitates the mobilisation, identified by Moore , of purity and impurity discourses in the justification of black-and-white and dehumanising world-views. --- Robbie

This article will propose a new interpretation of Pierre Bourdieu, as a theorist of purity and impurity. Bourdieu's writings indicate that through the adjudication of things or people as relatively impure or pure an image is constructed of their essential truth. Building from Bourdieu, we will show how themes of purity and impurity can be used to manage the tensions associated with attempts to impute an essence to human nature or to reality, ensuring that moral and epistemological significance of complexity is masked. This is the reason why themes of purity and impurity so often attend polarised worldviews, and why they are frequently mobilised for justifying and operating biopolitical processes of social stratification and regulation.

INTRODUCTION Gun violence remains a serious public health and safety problem in the USA. In 2009, a total of 9,146 people were murdered with firearms, and it is estimated another 48,158 were treated in hospitals for gunshot wounds received in assaults. 1,2 Furthermore, exposure to gun violence and homicide is associated with a host of negative health outcomes including PTSD, depression, psychobiological distress, anxiety, cognitive functioning, and suicide, as well as other negative social behaviors such as school dropout, increased sexual activity, running away from home, and engagement in criminal and deviant behaviors. [3][4][5] Leading social scientific explanations of gun violence commonly associate a heightened probability of gunshot victimization with individual , situational (e.g., the presence weapons, drugs, or Papachristos is with the Department of Sociology, Yale, New Haven, CT, USA; Braga is with the School of Criminal Justice, Rutgers University, Newark, NJ, USA; Braga is with the Program in Criminal Justice Policy and Management, Harvard University, Cambridge, MA, USA; Hureau is with the Department of Sociology, Harvard University, Cambridge, MA, USA. Correspondence: Andrew V. Papachristos, Department of Sociology, Yale, New Haven, CT, USA. alcohol), and community risk factors. [6][7][8][9] Yet, the majority of individuals in high-risk populations never become gunshot victims. Indeed, research suggests that gun violence is intensely concentrated within high-risk populations. 10,11 For example, recent studies in Boston found that from 1980 to 2008, only 5 % of city block faces and street corners experienced 74 % of gun assault incidents 12 and that 50 % of homicide and nearly 75 % of gun assaults were driven by less than 1 % of the city's youth population , most of whom were gang involved and chronic offenders. 13 To better understand how gunshot injury is distributed within highrisk populations, we conducted a study to determine whether the risk of gunshot victimization is related to characteristics of one's social networks. Studies on the health effects of social networks suggest that the clustering of certain health behaviors-such as obesity, smoking, and depression-is related not only to risk factors but also the contours of one's social network. [14][15][16][17] There are several reasons why the risk of gunshot victimization is related to one's social network. First, interpersonal violence tends to occur between people who know each other suggesting that the context of social relationships is important in understanding the dynamics of gun violence. 18,19 Second, the normative conditions surrounding gun use are transmitted through processes of peer influence, especially among young men with criminal histories. 20 Third, guns themselves are durable objects that often diffuse through interpersonal connections suggesting that obtaining a gun-a necessary precursor to using a gun-must also occur through interpersonal relationships. 21 Yet, despite the growing interest in social network analysis in the study of public health, no study has yet employed formal network models to understand how processes of peer influence and normative diffusion might influence the risk of gunshot injury. The present study analyzes the salience of social networks on differential risks of gunshot injury among a population of 763 individuals in Boston. We examine several aspects of individuals' social networks including: network density, the saturation of gang members in one's network, and the social distance between an individual and other gunshot victims. We hypothesize that the structure and composition of an individual's social network will influence one's exposure to and risk of gunshot injury and thereby better explain the concentration of gunshot injury within risk high-risk populations. --- METHODS --- Setting The present study examines gunshot victimization among a network of individuals from Boston's Cape Verdean community . Cape Verde is an archipelago of islands located off the West Coast of Africa that was a colony of Portugal until 1975. As of 2000, Boston is home to an estimated 35,000 to 50,000 persons of Cape Verdean descent. Boston's Cape Verdean population is concentrated in two communities-the Bowdoin-Geneva and Upham's Corner neighborhoods-that are associated with many traditional violent crime "risk factors." For example, in the Bowdoin-Geneva neighborhood 20 % of the population lives below the federal poverty line, 52 % live in a single-family household, and 42 % of the population has less than a high-school diploma. 22 As seen in Figure 1, and consistent with prior neighborhoodlevel research. 7,23 such socio-demographic characteristics tend to be higher in highcrime neighborhoods. Put another way, neighborhoods with high levels of socioeconomic disadvantage also tend to have higher levels of crime and violence. The study communities were selected for three main reasons. First, as just described, the study area exhibits many of the aggregate level risk factors commonly associated with elevated levels of crime and violence. Second, as seen in Map B in Figure 1, the community also exhibits some of the highest concentration of gunshot injuries in Boston and, in fact, this area is driving much of the city's current violence problem. In particular, the study neighborhoods struggle with youth street gang problems: fatal and non-fatal shootings involving Cape Verdean gang-involved youth more than tripled from 12 shootings in 1999 to 47 shootings in 2005. 24 Finally, although the study area is situated ecologically within the a larger predominately African American section of Boston, violence within the study area tends to occur almost entirely between members of the Cape Verdean population. 25 Thus, the study community represents a high-risk community in a larger urban area whose violence tends to occur within an identifiable population. --- Data Collection Data come from two sources provided by the Boston Police Department: Field Intelligence Observation cards and records of fatal and non-fatal gunshot injuries. FIOs are records of non-criminal encounters or observations made by the police; these reports include information such as: reason for the encounter, location, and the names of all individuals involved. Since these data include only observations by the police, the FIO data provide a conservative measure of one's social networks as individuals have more friends and associates whom the police do not observe. Ties between individuals were derived for all situations in which two or more individuals were observed in each other's presence by the police and recorded in FIO data-those two people observed by the police in the same time and place are taken to be "associates." Extant qualitative research in sociology, anthropology, and criminology suggests that "hanging out"-standing on street corners while associating with one's friends-is an important social behavior among young urban males as well as a key mechanism driving street-level violence. 20,[26][27][28][29] To generate the social networks of high-risk individuals in the study communities, we employed a two-step sampling method frequently used in the study of other high-risk populations such as drug users and sex workers. 30 The initial sampling seeds consisted of the entire population of Cape Verdean gang members known to the police . Step 1 entailed pulling all FIOs in the year 2008 for these 238 individuals to generate a list of their immediate associates. This step was repeated to gather the "friends' friends" of the original seeds to create a final social network of 763 individuals. Previous research suggests that such a two-step approach adequately captures the vast majority of information necessary to understand the underling social processes. 31 The FIO data were then merged with data on all known fatal and non-fatal gunshot injuries reported to the police enabling us to determine which individuals in our social network were the victims of gunshot violence in the years 2008-2009. During the study time period, two of the individuals in the sample were the victims of fatal gunshot wounds and 38 were victims of non-fatal injuries. --- Models We use rare event logistic regression 32 to model the determinates of gunshot victimization in the sample population. Two sets of models are presented. The first set presents the results on the entire population of 763 individuals, while the second set presents the results of a subsample of 579 of the population that comprise a single larger network. To account for temporal ordering, the network was constructed using data from 2008 and regressed on the victimization data for 2008-2009. Network calculations and visualizations were conducted using the statnet software in the statistical package, R. 33 Regression analyses were conducted using Stata 10. 34 Variables Table 1 shows the mean, standard deviation, and range for all variables used in our analysis. --- Dependent Variable The dependent variable is a binary indicator of whether or not an individual was the victim of either a fatal or non-fatal gunshot wound in 2008-2009. Approximately 5 % of the sample were victims of gun violence. The current study combines fatal and non-fatal injuries; analysis of only non-fatal shooting found no discernible differences in the results. --- Independent Variables Individual Level Covariates. Our models include several individual-level control variables associated with gun violence: age, gender, race/ethnicity, and whether or not the individual has ever been arrested. Age is consistently a strong predictor of violent victimization: rates of homicide victimization peak between 18 and 24, and decline steadily thereafter. We square age to capture this non-linear relationship. Gender is measured as a binary variable . The vast majority of network members are male . Ethnicity is measured as a binary variable indicating whether or not the subject was of Cape Verdean ancestry . Half of the study population is of Cape Verdean decent and the remainder is mainly African-American. Finally, we include a binary dummy variable to indicate whether or not the subject has at least one prior arrest with the Boston Police Department . A full third of the sample has at least one prior arrest. Network Measures On average, any individual in the network has ties to approximately three associates, though the standard deviation is equally large. This distribution of ties in the network-presented in the Appendix-is consistent with prior research that finds that most individuals in networks have a small number of ties, while a small number of individuals have an exceedingly large number of ties. 35 In the present data, however, some caution is in order as the ties themselves are based on police observations-i.e., the number of ties may be influenced how police go about their duties and investigations. 36 As such, we weigh our sample according to this degree distribution to account for any bias attributable to policing efforts. Four social network measures are included in the analyses: network density, the percentage of one's associates who are known gang members, the percentage of one's immediate associates who have been gunshot victims, and the average "social distance" from the subject to other shooting victims. Network density is a basic property that reflects the overall intensity of the connected actors: the more connected the network, the greater the density. 37 Dense networks are often associated with cohesive subgroupings and cliques. 38 Formally, network density is measured as the sum of ties that are present in the network divided by the possible number of ties. 37 Here, we measure the ego-network density as the density of ties in the immediate social network surrounding each individual. 37 We also measure the percentage of one's immediate associates who are gang members. This measure extends the prior research on the negative consequences of gang membership 39 by capturing a saturation effect: greater exposure to gang members in one's social network should also increase one's exposure to gun violence. Exposure to gunshot violence is measured in two ways. First, we measure the effect of exposure to gunshot violence in one's immediate social network as the percentage of an individual's immediate associates who were gunshot victims-i.e., someone whom they were observed associating with in public. Second, we extend this idea to include a measure of social distance to gunshot victims, measured as the average number of shortest paths from the subject to all gunshot victims in the social network. 37 In large social networks, individuals can be connected indirectly in many different ways and, therefore, information and influence can potentially travel different paths in the network between any two individuals. Research demonstrates that a wide variety of health and social behaviors are affected by people in the our social networks who are a few handshakes removed. 14,40 Formally, we measure social distance as the mean geodesic distance between each individual in the sample to all gunshot victims. The geodesic distance refers to the shortest path between two nodes, where the distance between two nodes n i and n j is measured simply as d. The shortest distance, then, is the smallest value of d. We calculated the measure in the following manner. First, we computed the entire distance matrix for the social network; in this case a 786 by 786 matrix where each cell value represents the distance d between two nodes. These data are symmetric meaning, that in all cases d=d. Next we complied a binary vector of the 786 individuals in the network indicating whether each of the individuals was the victim of a shooting . Finally, because there are multiple shooting victims and multiple paths connecting individuals to these multiple shooting victims, we calculated the mean average geodesic distance of all possible shortest distances to all shooting victims; such an approach thus allows us to capture all potential avenues of indirect exposure. For disconnected parts of the network, we calculate this measure within each component. Our basic working hypothesis in this study is that individuals who are, on average, closer to shooting victims will be at greater risk of becoming victims themselves. --- RESULTS The network of 763 individuals generated from the sampling method is presented in Figure 2. Each of the nodes represents a unique individual and each of the ties linking two nodes indicates at least one observation of two individuals observed socializing together. A total of 1,869 ties were extracted from the data. Gunshot victims are represented as the larger red nodes in the network. Figure 2 is comprised of 57 unique subnetworks , although 76 % of individuals are connected in the single large network consisting of 579 individuals . The majority of gunshot victims are found in the large component . The average geodesic distance between any individual in the network and a gunshot victim is 4.69. Taken together, these findings suggest that the majority of individuals in this network are connected in a single large network and, on average, any person is roughly five handshakes removed from a gunshot victim. --- Predicting Gunshot Victimization Table 2 shows the odds ratios and 95 % confidence intervals for models that regressed gunshot victimization on the full set of explanatory variables on both the entire network and the largest component. Examination of the individual-level predictors for both models show that, the odds of being a gunshot victim decreases with age and increases with prior contact with the criminal justice system . As expected, females in the network are less likely to be victims and those of Cape Verdean decent are more likely to be victims . However, neither of these variables attain statistical significance, reflecting the gender and racial homogeneity of the sample. Ego-network density in both models is negatively related to gunshot victimization suggesting that density may, in fact, be protective of victimization ; the p value of this variable, however, suggests that this effect is not significantly different than zero. When considering only the complete network, the saturation of gang members increases one's odds of being shot , although the statistical significance of this effect drops when considering only the large component . The magnitude of our two variables pertaining to network exposure to gunshot injuries differs slightly between the models in Table 2 of immediate alters who have been shot greatly increases one's odds of also being a gunshot victim ): a 1 % increase in the number of one's friends who are gunshot victims increases one's own odds of victimization by approximately 144 %. This effect diminishes in the model for only the large network and loses its statistical significance . This loss of statistical significance highlights the fact that individuals in smaller networks have fewer avenues of indirect exposure and, therefore, direct exposure has a much more potent influence. Both models in Table 2 support our main hypothesis that social distance is related to gun victimization: the closer one is to a gunshot victim, the greater the probability of one's own victimization, net of individual and other network characteristics. In the whole network model, every one connection away from a shooting victim decreases the odds of getting shot by 8.8 % . This effect is more pronounced in the large component: every one connection removed from a gunshot victim decreasing one's odds of getting shot by approximately 25 % . This relationship between distance to a shooting victim and probability of gunshot victimization is summarized in Figure 3 where the x-axis indicates the average distance to a shooting victim and the y-axis indicates the predicted probability of gunshot injury in the largest component model in Table 2. Because the large component is completely connected and contains the majority of shooting victims, the x-axis begins at four demonstrating that everyone in the large networks is indirectly connected to a shooting victim and the shortest average path to any victim is four connections. Figure 2 reveals two important features related to social distance. First, the association between social distance and the probability of gunshot victimization is more pronounced among gang members, suggesting that gang members may occupy unique positions within such networks that place them at greater risk. Second, for both gang and non-gang members, the effect of the risk begins to level off after approximately five network degrees. Regardless, the effect of indirect exposure to gunshot injuries is pronounced for both gang and non-gang members. --- DISCUSSION Our data on high-risk individuals Boston's Cape Verdean community reveals a social network of young men with a highly elevated risk of gunshot victimization. Network analysis shows the existence of a social network consisting of 763 individuals, the majority of whom are all connected in a single large network and, on average, individuals in this network are less than five handshakes away from the victim of a gun homicide or non-fatal shooting. Our findings demonstrate that the effect of this distance to a shooting victim greatly increases an individual's own odds of becoming a subsequent gunshot victim: the closer one is to a gunshot victim, the greater the probability that person will be shot. Indeed, each network step away from a gunshot victim decreases one's odds of getting shot by approximately 25 %. The findings of this study are limited in three ways. First, our sampling clearly does not identify all individuals at risk of gunshot victimization. Situations not visible to police investigation-such as unreported domestic violence incidents-would not be captured in our data. Second, the use of FIO data circumscribes our measurement of social networks to those ties witnessed firsthand by police and, therefore, we probably underestimate the extent of social networks. Third, our findings may also be confined to the unique character of Boston's Cape Verdean neighborhoods. However, these communities share many similarities with other high-crime and socially disadvantaged urban neighborhoods and recent research suggests that the network patterns described here extend to gang violence in Chicago. Limitations notwithstanding, these results imply that social networks are relevant in understanding the risk of gunshot injury in urban areas. The contours of our social networks-even when we cannot see them-affects our behavior. Furthermore, our findings suggest that the risk of gunshot victimization is not evenly distributed within high-risk populations. In the present study, those individuals in the largest social network, for instance, are at a much greater risk of victimization than either those in the smaller disconnected networks or of the general neighborhood population in large part because of the ways in which people are situated in social networks. How and why such networks affect the ways in which we assess the risk of gunshot injury is of importance for future research and public health. In particular, gun violence reduction strategies might be better served by directing intervention and prevention efforts towards individuals within high-risk social networks. --- APPENDIX A FIGURE 4. The distribution of the number of network ties. --- SOCIAL NETWORKS AND THE RISK OF GUNSHOT INJURY

Direct and indirect exposure to gun violence have considerable consequences on individual health and well-being. However, no study has considered the effects of one's social network on gunshot injury. This study investigates the relationship between an individual's position in a high-risk social network and the probability of being a victim of a fatal or non-fatal gunshot wound by combining observational data from the police with records of fatal and non-fatal gunshot injuries among 763 individuals in Boston's Cape Verdean community. A logistic regression approach is used to analyze the probability of being the victim of a fatal or non-fatal gunshot wound and whether such injury is related to age, gender, race, prior criminal activity, exposure to street gangs and other gunshot victims, density of one's peer network, and the social distance to other gunshot victims. The findings demonstrate that 85 % all of the gunshot injuries in the sample occur within a single social network. Probability of gunshot victimization is related to one's network distance to other gunshot victims: each network association removed from another gunshot victim reduces the odds of gunshot victimization by 25 % (odds ratio=0.75; 95 % confidence interval, 0.65 to 0.87). This indirect exposure to gunshot victimization exerts an effect above and beyond the saturation of gunshot victimization in one's peer network, age, prior criminal activity, and other individual and network variables.

INTRODUCTION One of the most important aspects affecting a nation's overall growth and development is its healthcare system. For a country to develop and compete on the international stage, its citizens must be educated and in good health. A population that is healthy is more productive and capable of working. Reducing the burden of disease and enhancing overall health outcomes require access to high-quality healthcare it is impossible to overstate the importance of health in a nation's development. It has an impact on social progress, educational achievement, and life quality in general. Sustainable development and prosperity are influenced by investments in healthcare systems, disease prevention, and public health. Good health enhances productivity and workforce participation, leading to increased economic output. Healthier individuals are more likely to be engaged in the labor force, leading to higher overall productivity. This concept is wellexplained in the Human Capital Theory [1]. A healthy population requires fewer healthcare resources and expenditures. This allows governments to allocate resources to other sectors, such as education and infrastructure, fostering overall development. [2]. Good health is linked to higher educational attainment. Healthy children are more likely to attend school regularly and perform better academically, contributing to a more educated and skilled workforce [3]. In the early stages of economic theory, Gross Domestic Product served as the standard metric for assessing a nation's economic progress. It was commonly believed that development and prosperity naturally increased with a rise in GDP per capita. However, this narrow focus on GDP as a development indicator overlooked the well-being of the population. High GDP alone does not guarantee greater prosperity among the people, as the equitable distribution of wealth and resources plays a pivotal role in the overall benefit to society. Esteemed economists like Paul Streeten, Frances Stewart, Mahbub Ul Haq, and Nobel Prize laureate Amartya Sen have criticized GDP as an inadequate proxy for development. Instead, they advocate the concept of human growth, which led to the emergence of Human Development in the late 1980s, championed by Drs. Amartya Sen and Mahbub ul Haq. The crux of development lies in creating an environment that fosters long, healthy, and productive lives for individuals. Although there has been a significant rise in government healthcare spending globally, increasing at approximately 6% in low and middle-income countries and 4% in highincome countries , a concerning trend persists. Each year, approximately 100 million people are pushed into extreme poverty because they are compelled to allocate over 35% of their income to access healthcare services [4]. Research has shown that simply increasing healthcare spending cannot guarantee universal healthcare and human development [5] Banna Banik et al. [6] examined how the quality of governance influences the relationship between healthcare spending and human development. Their findings revealed that political stability and absence of terrorism had the most significant net positive impact on health expenditure, contributing to the improvement of human welfare. Moreover, they observed that the interaction effect between expenditure and the Good Governance Index was negative but not statistically significant for low-income countries , negative and statistically significant for sub-Saharan African economies, and positive but not statistically significant for South Asian nations. Esen and Kecili [7] examined the relationship between health expenditure and economic growth in Turkey from 1975 to 2018. The study found evidence of a one-way causality running from health expenditure to economic growth in the short term. As a result, the study recommends that the government should prioritize increasing the allocation of health expenditure in the budget to enhance the overall health status of the population. Kaur investigated the relationship between health expenditure, health status , and economic growth in India from 1981-82 to 2015-16. The study confirmed a long-term relationship among these variables. It revealed that there was a one-way causality running from health expenditure to economic growth and from health expenditure to the infant mortality rate. The study also found that health expenditure significantly influenced both economic growth and infant mortality rate in India. Safitri [8] employed panel data from 23 districts over the period 2008-2014 and found only the spending on health has a significant impact on the HDI improvement. Arun and Kumar [9] investigated the relationship between per capita public health expenditure and GDP per capita in BRICS countries, including Brazil, Russia, India, China, and South Africa, from 1995 to 2013. The study revealed that causality ran from GDP to per capita public health expenditure, but no causality was observed from public health expenditure to GDP. Furthermore, the study concluded that economic growth played a crucial role in enhancing public health expenditure, leading to the provision of improved healthcare facilities. Rajeshkumar and Nalraj [10] conducted a study to explore the relationship between healthcare expenditure and economic growth in four Indian states: Madhya Pradesh, Orissa, Kerala, and Tamil Nadu, covering the period from 1991 to 2010.. The findings revealed that healthcare expenditure and economic growth were co-integrated, indicating a long-term connection between these variables. Additionally, the study confirmed the presence of unidirectional causality, with health expenditure influencing economic growth in Madhya Pradesh, Orissa, Kerala, and Tamil Nadu. Craigwell, Bynoe, and Lowe [11] conducted a study using panel data from 19 Caribbean countries. Their findings from the Granger causality test indicated that public spending on healthcare positively influences life expectancy, leading to an improvement in people's longevity. However, the study did not find any significant impact of public spending on healthcare on the level of education among the population. Razmi et al. [12] found no bidirectional relationship between health spending and the Human Development Index in Iran. However, their analysis using the Ordinary Least Square approach demonstrated that public health expenditure positively contributed to increasing the HDI. This was attributed to the allocation of funds for enhancing the healthcare system and promoting awareness among the population. Kerala is the southern Indian state which has a notable achievement in human development, particularly in health and education, set it apart and define the Kerala Model. Despite its low per capita income and limited industrial production, the state's Physical Quality of Life Index surpasses the national average, reflecting the irony of its development. While health and education improvements are crucial, unlocking economic opportunities is equally vital for sustainable development. Kerala's case serves as a striking example of this paradox, showcasing remarkable human development amid perceived poverty due to limitations on expanding productive capacity and entrepreneurial activities. According to Mujeeb Rahman AP and Bassam K [13] The Human Development Index of the Malappuram district in Kerala is lower than that of the state of Kerala, and in order to raise it, adequate steps must be taken in the areas of standard of living, health, and education. Provide more chances for income-generating activities to boost the per capita income. --- Theoretical Framework The major theoretical innovation in the study of the demand for health care has come from Grossman .Demand for healthcare is derived from a demand for health itself. Grossman's model assumes that individuals assess the benefits from outlays that will improve their health compared with expenditure on other goods and services in order to decide their optimal health state.Estimates provided by Grossman ,Rosett and Huang [14] and Phelps and Newhouse [15] all indicated a positive effect of income on health care expenditure. Grossman's results also indicate a negative relationship between income and the individual's number of healthy days per year. Education was found to be positively related to health stock .Leander and Garcia-Gómez in their study examined the contributing determinants of inequity in healthcare utilization in South Africa. The results indicate that the rich are more likely to use inpatient care than the poor, given the same level of need. Other factors like race, gender, education and the consumption level also contribute to inequity. --- Statement of the Problem --- METHODOLOGY This Section outlines the methodology employed to conduct the comprehensive analysis presented in this study, focusing on the relationship between individual health status, lifestyle factors, and healthcare expenditure patterns. The methodology encompasses data collection, variables selection, statistical techniques, and analytical frameworks adopted to address the research objectives. Data Collection: Primary data was collected through a structured survey conducted in 2023 within the Malappuram district. A diverse sample of 386 individuals was selected, representing various demographic segments. The survey questionnaire comprised sections that captured demographic information, health status perceptions, lifestyle choices, and healthcare expenditure patterns. The data collection process was carried out with a rigorous approach to ensure the accuracy and reliability of responses. Variables Selection: The study analyzed a range of variables to address the research objectives comprehensively. These variables included religious affiliations, socio-economic factors , lifestyle choices , general health status perceptions, and healthcare expenditure patterns. Analytical Techniques: Descriptive Analysis: The collected data was subjected to descriptive analysis, presenting frequencies, percentages, means, and standard deviations of different variables. This allowed for a preliminary overview of the dataset and highlighted initial trends. ANOVA and t-tests: Analysis of Variance and t-tests were employed to explore disparities in health status and healthcare expenditure patterns across various categories, such as religious affiliations, income levels, age groups, and gender. These techniques facilitated the identification of statistically significant differences. Regression Analysis: Multiple Linear Regression was conducted to assess the relationship between lifestyle factors and healthcare expenditure patterns. The model aimed to explain the variance in health expenditure based on selected independent variables. Coefficients, significance levels, and goodness of fit measures were used to interpret and evaluate the model. Ethical Considerations: Ethical standards were adhered to throughout the study. Informed consent was obtained from all participants, ensuring their willingness to participate. Data confidentiality and anonymity were maintained, and ethical guidelines were followed in data collection, storage, and analysis. Limitations: The study is not without limitations. The cross-sectional nature of the data limits causal inferences. The self-reported nature of data introduces potential biases, and unobserved variables could impact the results. --- RESULTS AND DISCUSSION Within the following analysis section, we embark on a comprehensive journey to explore multiple facets of healthcare within Malappuram district. Our objectives encompass a range of factors that collectively contribute to the understanding of health dynamics and healthcare utilization patterns among the district's diverse population. Through rigorous examination, we aim to shed light on the intricate interplay between socioeconomic variables, demographic characteristics, and healthcare outcomes, ultimately providing insights that can inform targeted interventions and policy enhancements. The objectives guiding this analysis are as follows: In this pursuit, we strive to unravel the health status of the district's inhabitants while carefully considering the potential influence of religious affiliations. By elucidating variations in health outcomes among different religious groups, we aim to contribute to a nuanced understanding of the interrelationship between cultural and religious factors and individual well-bein. The table presents a religious-wise comparison of the health status of respondents in terms of both general health and mental health. The health status is rated on a scale of 1 to 5, where 1 represents poor health and 5 represents excellent health. Upon analysis, several noteworthy observations emerge. Firstly, in terms of general health status, Christians exhibit a mean score of 3.0000, Hindus at 3.6870, and Muslims at 3.6502. The overall mean general health score for all respondents is 3.6425. The ANOVA results reveal an F-value of 2.258 with a corresponding p-value of 0.106. This nonsignificant p-value implies that the disparities in general health status scores across religious groups are not statistically significant. This might be attributed to the notion that general health status is influenced by a multitude of factors such as genetics, lifestyle, environment, and access to healthcare, which may not be strongly tied to religious affiliation. Turning to mental health status, the mean scores for Christians, Hindus, and Muslims are 3.2500, 3.7939, and 3.8601, respectively. The overall mean mental health score for all respondents is 3.8187. The ANOVA results for mental health status exhibit an F-value of 2.070 with a p-value of 0.128. Similarly, this non-significant p-value indicates that the differences in mental health status scores across religious groups are not statistically significant. This outcome aligns with the understanding that mental health is influenced by a complex interplay of psychological, biological, and social factors, wherein religious affiliation might not be the sole determinant. In summary, the statistical outcomes underscore that, within this dataset, religious affiliation does not seem to be a significant factor influencing health status perceptions in terms of both general health and mental health dimensions. Instead, other variables such as socio-economic status, access to healthcare, and individual lifestyle choices might play more substantial roles in shaping health outcomes. .7% nearly every day, and 10.7% for several days. The Chi-Square test is conducted to ascertain whether the differences in the distribution of interest or pleasure frequencies among religious groups are statistically significant. The Chi-Square result is 24.127 with 4 degrees of freedom and a pvalue of 0.000. This low p-value indicates that the differences in the distribution of frequencies are statistically significant at the conventional significance level of 0.05. The statistically significant Chi-Square result suggests substantial differences in the distribution of reported frequencies of feeling interest or pleasure among religious groups. These differences could be influenced by various factors, including cultural practices, social norms, and psychological wellbeing. It is possible that the ways in which individuals from different religious communities find enjoyment in activities may be shaped by their beliefs, traditions, and communal contexts. This table underscores the importance of considering not only negative mental health experiences but also positive emotions and wellbeing when examining mental health within religious contexts. The observed variations in interest and pleasure frequencies can inform culturally sensitive mental health interventions that address the diverse needs of different religious communities. Within the realm of healthcare analysis, the economic dimension stands as a pivotal focal point. Our mission revolves around unearthing the intricate ways in which medical expenditures are allocated across a spectrum of household income levels and socio-economic strata. Through an in-depth exploration of expenditure patterns, our objective is to illuminate potential disparities in healthcare service access. In doing so, we are dedicated to championing the cause of fostering a healthcare ecosystem characterized by equity and fairness. The objective of assessing variations in medical expenditure patterns based on household income levels or socio-economic status holds substantial significance within the realm of healthcare analysis and policy development. This objective addresses the vital issue of equity in healthcare access by unraveling how healthcare expenses are distributed across different economic strata. By scrutinizing expenditure patterns, we gain insights into potential disparities in accessing healthcare services. This knowledge is particularly crucial for identifying segments of the population that might face obstacles in affording necessary medical care, thereby exposing gaps in healthcare equity [17][18][19][20]. Nature of Family: The substantial difference in mean health expenditures between joint and nuclear families, along with the statistically significant t-test result, suggests that family structure plays a role in influencing health expenditure patterns. Individuals within joint families seem to allocate higher resources toward healthcare compared to those in nuclear families. This disparity could be attributed to factors such as shared financial responsibilities within joint families, which might lead to increased healthcare spending. In sum, the table underscores the intricate interplay between demographic characteristics and health expenditures. These findings reinforce the need for nuanced policy responses aimed at promoting equitable access to healthcare services. Recognizing the differential impact of demographics on expenditure trends is pivotal for crafting targeted interventions that bridge potential gaps, ultimately ensuring healthcare affordability and accessibility for diverse populations [20,21]. Table 6 encapsulates the Gini Coefficient, a measure used to quantify the inequality in the distribution of health expenditure across different segments of the population. The Gini Coefficient of 0.92 is indicative of a substantial level of inequality in the distribution of health expenditure. Such a high value signifies that there is significant disparity in how health expenditures are distributed among the surveyed population. The considerable gap between the Lorenz Curve and the line of perfect equality points to pronounced disparities. This result highlights the need for focused policy interventions aimed at reducing the inequality in health expenditure distribution. It suggests that a relatively small portion of the population might be shouldering a disproportionately large burden of health expenditure, while others have more equitable access to healthcare services. Addressing this imbalance could involve targeted measures to improve healthcare affordability for vulnerable and lower-income segments of the population, thereby working toward a fairer and more inclusive healthcare system. The exploration of the connection between lifestyle factors and health expenditure patterns holds significant importance for healthcare analysis and policymaking. This objective seeks to uncover how choices such as diet, exercise, Choice of Food: The analysis reveals that individuals who consume fast food have a slightly higher mean health expenditure of 2916.67, compared to those who opt for homely cooked food with a mean expenditure of 2800.24. However, the t-test result demonstrates no statistical significance , suggesting that the choice of food might not strongly influence health expenditure patterns. Timely Food Consumption: The data indicates that individuals who consume food on time exhibit a slightly higher mean expenditure of 2876.62, while those who do not consume food on time have a mean expenditure of 2606.06. The t-test result also reveals no statistical significance , indicating that the timing of food consumption might not be a significant driver of expenditure differences. Use of Intoxicants: The analysis shows that individuals' not using intoxicants have a mean expenditure of 2699.06, while those using intoxicants have a notably higher mean expenditure of 3229.63. that smoking might not be strongly correlated with expenditure differences. The table's findings suggest that while some lifestyle factors exhibit associations with health expenditure variations, others may not have significant impacts. The results emphasize the complex interplay between lifestyle choices and health-related spending patterns. While certain behaviors, like using intoxicants, appear to be correlated with higher health expenditures, other factors, such as food choices and smoking, seem to have less pronounced relationships. These insights contribute to a nuanced understanding of how lifestyle factors can influence healthcare costs. This table presents a comprehensive analysis that explores the relationship between individuals' general health status and various lifestyle factors. Choice of Food: Individuals who consume homely cooked food exhibit a slightly higher mean general health status of 3.8465, compared to those who opt for fast food with a mean health status of 3.7836. However, the t-test result indicates no statistical significance , suggesting that the choice of food might not strongly impact general health status. Timely Food Consumption: Individuals who consume food on time have a mean general health status of 3.7904, while those who do not consume food on time have a slightly higher mean health status of 4.1212. The t-test result indicates no statistical significance , suggesting that the timing of food consumption may not significantly affect general health status. Use of Intoxicants: Individuals not using intoxicants have a mean general health status of 3.7970, while those using intoxicants have a slightly higher mean health status of 3.8696. The t-test result indicates no statistical significance , suggesting that the use of intoxicants might not have a significant impact on general health status. The results indicated that the choice of food, timely food consumption, and smoking were not statistically significant predictors of health expenditure, while the use of intoxicants showed a positive and statistically significant relationship . The regression model demonstrated limited explanatory power in capturing the variation in health expenditure based on the examined lifestyle factors. The regression results suggest limited explanatory power of the selected lifestyle factors on health expenditure. The low R Square and insignificant ANOVA F-statistic indicate that the model as a whole does not significantly explain the variance in health expenditure. While the use of intoxicants appears to have a statistically significant positive relationship with health expenditure, the magnitudes of the coefficients are generally small. The lack of statistical significance for other lifestyle factors, like choice of food and smoking, suggests that these specific factors might not have strong predictive capabilities regarding health expenditure patterns. The hypothesis tests were conducted to examine whether various demographic factors have a significant influence on health expenditure patterns among individuals in the Malappuram district. The results suggest that certain demographic factors, including religion, income, marital status, occupation, age, gender, and family structure, are associated with variations in health expenditure patterns among individuals in the Malappuram district. These findings provide valuable insights into how these demographic characteristics influence individuals' healthcare spending behaviors. Similarly, the analysis suggests that while the use of intoxicants shows a statistically significant relationship with health expenditure, the other examined lifestyle factors, namely the choice of food, timely food consumption, and smoking, do not exhibit a significant relationship with health expenditure patterns among individuals in the Malappuram district [21][22][23]. --- CONCULISON In conclusion, the comprehensive analysis undertaken in this study has illuminated various aspects of the complex interplay between individual health status, lifestyle choices, and healthcare expenditure patterns. The study aimed to address a range of objectives, including studying the health status disparities across different religious affiliations, assessing variations in medical expenditure based on household income levels and socio-economic status, and exploring disparities in medical expenditure allocation based on age, gender, and family size. The analysis of health status across religious affiliations indicated that while slight variations existed, they were not statistically significant. However, this observation underscores the need for continued efforts to ensure equitable healthcare provisions across diverse religious communities. An important dimension of the study involved investigating the economic implications of lifestyle choices. While certain lifestyle factors displayed associations with health expenditure variations, the effects were often modest or statistically insignificant. This highlights the intricate nature of healthrelated spending patterns and the need to consider a broader range of determinants when addressing healthcare costs. Furthermore, the regression analyses conducted to understand the relationship between lifestyle factors and health expenditure provided valuable insights. The modest R-squared values and non-significant Fstatistics indicated that the selected lifestyle factors might not have substantial predictive power in explaining variations in health expenditure. However, the statistically significant relationship between the use of intoxicants and higher health expenditure underlines the potential impact of certain behaviours on healthcare costs. Throughout this analysis, it became evident that healthcare expenditure patterns are influenced by a multitude of factors, extending beyond individual lifestyle choices. Socio-economic factors, access to healthcare, and other unobserved variables can significantly contribute to the observed expenditure patterns. As a result, these findings emphasize the need for a holistic approach to healthcare policy and individual decision-making. Policymakers should consider not only lifestyle interventions but also broader systemic improvements to ensure equitable access to healthcare and effective cost management. Individual choices, while impactful, are just one facet of the intricate web of factors shaping healthcare expenditure patterns. In conclusion, this study underscores the complexity of healthcare expenditure, the multifaceted nature of individual health status, and the interplay of lifestyle choices with broader socio-economic and systemic dynamics. --- COMPETING INTERESTS Authors have declared that no competing interests exist.

This study investigates the relationships between demographic factors and health expenditure based on 387 households in Malappuram district between September 2021 and April 2022. Rigorous statistical methodologies, including ANOVA and t-tests, were applied to scrutinize and quantify these relationships. The findings indicate that health status variations across religious affiliations were present but not statistically significant. Similarly, the regression analysis showed limited predictive power of lifestyle factors (low R-squared, non-significant F-statistics) in explaining health expenditure variations. However, it did reveal a significant correlation between intoxicant use and increased healthcare costs, emphasizing the influence of specific behaviors on healthcare expenses. furthermore, the analysis emphasized the multifaceted nature of healthcare spending, influenced by socio-economic factors, healthcare accessibility, and unobservable variables,

asthma prevalence at lower incomes. Despite a long history of research on asthma, the disparities still exist. Original research focused on the causes and treatments at the individual level. More recent research has begun to investigate family, neighborhood, and societal factors that may influence childhood asthma. This integrative review analyzes research from the year 2010 to 2016 with regard to the quick pace of developments in the field and allows a few years for the 2007 National Asthma Education and Prevention Program guidelines to be implemented. Neighborhood-level factors are being investigated using sophisticated statistics, accounting for several domains in their modeling. Neighborhood is defined as a "bundle of spatially based attributes associated with clusters of residences, sometimes in conjunction with other land uses" . Neighborhood-level factors include structural characteristics of buildings , infrastructural characteristics , demographic characteristics of residents, class status characteristics , tax/public service package characteristics , environmental characteristics, proximity characteristics , political characteristics, social-interactive characteristics, and sentimental characteristics such as how residents identify with their neighborhood . Defining neighborhood-level factors demonstrated several areas that might be of interest to school nurses including schools themselves. Both social and physical aspects of neighborhoods affect health. The neighborhood still remains an important area to investigate because it is linked with more severe asthma and greater risk of death. Although there is not one definitive causal factor, it has been proven that residential segregation, both past and present, leads to disparities in socioeconomic status , causing racial health disparities between African Americans and Caucasians . Racial and ethnic minority groups and those living in low-income communities are exposed to multiple physical environmental triggers as well as social stressors related to poverty and inequality . These physical and social affects have negative health consequences for children with asthma. According to the NAEPP Expert Panel Report 3 , one risk factor for asthma-related death is being an inner-city resident or of low SES. This statement emphasizes how important it is to investigate the impact of neighborhood-level factors on children's asthma. Analyzing current research in this area will allow scientists to develop interventions to combat asthma at this level. As is mentioned in previous reviews on this topic , a multilevel framework would support and encourage multidisciplinary investigation in asthma-related research. This review also includes neighborhood-level factors that may be important for school nurses to consider when treating students with asthma. The purpose of this integrative review is to analyze and critique research from 2010 to 2016 that investigated the relationships between neighborhood-level factors and asthma in children living in urban areas. --- Literature Search Method Integrative reviews present the state of the science by compiling recent empirical or theoretical literature to provide a comprehensive view of a health topic . The purpose of this integrative review was to critique and summarize research examining the relationship between neighborhood-level factors and asthma in children living in urban areas. A literature search of CINAHL, Pubmed, and PsycINFO databases was conducted in March 2016 with the assistance of a university librarian. Search terms included the Medical Subject Headings terms "neighborhood," "asthma," "child," and "urban." See Figure 1 for a screenshot of the search strategy. The search was performed for the years 2010 through March 2016. Articles were found from this search as well as a hand search after speaking with experts in the field and reviewing reference lists to identify relevant articles. There were no restrictions on primary author discipline, location of the study, or methodological orientations. Due to the nature of neighborhood-level factors being relatively difficult to manipulate, all of the studies included were nonexperimental. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram illustrates the search process. The inclusion criteria were full text, published, peer-reviewed, and English language. Studies based on adult asthma, trialing interventions, dissertations, reviews, opinion articles, and articles published prior to 2010 were excluded, leaving 26 studies after 3 additional studies were added from a hand search. The presentation of results is organized conceptually discussing research on prevalence as well as physical and social neighborhood-level factors that are associated with asthma. The next section is a critique followed by implications for future research. --- Results Articles were analyzed by the type of study; method; results; asthma measurement; inclusion of sex, race, and ethnicity; and SES. See Supplementary Tables 1-6 for a detailed analysis of each individual article. During analysis, articles were evaluated critically regarding theoretical and methodological features. After analysis of all articles, six themes emerged organically from the data. The results are presented according to the theme below, beginning with research of asthma prevalence followed by physical then social neighborhood-level factors. --- Prevalence Several recent studies demonstrate that asthma prevalence varies by race as well as SES. Two large national studies demonstrated that asthma rates are significantly higher for children who are African American or Puerto Rican . These studies also analyzed data by SES and one demonstrated that a one-unit decrease in the household income to poverty ratio was associated with a 7% increase in the odds of prevalent asthma . A similar study was done with data from the Latin American International Study of Asthma and Allergies in Childhood to analyze multiple factors relating to asthma prevalence . This study also found that the Gini index, which is a measure of income distribution similar to SES, along with crowding, sanitation, infant mortality rate, and homicide rate, explains 75% of the variability in asthma prevalence between 31 Latin American urban centers . This study, along with the two discussed earlier, utilized a cross-sectional design that prevents any causational inferences to be drawn. Fattore, Santos, and Barreto measured socioeconomic and environmental determinants of wheezing in children in Latin America as described above. Similarly, Holt, Theall, and Rabito and Keet et al. analyzed data on the sociodemographic, neighborhood, familial, and individual level. Holt et al. found that the mother's race and ethnicity, child's insurance coverage, child's sex, allergies, the physical condition of the exterior of the home, and the percentage of the population with a bachelor's degree were significantly associated with an asthma diagnosis by age 5 in children . The variables were significant when analyzed in a multivariable model including individual-level sociodemographic, housing characteristics, interviewer observed/ participant-reported neighborhood characteristics, and census tract-level neighborhood variables. Keet et al. sampled children living in various census tracts around the United States and found that race and ethnicity were strongly associated with asthma prevalence, but in a multivariate analysis they found that neighborhood poverty and urban/rural status were not risk factors. In a similar study, Ownby et al. demonstrated that asthma prevalence, when controlling for SES, is similar among rural and urban African American teens. These data also included asthma symptoms and demonstrated that the teens living in urban Detroit reported more symptoms and hospitalizations in the past 30 days compared to the teens living in rural Georgia . The cross-sectional data were from self-report questionnaires that were sent out to students through public schools in each area during the years 2007-2008 in Detroit and2009-2010 in Georgia. These two samples were taken 2 years apart and had differing sample sizes which weaken the strength of the authors' argument. Although there have been unique findings as demonstrated by Ownby et al. , methodologically stronger research should be done to support the evidence that urban versus rural location are not associated with asthma prevalence. Taken together, this body of research demonstrates that asthma disproportionally affects African American and Puerto Rican children in the United States, and it seems to correlate with SES nationally and internationally. There is a discrepancy as to whether or not ones neighborhood is associated with increased asthma prevalence, however, being an "inner-city resident" is cited as a risk factor for asthma mortality, and it is important to continue to investigate the elements of this environment that are related to asthma prevalence and control. --- Physical-Outdoor Air Research of outdoor air-quality ranges from analyzing pollen exposure related to asthma emergency department visits to the impact of weather and particular air pollutants on asthma and respiratory health. Some of the research has mixed results. Jariwala et al. found a correlation between pollen exposure and asthma ED visits but could not demonstrate a relationship between nitrogen oxides or ozone peaks and peak visits for asthma to the ED. Glad et al. utilized a case-crossover design to demonstrate a 2.5% increase in asthma ED visits for every 10 parts per billion increase in O 3 level . The case-crossover design may have strengthened the method, but the pollution data were obtained from a limited number of monitors and therefore did not capture variability across regions. Through a retrospective case-control study, Nishimura et al. demonstrated a 5 ppb increase in average nitrogen dioxide during the first year of life was associated with an OR of 1.17 for asthma . This study also demonstrated that Chicago and New York had the highest level of NO 2 , considered to be a traffic-related pollutant. The study design also addressed the temporality of air pollution, but the data were analyzed as large areas rather than particular homes of the study participants. Svendsen et al. demonstrated in Texas that a 10 ppb increment increase in NO 2 levels was associated with current asthma for children in public schools at elevations above 1,170 m. In valley schools of lower elevation, the relationship was not significant. The environmental data were collected in 1999 and the questionnaires from the schools were obtained in 2001. The final analyses focused on children who had lived in their current residence for over 1 year. These analyses controlled for indoor sources of exposure including gas stoves and secondhand smoke. Buonanno, Marks, and Morawska and Vieira et al. were able to measure individual pollutant exposure, thereby strengthening their inferences. In Italy, Buonanno et al. demonstrated that children in urban areas are exposed to particle concentrations 25% higher than the overall mean exposure . The author also demonstrated that home was a major contributor to daily dose of airborne particles and that there is an association between personal dose and respiratory health effects. Vieira et al. demonstrated a low-cost method to measure NO 2 and O 3 exposure among children by using small air filters in their homes, balconies, and hooked onto their backpacks. Although the sample size was small , the study found that the NO 2 in indoor air and personal exposure to O 3 were independently associated with asthma and wheezing at any time . These data support research that air pollution in the form of NO 2 and O 3 do relate to asthma, and their exposure early in life may increase the odds of an asthma diagnosis. Overall there appears to be strong evidence supporting an association with NO 2 and asthma, although not as much information on O 3 , and limited research on the association between pollen and asthma. More research should be done on individual-level exposures possibly by the methods used by Vieira et al. to measure outdoor, indoor, and personal exposure. --- Physical-Outdoor Air-Traffic In analyzing traffic-related pollution, Cook, deVos, Pereira, Jardin, and Weinstein and Li et al. used proximity of houses to major roads and estimated traffic counts to analyze asthma. They both used case-control studies to demonstrate increased risk of asthma-related events such as ED visits. Cook et al. demonstrated that in Perth, Australia, there was a 24% increase in the risk of experiencing multiple emergency contacts for asthma for every log unit of traffic exposure . Li et al. demonstrated that in Detroit, MI, for every 1-km increase in distance away from a primary road, the OR of an asthma event is 0.97 . Although they used large sample sizes, N = 434 , and N = 14,646 , neither study took into account SES, sex, or ethnicity, which might explain some of the variation in asthma events. These data did not account for other exposures such as secondhand smoke, family history, or indoor air pollutants and allergens because they were deidentified from hospital records in Australia and Medicaid data in Detroit. Although this information does not demonstrate causation, these data about specific pollutants appear to have an association between outdoor air quality and asthma. --- Physical-Indoor Air-Triggers and Housing Research investigating indoor variables and asthma includes types of housing , specific triggers , and behaviors aimed at decreasing triggers . Northridge, Ramirez, Stingone, and Claudio determined that there is a high prevalence of asthma in public housing in New York City after adjusting for community-level demographic and economic factors. Of those sampled, those in public housing were more likely to report the presence of cockroaches compared to other types of housing. This finding demonstrates differential exposure to asthma triggers by housing type. The researchers analyzed questionnaires from 4,853 children representative of the 5-to 12-year-old population of children attending New York City public schools. Olmedo et al. used a case-control design to analyze high asthma prevalence neighborhoods compared to low asthma prevalence neighborhoods . Bed dust was tested for allergens, and in HAPNs there were more cockroach , mouse , and cat allergens and lower dust mite allergens . The author theorized that air conditioning was less common in HAPNs, and therefore, the dry air was less conducive to dust mites. This idea was supported by Northridge et al. who found that public housing residents, who had higher asthma prevalence, were less likely to report the use of an air conditioner compared to the residents of family dwellings. Another asthma trigger analyzed by Butz et al. is exposure to secondhand tobacco smoke measured through urine cotinine levels. Although no significant associations were found, it should be noted that children with a greater number of symptom days had the highest cotinine concentrations. Data also indicated a twofold increase in urine cotinine concentrations, when the child's caregiver smoked as compared to other people living in the house. The sample consisted of children with uncontrolled asthma, so the lack of variability may have led to the lack of statistical significance. The study also demonstrated that urine cotinine, as opposed to passive nicotine sampling badges in the environment, should be considered the gold standard for measuring secondhand smoke exposure in children. Martin et al. also utilized cotinine data in saliva to analyze asthma trigger behaviors for Puerto Ricans in Chicago . Caregivers of children with asthma were scored through questionnaire, home assessment, and cotinine levels to determine their trigger behavior summary score. Each item was coded as 1 for affirmative and 0 for negative, and they included dust in the bedroom; dust in the home; lack of an air cleaner; lack of allergen covers; pets in the bedroom; pets in the home; presence of roaches, rodents, cracks/holes; strong cleaning smells; smoke exposure; fragrance; and/or mold. The average trigger score was 6.0 . All items on the trigger behavior summary score represent modifiable behaviors, so these data have identified areas for future environmental interventions. Fragrance use was the most common behavior followed by presence of mold . Although the study by Martin et al. identified triggers, Yinusa-Nyahkoon, Cohn, Cortes, and Bokhour analyzed qualitative data to address family routines and asthma management in African American families in Boston, MA. Two of the four main themes were that caregivers worked to manage the air quality in their home and frequently cleaned the home to remove triggers. Over half of the study sample lived above the poverty line, almost half lived in two-parent houses, and all had health insurance and a regular source of ambulatory care. Although this study may not be transferable to "inner-city" populations, it would be interesting to interview families of lower SES to understand how their family routines and asthma management differ. --- Social-Outdoor-Safety Measuring safety was done both directly by looking at crime data and indirectly by measuring the perception of safety . Gupta et al. utilized surveys of children from kindergarten through eighth grade . The surveys asked the child's caregiver about demographic information as well as whether the child had been diagnosed with asthma, age at diagnosis if applicable, current asthma status, and presence of a family history of asthma. In total, 5,908 children reported an asthma diagnosis. After geocoding all addresses into 247 Chicago neighborhoods, they were divided into quartiles based on asthma prevalence. These data were analyzed with crime data from the Chicago Police Department. Criminal activity was significantly higher in neighborhoods with high asthma prevalence. After adjusting for community-level race and ethnicity, violent crime continued to explain 15% of the variation in childhood asthma . These data only account for reported crime, so prevalence of asthma may be underestimated. This study demonstrated the utility of directly measuring crime, while Coutinho, McQuaid, and Koinis-Mitchell and Vangeepuram, Galvez, Teitelbaum, Brenner, and Wolff measured perceived safety. Coutinho et al. utilized mixed methods to analyze caregiver/child dyads , and their perception of safety related to family asthma management. There were significant ethnic group differences in poverty status, F = 6.49, p <.01, with Latino and African American dyads reporting greater poverty than the non-Latino White dyads. There were also differences in perceived safety by ethnicity, F = 6.29, p < .01, with non-Latino White dyads reporting higher perceived home and neighborhood safety compared to the Latino and African American dyads. Ratings of effective family management positively correlated with home and neighborhood safety and negatively correlated with perceived discrimination and acculturative stress . Vangeepuram et al. used cross-sectional data from parents of 6-to 8-year-olds with asthma, in New York City. They found that parents who reported feeling unsafe walking in the neighborhood were more likely to have a child diagnosed with asthma . An interesting finding in these data is that the neighborhoods sampled were relatively similar, but it was the variation in parent perception of safety that related to asthma prevalence. All three of these studies demonstrate that the direct and indirect measures of safety may be correlated with asthma. Koinis-Mitchell, Kopel, Salcedo, McCue, and McQuaid used a mixed methods exploratory study with cross-sectional data from two separate studies to analyze children's stress perception related to their asthma. Children with poorly controlled asthma reported a higher rate of neighborhood stress compared to children with wellcontrolled asthma, F = 4.5, p = .04. The family moving was most frequently mentioned as a stressor for respondents. Children with poorly controlled asthma reported higher levels of stress related to being afraid to go outside compared to children with well-controlled asthma, F = 6.1, p = .02. The study authors reported that a limitation was that there was not a parent instrument to corroborate children's perceptions of concepts, but the mixed methods nature of the study allowed the children to explain their answers. --- Social-Indoor-Family Stress Family stress is analyzed in several different ways, taking into account the family and the parent . Family-level data were obtained by Koinis-Mitchell et al. and Sampson et al. using mixed methods. The data Sampson et al. presented demonstrated that caregivers of children with asthma from low-income families in Detroit, MI, do not see caring for an asthmatic child as a source of stress. Rather, they reported high stress with regard to asthmarelated change, uncertainty, control, and anxiety. Koinis-Mitchell et al. calculated a cumulative risk index to quantify the risk factors for asthma. The CRI score was formulated by adding a 1 for the presence of six risk factors including neighborhood stress, poverty, acculturative stress, discrimination, asthma severity, and asthma triggers within the home. The study showed that cumulative risks were associated with increased functional limitation and risk for an ED visits in the past 12 months . Quinn, Kaufman, Siddiqi, and Yeats analyzed both parent perceptions of collective efficacy and neighborhood order for a sample of lowincome Chicago families with a child who has asthma. Collective efficacy is the mutual trust between neighbors and the willingness to intervene on behalf of the public good . Neighborhood order was defined as observable physical and social decay operationalized as presence of litter/trash, graffiti, vacant houses, and asking about trust between businesses and residents. Parent health was associated with collective efficacy . General health outcomes of the family, including parents, were associated with collective efficacy, but child respiratory health outcomes were associated with neighborhood order. No biological or behavioral risk factors were accounted for in the statistical analysis, and the authors note that reverse causality may be a limitation since asthma may cause stress. Due to the cross-sectional design of this study, causality cannot be demonstrated. Both Mathilda-Chiu, Coull, Wooley, and Wright and Otsuki et al. utilized prospective designs to demonstrate temporality between maternal stress and depressive symptoms and wheeze and asthma morbidity. Mathilda-Chiu et al. analyzed maternal stress during the prenatal and postnatal period and then subsequent wheeze as reported by predominantly Latina and African American mothers in Boston, MA. Analysis of 989 mothers revealed that children born to mothers with high stress in both pregnancy and the postpartum period were significantly more likely to have repeated wheeze compared with children of mothers reporting low stress in both periods. The study followed children up to 2 years old and assumed that wheeze was a proxy for those that may be more likely to develop asthma or have compromised lung function. Otsuki et al. analyzed data at baseline and 6 months regarding maternal depressive symptoms and child asthma morbidity. This large sample of African American mothers of children with asthma demonstrated that maternal depressive symptoms at baseline predicted children's asthma symptoms 6 months later . It was shown that maternal depressive symptoms at baseline did not predict asthma ED visits at 6 months, but the sample was recruited from the ED during high morbidity, so these data may not be the best outcome measure. --- Discussion --- SES, Race and Ethnicity, and Sex In general, most international articles did not identify the race and ethnicity of their samples. Two of the four international articles did mention the SES of their sample because they were looking at social inequity as the main variable related to asthma . The data from the United States also frequently accounted for SES. It was operationalized in different ways; Northridge et al. , Holt et al. , and Keet et al. used census data as a proxy for neighborhood SES, some discussed if sample participants lived above or below the poverty line , one was not able to get SES information due to the type of data they were using , and several mentioned "controlling for SES" in analysis . More than half of the studies from the United States sampled specifically African American and/or Puerto Rican children due to the high prevalence of disease among this race and ethnicity. Several of the studies analyzed sex as it related to the outcome variable. A few of them had small sample sizes, so this stratification was not possible. Several of the stress studies sampling caregivers ended up with predominantly female caregivers as their sample. Because there was always a high percentage of caregivers who were female it was noted but data were never analyzed separately depending on the sex of the caregiver. When analyzing neighborhood-level factors related to asthma in children, it is important to account for unique contextual variables in analysis. As mentioned above, some of the reviewed articles did not address these key variables, thus adding complexity to the interpretation of their variables. --- Method As mentioned in Results section, a common research design for these studies is cross sectional. Data were frequently obtained from baseline questionnaires of intervention studies or other types of secondary data. These data can demonstrate correlations, but the design prevents an inference of causation between variables. They were also frequently gathered retrospectively and missing measurement of important variables. The few prospective studies were helpful to analyze maternal mental health and wheezing and asthma morbidity . Prospective data add temporality to relationships thus strengthening the evidence, but without an experimental study, causation cannot be inferred. The article by Otsuki et al. was also the only article to use path analysis to model complex relationships between several variables over time. Another frequent analytic technique was logarithmic regression, which is appropriate, given binary outcome data. Although logarithmic regression can test for confounding or effect modification, it does not allow for modeling complex relationships like path analysis or mixed effects modeling. Another statistical trend was to dichotomize or group data originally gathered continuously. This change in the level of statistical analysis from parametric to nonparametric weakens the strength of the evidence. Some authors noted this as a limitation, but others accepted it as a requirement of their study due to non-normal distributions of results. Sample selection also varied across studies as shown in the studies that sampled from medical records such as ED data, primary care clinics, Medicaid data, and urgent care call centers. Each type of medical data lends itself to a different sample and therefore results should be interpreted within that sample and are not generalizable to other samples. Several studies also sampled from schools that have implications depending on whether they are public versus private, and the specific demographics of their students. A few of the studies were explicit in their reasoning for sampling at a specific school, but others neglected to present this information in their article. The sampling methods differed depending on the purpose and pragmatic challenges of the study. A critique of the methods for most of the air-quality studies, aside from the study by Vieira et al., was the lack of individualized data for each study participant. The authors mentioned that using generalized air pollution data could inaccurately estimate the amount of exposure for an individual. Vieira et al. used a unique and inexpensive method for attaching air filters to children's backpacks and placing them around their homes for a more precise estimate of exposure. --- Asthma Operationalization Supplementary Tables 1-6 also present how asthma is operationalized in each study. Some studies utilized health-care data and International Classification of Diseases, Ninth Revision or ICD-10 diagnosis of asthma or "status asthmaticus." Several studies asked for parents to report whether or not their child had physician-diagnosed asthma and their symptom frequency. Self-report can be limiting if the caregiver cannot recall this diagnosis or frequency of symptoms. This might happen if the child is in day care, if the parent works long hours, or if more than one child has asthma. Of note, other health-care providers, nurse practitioners for example, also frequently diagnose asthma. Some further stratified type of asthma by analyzing parent report of control, severity, and functional limitation. Another common theme was to look at ED visits for exacerbation. Several studies cited the NAEPP Expert Panel Guidelines for how they measured asthma. This resource from the National Heart Lung and Blood Institute of National Institutes of Health provides details on how to diagnose asthma severity, assess persistence and control, and provides a good universal method to operationalize this concept. --- Limitations This review has some limiting factors. One possible limitation of the inclusion criteria is that articles were from 2010 to March 2016. Although this time frame was justified considering frequent advances in the field, some important articles might have been excluded. Research from 2010 to March 2016 appeared to build on previous research and allowed examination of results after dissemination of the NAEPP guidelines. Thus, expanding the search time frame might not have added any significant articles. There is also a chance that search terms related to neighborhood-level factors were not included in the search. For this reason, a university librarian was consulted to help design the search. MeSH terms as well as search terms were used to attempt to capture all of the required terms without being too specific, but there is a chance that an important term was excluded. Another possible limitation was the use of the term "urban" in the inclusion criteria. Some of the articles in the review demonstrated that urban versus rural status may not be an important factor in asthma prevalence. Although this may be true, the focus on this review was urban children, and therefore it was a critical search term to be included. --- Implications for Future Research There are specific as well as general suggestions for future research in the above areas. Research into prevalence should be clear about what is being measured. Evidence presented argued that SES, and not place, drives asthma disparities. However, as the connection between SES and asthma was not the specific focus of this review, analyzing more research in this area would help support or refute these findings. Fattore et al. recommended continuing prevalence studies discriminating between phenotypes of asthma, addressing the geographic and individual level, and incorporating variables related to socioeconomics and urbanization. On that note, as it was suggested several times, outdoor air pollution data need to use more individualized methods . If these small-scale studies are done frequently in differing populations, climates, and geographic areas, then a meta-analysis would help to draw conclusions about the overall risk for humans . Nishimura et al. also reported that increased exposure to NO 2 in the first year of life statistically significantly increased the odds of being diagnosed with asthma, but the regions where study participants lived had NO 2 levels lower than required by the Environmental Protection Agency air-quality standard. This suggests the need for further research to specifically test NO 2 levels. Current EPA requirements may not be sufficient to protect the health of children, and the EPA standards might need to be increased. The studies relating proximity to traffic and asthma also represent areas where public health indicated the requirement for changes in policy. Some policy changes are already underway. Currently, the EPA has a Clean School Bus program to reduce diesel exhaust from school buses. They have developed tougher emission standards for new school buses and fund projects that reduce emissions from existing diesel engines. One of the simplest ways to decrease emissions is to reduce idling in school parking lots where children are exposed . This environmental health issue should be investigated further particularly in the school environment where school buses park emitting exhaust while waiting to load children. School nurses should lead the charge to implement policies to reduce the exhaust exposure among school children. Several of the studies discussing indoor air suggest a multipronged approach to reducing home triggers. There should be more research and interventions targeting the individual level whether that is smoking cessation or discussion of how to eliminate other triggers such as cockroaches and mold, using integrated pest management practices as well as policy changes around housing, pollution control, and tobacco legislation. Northridge et al. noted that because many of the houses are multiunit dwellings, the families living inside them have limited control over the maintenance of the spaces and therefore limited ability to control triggers that affect their children. Therefore, policy changes around housing for those on government assistance should take into account the health of children in determining what type of building meets their requirements. The safety research demonstrated correlations between actual crime in neighborhoods , perceived safety , and asthma. All three studies recommend additional research in this area to investigate the mechanisms behind the associations , specific psychological measures related to violence exposure , and control for potential confounders in the built environment . More data are needed in this area to develop interventions to treat this problem. Regarding other sources of stress, future research should continue to investigate the mechanisms of how chronic stress effects children over time. The range of studies recommended looking at maternal depression starting in pregnancy, evaluating interventions to treat the depression, and further investigating the relationship between prenatal depression and future wheeze. Stress studies need to analyze psychological and physiological effects of stress while taking into account the other variables noted to affect asthma including the physical environment. Analyzing the current research in neighborhood-level factors affecting asthma clearly shows that the next step is multilevel research into factors associated with asthma. As mentioned earlier, mixed effects modeling would be the optimal statistical analysis due to complex relationships between many differing levels of variables. SES, for example, is related to asthma and demonstrated by research in this review and previous research, but it acts through different mediators such as housing. To appropriately measure the relationship between SES and asthma, several pathways need to be included in the model to examine multiple associations between these variables, and therefore mixed effects modeling is the best statistical analysis to use with these types of data. This review has analyzed the latest research on neighborhood-level factors that affect asthma in children living in urban environments. The current research has demonstrated a gap in studies analyzing multiple levels of influence through advanced statistical methods. These data can be used to better understand relationships of multiple factors as well as to influence health policy to improve the neighborhood environment for those children most affected by asthma. --- Implications for School Nurse Clinicians This review directly applies to school nurse clinicians who frequently monitor and care for children with asthma. As health-care providers, working in the community where high-risk children reside, they should be informed of the physical and social neighborhood-level factors that contribute to childhood asthma. They have a unique opportunity to target educational and behavioral interventions as well as policy decisions toward these determinants of health. This review also serves as a reminder to assess children with asthma holistically, within the context of their physical and social neighborhood environment. --- Acknowledgments --- Supplementary Material Refer to Web version on PubMed Central for supplementary material.

Asthma disproportionately affects children who are non-White and of low socioeconomic status. One innovative approach to address these health disparities is to investigate the child's neighborhood environment and factors influencing asthma symptoms. The purpose of this integrative review is to critique research investigating the relationships between neighborhoodlevel factors and asthma morbidity in urban children. Three literature databases were searched using the terms "asthma," "child," "neighborhood," and "urban." The articles included were organized into six themes within the larger domains of prevalence, physical, and social factors. Literature tables provide in-depth analysis of each article and demonstrate a need for strengthening analysis methods. The current research points to the necessity for a multilevel study to analyze neighborhood-level factors that are associated with increased asthma morbidity in urban children. School nurse clinicians, working within children's neighborhoods, are uniquely positioned to assess modifiable neighborhood-level determinants of health in caring for children with asthma.

Introduction Schizophrenia is one of the most debilitating mental disorders. For a significant portion of individuals who suffer from this disorder, onset occurs in young adulthood, arresting important social and educational development that is necessary for future successful labor force participation. As a result, schizophrenia is often associated with an eventual downward spiral ending in poverty and isolation [1]. To avert this dismal future, the focus on the first psychotic episode is becoming a priority for mental health care globally . There is also growing evidence surrounding the low employment rates of people with FEP. Employment rate estimates range from 13 to 55% [3][4][5]. Reviews of the literature [3,6] note that there are a number of factors associated with employment along a number of dimensions [7][8][9]. Along with cognitive impairment and symptoms, educational attainment, family socioeconomic status, social benefit structures, and labour market conditions also potentially affect employment of young adults with first episode psychosis [3]. There appears to be a complex mix of both individual and environmental factors linked to successful labour market participation. Although it is a multifaceted challenge, there is increasing evidence for the effectiveness of interventions such as Individual Placement and Support programs and the interventions of vocational specialists for this population [4,6,[10][11][12][13]. As such, the International First Episode Vocational Recovery Group [14] released a consensus statement advocating for the "right to education, training and employment" for young people enrolled in early intervention programs. The need for the iFEVR consensus statement is motivated by the mounting evidence that employment is often not a priority for providers who serve and support FEP clients. For example, Rinaldi and colleagues [6] note that few FEP studies have focused on employment and education outcomes. In fact, employment is often discouraged by well-intentioned medical professionals and family [6,10,15]. In addition, governments have been less than supportive of employment programs for people with mental illnesses [16]. The purpose of this paper is to contribute to the literature about clients enrolled in first episode psychosis programs and psychosocial outcomes by examining the factors associated with paid employment among young adults who have experienced their first psychotic episode. In this paper, we consider the association of socioeconomic factors with employment. --- Background 2.1. Characteristics of FEP Programs. FEP programs are designed to facilitate recovery from psychotic illnesses that first arise during youth, most commonly schizophrenia and bipolar disorder, and have historically been most associated with very high levels of disability [17]. Programs are designed to provide services to both clients and their families. Services include comprehensive diagnostic assessment, treatment, psychosocial supports, as well as family education and support. There is emphasis on a multidisciplinary team approach that integrates medical treatment, counseling, case management, substance abuse treatment, cognitiv behavioural therapy, in addition to psychoeducation for families. --- Characteristics of the Study FEP Programs. The six FEP programs included in this study were located in six regions throughout the province of Ontario, Canada. Each program accepted youth who were experiencing their first episode or early stages of psychosis. Each early intervention program was developed to exclusively provide outpatient services. Five of the six FEP programs reported that nearly threequarters of their clientele were males. Three of the FEP programs engaged clients aged 14 to 35 years old, while the remaining three programs limited access to people who were at least 16 years old. Two FEP programs engaged transitional aged youth 16 to 23 years old. All but one of the six programs were located in established community mental health agencies. The exception was a community-based program that was part of an acute care hospital. The number of staff members in each of the six programs varied considerably, ranging from three parttime positions to 10 full-time equivalent positions. Each of the FEP programs was developed in accordance with the guidelines and standards set forth by the International Early Psychosis Association and other pioneers in the field [17-20]. --- FEP Programs and Vocational Outcomes. There is evidence that compared to FEP clients who receive usual care, those enrolled in early intervention programs have significantly better educational and vocational outcomes [21]. The relatively better vocational outcomes associated with FEP programs may be associated with the shorter duration of untreated psychosis for these clients. Norman et al. [22] found that shorter duration of untreated psychosis and greater social support were significantly associated with either more full-time competitive work or full-time enrollment in school at three-year followup. In their review, Rinaldi and colleagues [6] note that when compared to those enrolled in community mental health teams, clients enrolled in FEP program seem to experience a relatively smaller decline in employment and education. They attribute this to a protective role played by FEP programs. Major et al. [5] observed a decrease in unemployment among FEP clients after 12 months even in the absence of a vocational focus. However, there is less evidence regarding the long-run outcomes for FEP programs versus usual care. After 18 months, the LEO study did not find [21] a significant difference between clients in specialize programs versus those who were not. After 5 years, the OPUS trial did not find significant differences in employment between clients who were in FEP versus usual care [23]. --- Work History and Employment. In the literature focusing on vocational outcomes in the adult population with severe and persistent mental disorders, it has been observed that work history is one of the most consistent predictors of employment [24]. However, because of their stage in life, young adults experiencing their first psychotic episode have not had the opportunity to accumulate labor force experience. This can place FEP clients at a disadvantage in finding paid employment. --- Educational Attainment and Employment. The literature also indicates that educational attainment is significantly related to employment status. People who have no high school diploma are more likely to be unemployed [25,26]. If young people have their education interrupted by their illness, they can be faced with another disadvantage in finding employment. --- Disability Benefits and Employment. There is also an association between receipt of disability benefits and unemployment [10,27]. The fear of losing disability benefits has been identified as one of the barriers to obtaining and maintaining employment for clients with severe mental illness [3,24,28]. As a result, those who receive benefits often have poorer employment experiences. Thus, there is accumulating evidence suggesting that in addition to treatment in FEP programs, socioeconomic factors such as educational attainment and income sources also play a role in employment outcomes. This type of evidence can help to inform future directions for the enhancement of psychosocial programs in FEP models. Interview participation criteria included willingness to be contacted by a study interviewer, ability to understand and give informed consent to be interviewed, and enrollment in one of the six participating FEP programs. Data were collected using face-to-face structured interviews In 2005, the participating FEP programs referred 45 clients of the total 161 enrolled, to be contacted by the study; of these, 33 were successfully interviewed. They represented 20% of the total clients enrolled. In 2006, 106 of 302 early intervention clients were eligible to be contacted by the study. There were 75 who were successfully interviewed. They represented 25% of the total clients. In 2007, 162 of 370 early intervention clients were eligible to be contacted for interviews. Of these, 107 were successfully interviewed. These clients constituted 29% of the total early intervention clients enrolled in the six programs. --- Methods --- Dependent Variable. An indicator variable was created to capture whether or not study participants had paid employment during the previous 12 months. --- Study Participant Characteristics. Indicator variables were created to capture whether the respondent was living with family , sibling, and spouse), education , gender , age at time of interview, receipt of provincial disability benefits as the primary income source and enrolled in school during the previous 12 months . An indicator variable was also created to capture whether participants were enrolled in their FEP program for more than a year. Three region indicator variables were created to capture the population density of the region in which the participant lived. Region 1 indicated that the participant resided in a region in which the population density was less than 100 people/km 2 . Region 2 indicated that the region of residence had a population density between 200 and 450 people/km 2 . Finally, Region 3 indicated that the region of residence had 3,929 people/km 2 . --- Analyses. Descriptive statistics were used to test for significant differences in the characteristics of people who had paid employment during the previous 12 months. Chisquare and Fisher exact tests were used to test the differences between the categorical variables. Logistic regression analysis was used to identify the factors associated with having had paid employment. The most parsimonious model was used based on the factors described in the literature that are associated with employment for people with serious mental illnesses. --- Results About 61% of the sample had paid employment during the previous 12 months. In Table 1, compared to those who did not have employment, among those who had employment, there was a significantly greater proportion of people with at least a high school diploma = 6.21, P = 0.013) and a significantly lower proportion who identified provincial disability benefits as their primary income source = 23.31, P < 0.0001). Table 2 contains the results of the logistic regression analysis. The regression model produced 83.2% concordant and 16.6% discordant predicted probabilities and observed responses. The Hosmer and Lemeshow Goodness-of-Fit Test The results suggest that the odds of having been employed in the past 12 months are significantly lower for people who have not completed high school and for clients who identified provincial disability benefits as their primary income source . There is also a significant positive association between employment and being enrolled in the FEP program for more than a year . --- Discussion In our sample, 61% of respondents had paid employment during the previous 12 months. This proportion is greater than that reported in the literature. For example, Singh et al. [29] observed that among the cohort of FEP clients, the past year employment rate was 31.3%. Similarly, Norman et al. [22] reported employment rates of 44%, while Turner et al. [30] reported an employment rate of 46%. We also observed a significant difference between the educational attainment of those who were employed and those who were not. Educational attainment is a significant concern among FEP clients [10,11]. Norman et al. [22] observed that 46% of clients in their sample had attained less than a high school education at time of program entry. Assessing the three-year outcomes of FEP clients, Singh et al. [29] reported that 53.6% participants had not attained their secondary education certificates. Our regression results underscore one of the major problems associated with the lack of educational attainment; educational attainment is associated with the ability to attain employment. Our results indicate that those who did not complete high school were less likely to be employed. This association corroborates what has been reported in literature [5,31]. These results echo the need highlighted by Cook [32] in her review of the literature that a significant proportion of people with serious and persistent mental disorders have their education interrupted by the onset of mental disorders. That is, people with severe mental illness are likely to have lower educational attainment than their counterparts who have not experienced a mental disorder. In turn, this places them at a disadvantage to successfully compete in the labor market. One of the arguments for early labour market participation is because of its effect on future employment. In their longitudinal study of people with severe mental illness, Bush and colleagues [33] found that those who were employed at the beginning of a 10-year period were more likely to remain steadily employed throughout. In contrast, those who were not employed at the beginning were more likely to remain unemployed throughout the period. It should be noted that Bush et al. [33] looked at a group of people with severe mental illness. It will be important for future studies to examine effects of early employment in the FEP population. Our regression results also indicate that enrollment in a FEP program for more than a year is significantly associated with employment. That is, clients who have been enrolled in a FEP program for more than a year are three times more likely to have been employed during the year than clients who were enrolled for less than a year. This is a trend that has been reported in the literature. After a three-year followup of FEP clients, Singh et al. [29] saw past 12-month employment rates increase from 25.3% at intake to 31.3% at followup. This could be reflective of the protective effect provided by FEP programs [6]. None of the FEP programs that participated in this study had vocational specialists. This suggests that vocational outcomes were not necessarily one of the main foci of the programs. Yet, there was improvement in vocational outcomes. At the same time, compared to FEP programs without a vocational focus, those with IPS programs and vocational specialists have reported relatively better vocational outcomes [4,6,[10][11][12][13]. Thus, although FEP programs appear to provide a protective effect, there is evidence that with a vocational focus, they can be more than protective. Finally, these results suggest that enrolment in a public disability benefit program also decreases the likelihood of being employed. Our results may reflect the fact that the people who receive disability benefits are too ill to be employed. Alternatively, the results could also be related to incentives associated with receipt of disability benefits. In her review of the literature, Cook [32] points out that few people leave public disability benefits due to employment. Our finding corroborates findings that disability benefits are inversely associated with employment rates. That is, as disability benefit levels decrease, employment rates increase, and the use of unemployment benefits decreases [34]. This suggests that publicly funded benefits may inadvertently create disincentives to work. For example, people might perceive that they will be penalized for working. In Ontario, the province in which this study was conducted, when people are employed and also receive benefits from the provincial disability support program, the program calculates half of the client's net monthly earnings, deducts part or all of his/her monthly child care and disability-related work costs, and subtracts this amount from the client's total income support [35]. Although in theory the earned income should substitute for income support, people who receive disability benefits and who work will see their disability income support decreased but not necessarily associate the decrease with their pay check. At the same time, the provincial disability benefit program also seeks to create an incentive for paid employment by offering clients an additional monthly $100 work-related benefit [35] if they are employed in a paid position. Yet, the incentive to work may not be sufficient if people are unsure about their long-term ability to maintain employment. There may still be a fear that if they lose their disability benefits, they will be without a safety net if they lose their jobs [36][37][38]. It will be important for future research to explore the mechanisms of designing safety nets that ensure that when people are unable to work, they will have income while encouraging employment when it is possible. It will also be a challenge for FEP programs to design programs that help clients to receive training for jobs in business sectors in which they can earn a living wage [32] and that can accommodate episodes of illness so clients can accumulate successful employment histories. Limitations. The results of this study should be considered in the light of its limitations. One of its major limitations relates to its generalizability. The clients who participated in these interviews may not necessarily be representative of all clients in FEP programs. Given that only clients who were able to provide informed consent were asked to participate, those who were the most sick would have been omitted. Thus, our results are a conservative estimate of the proportion who were unemployed if the most severely ill were more likely to be unemployed. Likewise, the regression results reflect associations for less severely ill people and may differ for the more severely ill; however, the latter group may also be less likely to be seeking employment. In addition, the sample only included people who were enrolled in early intervention for psychosis programs. Results may be different for people experiencing their first psychotic episode who do not receive services from a program specializing in first psychotic episode cases. --- Conclusions There is little in the literature focusing on employment in the FEP population examining the contribution of socioeconomic factors to employment status. The results of our analyses indicate that receipt of public disability benefits and high school education are important factors related to employment. They also suggest that if paid employment is to be used as one measure of psychosocial outcomes, it is an outcome that may require cross-sector collaboration among health, education, and social services.

Schizophrenia is one of the most debilitating mental disorders. For a significant portion of individuals who suffer from this disorder, onset occurs in young adulthood, arresting important social and educational development that is necessary for future successful labor force participation. The purpose of this paper is to contribute to the literature about clients enrolled in first episode psychosis programs and psychosocial outcomes by examining the factors associated with paid employment among young adults who have experienced their first psychotic episodes. In this paper, we consider the association of socioeconomic factors to employment. Our results suggest that in addition to treatment, socioeconomic factors such as receipt of public disability benefits and educational attainment are associated with employment status. These results can help to inform future directions for the enhancement of psychosocial programs in FEP models to promote paid employment.

Background Approximately 68,000 indigenous farmworkers from Mexico currently live in Oregon. National data suggest that about 20% of farmworkers are women, but there are no reliable estimates of the number of female indigenous Mexican farmworkers. Indigenous farmworker women often do not speak English or Spanish, but one of many indigenous languages like Zapotec, Mixtec and Triqui which are common in villages in southern and western Mexico. In recent years female farmworkers in Oregon's Willamette Valley have reported sexual harassment and sexual assaults at the workplace to local community service providers and farmworker advocates. In April 2009, the Equal Employment Opportunity Commission and the Oregon Law Center filed a sexual harassment lawsuit against Oregon's Willamette Tree Wholesale Company in response to complaints of sexual harassment and rape. EEOC filed the lawsuit in Oregon after successful prosecution of a sexual harassment case against a farm in California that resulted in a large settlement for the Latino woman farmworker. The verdict was reaffirmed on appeal in 2008. 1A review of the scientific literature on sexual harassment of indigenous Mexican farmworker women revealed no results. A wider search on sexual harassment of farmworkers yielded a 2010 qualitative study by Waugh on sexual harassment among immigrant non-indigenous Mexican farmworker women highlighted a lack of workplace protections against sexual harassment. Women reported engaging in tactics including ignoring or even pretending to consent to harassment, worried that reporting the behavior would lead to losing their jobs. A broader search included important work by Krieger et al. , which demonstrated high prevalence of sexual harassment among low income workers, and places it in a continuum of harassment and other structural sources of oppression that can result in health problems like high blood pressure. Because of this lack of information about the problem and strategies for prevention and intervention, community and academic partners collaborated to implement a communityacademic participatory research study to investigate the experience of sexual harassment among indigenous Mexican farmworker women in the Willamette Valley of Oregon.2 --- Methods --- Community-based participatory research The study methods follow the goals of community-based participatory research, where the research team includes community members who are integral to all aspects of the research process. Community and organization members and researchers engaged in a process that was designed to ensure shared decision-making and mutual ownership of the research procedures and results . Indigenous farmworkers were already permanent members of partner organizations, working as community educators. They served an essential role in this study by participating in all planning activities of the study and by moderating focus groups. They also participated in data analysis and interpretation. --- --- Data Collection Interested women were invited to participate in focus group interviews that were held at non-workplace sites in the area, and occurred in the evening or other non-working hours. Confidentiality was protected. When language was not a match for participants and the facilitator, interpreters in additional indigenous languages joined the group. Interpreters and group leaders were long-term members of the study team, and many were indigenous former farmworkers themselves. Each focus group was audio-recorded and transcribed into Spanish from the indigenous language if the focus group discussion was not in Spanish. Then, all Spanish transcripts were translated into English. Participants received $25 grocery cards for participation. Institutional Review Board approval was given by Oregon Health & Science University, University of Pennsylvania, and Johns Hopkins University. --- Measures Table 1 shows questions moderators used to lead focus groups, corresponding to the aims of the research. --- Analysis The analysis used techniques of open and theoretical coding from grounded theory analysis. The group of collaborators reviewed all transcripts individually, and through multiple discussions the group developed themes and interpretations in a collaborative and iterative effort. Moderators did not regularly poll participants in answering questions, since group members were promised anonymity and were thus not necessarily identified in transcripts. Instead, moderators encouraged open discussion in response to interview questions. As a result, it was not possible to code individual responses and produce an exact number of responses to questions. --- Results There were seven focus groups; five were held in 2006, and 2 in 2008. 49 women participated in the 2006 groups, 38 indigenous and 11 "Latina." There were 10 participants in the 2008 groups, 7 indigenous and 3 "Latina." Table 2 shows details of group composition, with the languages spoken. Of the 49 participants, 20 women required translation from an indigenous language to Spanish. --- Themes --- Knowledge about workplace policies relating to sexual harassment Now, if we get hurt, cut ourselves, we have to let somebody know so they can dress the wound and not get blood in the fruit. We have to report that. But it doesn't say anything about sexual harassment. No. All the women reported receiving some kind of supervision in the workplace, usually by Spanish-speaking foremen. Foremen were usually former farmworkers who were trained to oversee workers directly. At work, the women usually reported to a foreman, although in the absence of such a person they would report directly to the supervisor of the field or workplace. Women reported receiving training about safety and other workplace policies, and many spoke about the workplace's greater concern for the farm product than for their own welfare. The trainings were conducted in Spanish; none occurred in any indigenous language. Workplace policies: sexual harassment -Where I work, [sexual harassment] is never explained. The only rules we hear have to do with the chores that we have to do but they have never explained anything about sexual harassment. Women in the focus groups sometimes reported sexual harassment training, but most of the women in the groups had not heard of any training. One woman reported an anonymous reporting system using papers put into a closed box, where women could name harassers. Another woman saw a video in Spanish explaining company policy about reporting of sexual harassment. Others reported bosses encouraging them to report sexual harassment to them directly, not to foremen, perhaps because they viewed foremen as potential offenders. --- Experience and witnessing of sexual harassment of indigenous women farmworkers [The foreman] does not treat us all the same; he notices a few younger women and sometimes may force them or grab them. As I said, it hasn't happened to me but I've seen it happening. Those who give into it get an easier job, and I get a harder one because I don't give into it. Many women had seen sexual harassment, and a common manifestation was that of supervisors picking a favorite woman and giving her lighter work. Women expressed feelings of resentment and anger, both towards the supervisors and to the women who "play along" . The women reported feeling "bad" , and that it was unfair that "we have to do the rest" . While focus group participants spoke with frustration about the women who "really don't work very hard" and "like to chat with the men" , they also spoke about women who "for fear of losing their jobs participate and pay attention and chat using dirty words" . Fear and intimidation surrounded sexual harassment that women officially complained about: "...when it comes to supporting each other ...they are afraid" . Despite this sometimes the women provided glimpses of hope. Here, one described support from her manager despite death threats she had received: What happens is that the new manager is a good person; he did not have to believe us because he saw it and kept it confidential... So he knew how to handle it because he watched from a distance and that man had no way of denying it. Later they said they were going to kill us once they were to find out who said it. I said it did not matter and I was going to speak up anyway that he was going to hit her. I told her she had to talk to the manager because you can't keep silent. Backed by a manager who responded promptly and dealt with harassment appropriately, the woman who witnessed harassment spoke up and encouraged another woman to report abuse. Other women identified single women with children as a particular target for sexual harassment. While not a consistent finding throughout all of the groups, a few women spoke about single mothers' vulnerability because of their need to stay at a job despite harassment and aggression they experience there. In the group of Triqui women, moderators asked whether or not the women had experienced sexual harassment and received resounding "No's". When moderators asked for women's specific experiences, the women spoke more: Moderator: Let's say that men are saying vulgarities. What do you think about that? -Sometimes we feel that we can't feel comfortable at work because they are talking about us. -All we tell our husbands is that those people sure use dirty language and continue with our work so we don't get behind. -They talk like that, but we keep working. --- [...] -What we hear they are telling us is that we are ugly. -No. We don't know any. -No. When the moderators used the term "sexual harassment," women did not seem to know what it was. When other terms like "vulgarities" were used, Triqui women could identify examples. --- Indirect effects of sexual harassment Voice 4: Yes. They say something like wanting to get better hours and going with the foreman. I only hear people talking. I don't know how that works. One can talk with more coworkers. People can help. Most women reported that they had experienced or witnessed sexual harassment. Women reported damage to goodwill in the workplace community even when they had no direct experience of harassment. As one woman complained, "The hard [work] is for those other ones who are ugly" . Then he started doing it to me too, and I went to tell the manager and he wanted to fix things in private [...] I never found out because I was pregnant and I left [...] they did not listen to us. They did not do anything. And I think that man is still there. --- Harassment: direct experience Moderator: [...] So at the end you had to leave so that you would not have to put up with the situation. Voice 12: Yes. I left. Women described situations of unwanted sexual advances and even menacing behavior as seen in the passages above. Some women recounted stories of assertive behavior that relieved the abuse, like the woman in the first quotation captioned above. More frequent, however, was the kind of story in the second quotation, where women who experienced harassment often faced disbelief and inaction. Leaving the worksite was a common response. --- Aim 3: Attitudes about increased risk of sexual harassment of indigenous Mexican women farmworkers Moderator: Do you believe it is worse for indigenous women? -Yes. -[...] Because it's like I tell you. Because of how they see us, because we don't speak Spanish well, or they may think that they can say things to us and we are not going to understand and then it [inaudible] easier. [...] -So, [inaudible] obey or sometimes just so they can work, right? For the money or out of necessity, I think. I don't see it but I have heard. -We see it. -Sometimes because we can't speak Spanish well, like she says, or we just arrived and don't know. -Or because we don't know English we can't speak, right? -No other way. I have to obey and if they fire me well they fire me [...] In every group, the consensus held that indigenous women who did not speak Spanish were more vulnerable to sexual harassment partly because of language. Women also mentioned that indigenous women are less educated, leading to vulnerability: "They don't know how to read and write and barely recognize money" . Participants linked sexual harassment to threats of losing, or needing to leave, farm jobs. They often referred to the double burden of language isolation and a desperate need to work. In these data, the two contributed to indigenous farmworker women's sense of increased vulnerability to sexual harassment. --- Discussion In these focus groups, farmworker women reported widespread awareness of sexual harassment behaviors that they might not label as "sexual harassment." They did, however, recognize the behaviors and saw the deleterious effects of harassment on their lives. When women experienced harassment directly, it caused distress and disruption. When they witnessed it, it could lead to workplace conflicts. In general, women reported that these experiences made the workplace feel unsafe and unfair. While there was evidence that some workplaces provided sexual harassment training, many women reported that they had no idea how to deal with sexual harassment at work and the company did not inform them what to do. Women in these groups said that the vegetable or fruit product was more important than their own safety and welfare at the workplace. Women felt discouraged from speaking up because they did not know with whom to speak or report regarding sexual harassment, or did not think they would be believed. Language isolation and poverty were consistent themes. In the farm work setting, women who did not speak Spanish or English were more vulnerable to harassment, sexual and otherwise. Poverty was identified as a key reason why women tolerated sexual harassment in order to better their situation at work. Farmworkers' social networks and traditional societal norms may have also isolated women who experienced harassment. These data demonstrate that social and economic pressures interacted with social networks to amplify the impact of sexual harassment in the agricultural workplace. --- Limitations Focus groups by their very nature are designed to solicit general attitudes and opinions of participants, but in this study do not provide quantitative information about these attitudes and opinions. This is as expected for focus group studies. Rigor in this research is assured by transparency of analysis and confirmation of findings with community partners, and review of developing codes with partners and qualitative experts within the advisory group. In addition, purposive sampling as employed here ensured that the target community is sampled. Because this is not probability sampling, results cannot be generalized to the entire community of indigenous farmworker women either locally or nationally. New contribution to the literature-This research demonstrates the strengths of community-based participatory research to illuminate a problem facing a vulnerable and hard-to-reach community. In particular, it shows the importance of using the right questions when querying women about sexual harassment. Women in this study knew what sexual harassment was, but often did not use these words to describe it. This is essential information for planning services and future research on this topic with this population. This study has implications for practice and policy. Clinicians and service providers need to be aware of and plan for communication issues arising from language, and social dynamics in the workplace that limit conversation about sexual harassment among indigenous farmworkers. Policy should be implemented to create provisions for requiring appropriate training about sexual harassment for farmworkers in their own languages, with sensitivity towards women's fears of reporting or even acknowledging harassment. Workplace protections against sexual harassment and abuse must be monitored by law enforcement, and intervention for complaints should be swift and consistent. --- Conclusion To our knowledge, this study is first study of indigenous Mexican farmworker women's experiences of workplace sexual harassment and sexual assault. It demonstrates that women are negatively affected by sexual harassment and assault, even if it does not happen directly to them. Indigenous women and single women with children may be particularly vulnerable. In these interviews, women said repeatedly that they want and need to work. This study and future studies will provide evidence to inform the creation of interventions and services to benefit indigenous women farmworkers, but also the entire farmworker community. As community partners reach out to involve all partners in the farmworker community, including farm owners, health service providers, farmworker advocates and farmworkers themselves, their continuing partnership will contribute to preventing future sexual harassment and assault.

In order to examine the experiences of sexual harassment and sexual assault among indigenous and non-indigenous Mexican immigrant farmworkers in Oregon's Willamette Valley, a community-academic participatory research partnership initiated a study, which included focus groups, conducted and analyzed by skilled practitioners and researchers. The themes that emerged from the focus groups included direct and indirect effects of sexual harassment and sexual assault on women and risk factors associated with the farmworker workplace environment, and the increased vulnerability of non-Spanish-speaking indigenous women due to low social status, poverty, cultural and linguistic issues, and isolation. Recommendations for prevention and improved services for vulnerable women will be discussed as well as limitations and future research directions.

Introduction Basque is a low resourced language, spoken by 28.4% and understood by 44.8% of the population of the Basque Country . Thanks to its official status, it is a language with presence in the regional public administration, education system and in some news media. Thus, in EiTB , the Basque public radio and television broadcaster, it is possible to find radio and television channels in which all content is entirely broadcasted in Basque. Furthermore, there are other independent media such as Berria , Argia and HamaikaTB , in which Basque is the vehicular language. Still, the presence of Basque in traditional television and news media remains quite low, particularly when compared with those available for Spanish. In this context, the increasingly used social networks such as Twitter are of particular importance for a low resourced language such as Basque. Thus, it is possible to find a strong and active community of Basque speakers in Twitter which generates, for a low resourced language, a large amount of textual content written on Basque. Furthermore, as users create both explicit and implicit relations and communities, this data is useful to do social research using methods that may complement those traditionally used in sociology . Following this, a promising and relatively new avenue of research in social and demographic analysis combines the study of social structures created in media such as Twitter with the automatic analysis of texts via NLP. For example, previous work has focused on Twitter to study the spread of rumours , the detection of political stance or hate speech . In this paper we will take into account both social and linguistic aspects in order to perform demographic analysis by processing a large amount of tweets in Basque language. The study of demographic characteristics and social relationships will be approached by applying machine learning and modern deep-learning NLP techniques, combining social sciences with automatic text processing. More specifically, our main objective is to combine demographic inference and social analysis in order to detect young Basque Twitter users and to identify the communities that arise from their relationships or shared content. By "Basque Twitter users" we refer to those that write at least 20% of their tweets in Basque. This social and demographic analysis will be entirely based on the automatically collected tweets using NLP to convert unstructured textual information into interpretable knowledge. Current work substantially improves and extends the preliminary experimental work presented in . These improvements have led to a number of contributions. First, and taking as a starting point the Heldugazte corpus containing 6M tweets in Basque language , we devise a whole new methodology to classify users by life-stage . This new method generates a new dataset, Heldugazte-Age, containing 80K tweets semi-automatically annotated at young/adult level. Second, we explore the application of modern pre-trained large multilingual and monolingual models in order to identify young and adult users. Third, we perform a qualitative analysis comparing human performance vs life stage classifiers for classifying Basque users into the young or adult categories. Four, we use recently developed deep learning techniques for community detection, achieving better detection and visualization of the communities, as well as providing information of the relations among them. We believe that the methodology presented in this paper might be of interest for other NLP tasks and other types of social and demographic studies. Finally, we publicly distribute every resource to facilitate further research for low resourced languages such as Basque1 . The rest of the paper is structured as follows. In the next section we describe related work in computational sociolinguistics and NLP. In Section 3 we present our method to build the Heldugazte-Age annotated dataset to train classifiers for life stage detection. Section 4 presents systems used to train classifiers for life stage detection. These classifiers are evaluated in Section 5 and applied to perform social network analysis in Section 6. We finish with some concluding remarks and future work. --- Context and Related Work Social media offers the opportunity to express beliefs, sentiments or opinions in a variety of formats, including text, image, audio and video. Social media publications express conscious and/or subconscious manifestations of our social, emotional and rational condition. Previous work in sociolinguistics argues that our writing style can even be a reflection of demographic characteristics . Considering the fact that language is a social phenomenon and thanks to the ever-growing capacity in the NLP field to collect and process large-scale amounts of texts, computational sociolinguistics is becoming increasingly popular. The widespread use of Twitter has in fact benefited such approaches as it is possible now to mine large amounts of texts also for less resourced languages. Twitter is widely used in NLP for tasks such as mining opinions about specific products or topics , detecting political stance and hate speech or for basic tasks such as POS tagging , named entity recognition , normalization and language identification . NLP techniques specifically adapted for Twitter have also been used to infer demographic characteristics such as gender, age or location . Moreover, relationships, style shifting and community dynamics can also be inferred from language analysis . Of particular interest to us is the body of work performed with the objective of age or life stage detection for Twitter users. Previous works usually generate their own manually annotated datasets, covering languages such as Dutch, English or Spanish for a user range between 300 and 3000. The best performing systems are those that model life stage classification as a binary or ternary task. In relation to the study of the social relationships that are generated within the network, closer to us are those studies that have aimed to identify communities of users based on their retweets. Among these, one can find studies about political polarization , political affiliation detection or even studies about identifying communities in movements for independence . Finally, there are different research works investigating the use of low resourced languages within social networks. An investigation about Welsh speakers and Twitter, shows that speakers of this language are also active in social media . Additionally, there is another work that extracts and analyzes more than 80k tweets in Irish to do content, sentiment and network analysis . It is also interesting a study combining Welsh, Irish and Frisian to investigate the use of hashtags across 3,000 different tweets . All these works show the potentiality of Twitter to provide text data even for low resourced languages, giving the chance to find and study a huge variety of languages and cultures. 3 Heldugazte-Age: A New Dataset for Life Stage Classification In this section we propose a new methodology to semi-automatically obtain labeled data to develop life stage classifiers. The result is a new dataset for to train classifiers for Life Stage Detection, namely, the Heldugazte-Age corpus. The first step to identify online communities of young Basque speakers is collecting the data. As a starting point we will use the Basque corpus Heldugazte , which consists of 6M Basque tweets from 8,000 users, collected in May 20182 . In this collection, the last 3,200 tweets from each user were retrieved , including personal tweets and retweets. The Heldugazte corpus will be used to semi-automatically generate a labeled subset of the corpus, 80K tweets, to train classifiers to detect young/adult users. The obtained classifiers will then be applied to the rest of the Heldugazte corpus to obtain a large number of tweets written by young users. This data will be used to detect the communities between young users. In order to obtain a young/adult classifier we need some labeled data for training and evaluation. However, labeling users' tweets by life stage is a difficult task, due to two main reasons: users age hardly ever appear in the tweets metadata and, manually annotating tweets by life stage is far from being trivial. Examples below illustrate the difficulty of manually labeling individual tweets by life stage and without any additional context. "Zarauzko triatloian izena ematea lortu gabe, motibazioa falta" I have not managed to sign up for the Zarautz triathlon, I am unmotivated "A zer nolako eguraldi kaxkarra ez al du gelditu behar edo" What a bad weather, shouldn't stop or what. "5 mila euro, bideo kamera eta telefono mugikor bat eroan dituzte lapurrek" 5,000 euros, a video camera and a cell phone were taken away by the burglars. In order to overcome this problem, previous sociolinguistic work has argued that writing style could be associated to author's life stage, assuming that young people's style is more informal than that of adults . Based on these previous works, Fernandez de Landa et al. trained various classifiers to distinguish between formal and informal writing style in tweets. Every tweet for every user in the Heldugazte corpus was automatically tagged, projecting from formal/informal to young/adult classification depending on the concentration of formal/informal tweets in each user's timeline. The problem with this procedure was to objectively define a threshold for the proportion of formal/informal tweets required to classify a user as young or adult. The proposed ad-hoc solution, establishing that if 45% of the tweets were labelled informal then the timeline was to be classified as young was far from ideal. Figure 1: Ranking users by proportion of formal and informal tweets. In this paper we propose a new method to objectively and semi-automatically obtain the labeled data required to train young/adult classifiers. The procedure is illustrated in Figure 1. First, we automatically tagged the 6M tweets in the Heldugazte corpus using the formal/informal classifiers developed by Fernandez de Landa et al. . Second, we ranked users according to the proportion of informal tweets in their timeline. The top users would contain mostly informal tweets whereas the users at the bottom of the rank would consist mostly of formal tweets. Third, a manual inspection of 100 timelines established that it was feasible to manually annotate users at both ends of the ranking as young/adult. In this step it was particularly helpful to perform the annotation at user-level because a full timeline provides more contextual information to characterize a specific user. Fourth, we took the 500 users at the top of the ranking to be young users and the 500 at the bottom to be adult. As a result, following this new method we obtain a set of 1,000 users classified as adult/young based on the initial formal/informal manual categorization of 1,000 tweets provided by Fernandez de Landa et al. . The final step consisted of randomly sampling a number of tweets per user. The idea was to vary the number of tweets and topics available per user providing a sample general enough to train robust young/adult classifiers. With this objective in mind, we picked 100 random tweets per user assigned to each of them the label attributed to the user . As it is shown in Here we present the two main systems used for life stage detection: an off-the-shelf system based on linear classification and clustering features , and a deep learning approach based on learning contextual, vector-based word representations and the Transformer architecture . Previous approaches address life stage detection as a supervised text classification task . This means that classifiers will learn, from annotated data, that a given tweet is written by a young or an adult person. An example of the dataset annotations used for training can be seen in Table 2. The Heldugazte-Age dataset developed in the previous section will therefore be used to train three different text classifiers: IXA pipes multilingual BERT and, BERTeus . --- Label Content adult Taldeak mikel laboaren lanean oinarritu du bere hurrengo diskoa. The band has based their next album on the work of Mikel Laboa. adult Gure herriko ateak zabalik dituzu. The doors of our town are opened. young Buaa q follaa eun guztia eon zea ikasi ordez jolasateeenn jajaja. How lucky! You have been all day playing instead of studying hahaha. young Batzutan ze gutxi aguantatze zaituten. Sometimes I can't stand you. Table 2: Examples taken from the Heldugazte-Age dataset. --- IXA pipes IXA pipes is a set of tools with a multilingual approach across NLP tasks. This system has been successfully used in several sequence labelling tasks for various languages, including Named Entity Recognition , and Opinion Target Extraction . The general objective of IXA pipes is to provide a general semi-supervised approach that performs well across languages and tasks. This approach consists of two different components. In the first one a set of linguistically shallow features are extracted from the local context; these features are based on orthographic and ngrams and character-based information to capture multi-word patterns and prefixes and suffixes of words, which has proven useful to work with an agglutinative language such as Basque . The second, semi-supervised, component injects external knowledge previously obtained from the unsupervised induction of clustering models over large amounts of texts. This component provides several benefits. First, it generates denser document representations, given that a document is represented with respect to the number of dimensions specified in the obtained clustering model. Second, by training the clustering models on source data from different domains and text genres it is possible to inject domain-specific knowledge into the system. Finally, IXA pipes includes the possibility of including features from three types of clustering models , which helps to represent domain-specific information via complementary semantically induced knowledge. More details can be found in . For this particular work we train the IXA pipes document classifier using the same experimental setup used in . --- Transformer Models As for many other NLP tasks, current best performing systems for text classification are large pre-trained language models which allow to build rich representations of text based on contextual word embeddings. Deep learning methods in NLP represent words as continuous vectors on a low dimensional space, called word embeddings. The first approaches generated static word embeddings , namely, they provided a unique vector-based representation for a given word independently of the context in which the word occurs. This means that polysemy cannot be represented. Thus, if we consider the word 'bank', static word embedding approaches will generate only one vector representation even though such word may have different senses, namely, 'financial institution','bench', etc. In order to address this problem, contextual word embeddings were proposed. The idea is to be able to generate word representations according to the context in which the word occurs. Currently there are many approaches to generate such contextual word representations, but we will focus on those that have had a direct impact in text classification, namely, the models based on the Transformer architecture and of which BERT is perhaps the most popular example . There are several multilingual versions of these models. Thus, the multilingual version of BERT was trained for 104 languages. More recently, XLM-RoBERTa distributes a multilingual model which contains 100 languages. Both include Basque among the languages. These multilingual models perform very well in tasks involving high-resourced languages such as English or Spanish, but their performance drops when applied to low-resourced languages . Although this is still an open issue, a number of reasons can be found in the literature. First, each language has to share the quota of substrings and parameters with the rest of the languages represented in the pre-trained multilingual model. As the quota of substrings partially depends on corpus size, this means that larger languages such as English or Spanish are better represented than lower resourced languages such as Basque. Moreover, multilingual models also seem to behave better for structurally similar languages . BERTeus is a language model trained in Basque language following BERT's architecture . They show that training a monolingual Basque BERT model obtains much better results than the multilingual versions. In this paper we will compare the performance of multilingual BERT and BERTeus for life stage detection using the same hyperparameters as in Agerri et al. . --- Life Stage Detection In this section we will use the Heldugazte-Age corpus to train the classifiers previously described. The best classifier will then be applied to the whole Heldugazte dataset in order to obtain a young/adult classification of the 8K Basque tweet users contained in the corpus. Additionally, an analysis of the results is performed to better understand the quality of the semi-automatically obtained annotations. --- Experimental Results It should be noted that, in contrast to our previous work , the Heldugazte-Age corpus allows us to directly classify users as young/adult, without having to perform the formal/informal step. We perform minimal pre-processing on the tweets; we remove URLs, hashtags and usernames, leaving label-tweet pairs such as the examples shown in Table 2. This procedure has proven to be useful in previous text classification works with tweets . Table 3 reports the results obtained using the three systems described in Section 4. The high scores show that our semi-automatic method to obtain young/adult training data produces good quality annotations. Furthermore, the differences between the systems are not that large, although BERTeus is consistently the best scoring model. --- System Accuracy In order to further test the robustness of our semi-automatic method, described in Section 3, we decided to manually annotate 200 randomly selected tweets. Two human annotators labeled the 200 tweets and we calculated an agreement between the annotators of 0.78 and a Kappa score of 0.55, showing a moderate agreement between them. Furthermore, the accuracy of the two annotators are 0.795 and 0.775 respectively. When comparing these scores with the results reported in Table 3, it is clear that manually annotating young/adult at tweet level is a very difficult task. These results also show the effectiveness of our method to obtain the Heldugazte-Age corpus. In the rest of this paper we will use the BERTeus fine-tuned model to automatically annotate the whole Heldugazte corpus. It should be noted that the classifier works at tweet level . This means that once every tweet is automatically annotated, we still need to decide whether each of the user timelines corresponds to a young or adult user based on the number of individual tweets classified as young/adult. --- Labeling the Large Corpus Once the tweet classifier is ready to use, we apply the following strategy to automatically annotate the tweets in the Heldugazte corpus. First we assign a discrete young or adult label to each tweet. We then obtain a single score by averaging the number of the young/adult classified tweets of each user's timeline. The last step is to decide whether a given timeline corresponds to a young or adult user based on the score obtained from the classification of the individual tweets. In order to avoid establishing an adhoc value as a threshold, we introduce a third class for classification. In other words, a new synthetic category, "underdetermined", is created thus transforming a binary task into a ternary one. Based on the new ternary task, two thresholds are used instead of one, located at 60% and 40% of the number of tweets annotated as young in each timeline. Thus, if the proportion of labels or the average probability is over 60%, the user will be defined as a young user. On the other hand, if those values are lower than 40%, the timeline will be considered to be from an adult user. Finally, if the value is between 40% and 60% we will consider the timeline to be "underdetermined", meaning that we do not have enough evidence to decide the life stage of the user. Adding the underdetermined class has the benefit of avoiding to commit ourselves to classify difficult cases as young/adult. We are also interested in comparing the distribution of young/adult users obtained using the described procedure with those that are obtained using our previous method . As a reminder, in our previous work each tweet is classified as formal/informal and then, based on the number of informal tweets we decide whether the user is young or adult. However, for a fair comparison we will adapt it to use two thresholds and three classes, as it has been described above. Adult Underdetermined Young Previous system 5,213 911 962 New system 4,472 980 1,635 Table 4 shows the number of timelines classified as young/adult or underdetermined using our new and old methods. It can be seen that the main difference corresponds to the quantity of young users obtained by each of the methods. In the next subsection we further look into this issue. --- Comparison of Methods In this section we look at those variations in the automatic annotations assigned by the old method with respect to the one presented in this paper. Table 5 shows the differences of classifying the timelines using the old with respect to the the new method . After a superficial look to the variations, it can be seen that 21.79% of the labels were differently labeled from previous to new system, a substantial difference. Besides, one of the most significant variations is the increase in the amount of users labeled as young. Figure 2 Two of the three most important variations, marked with an asterisk, refer to the transfer of timelines to the young class. It is also important the transfer from adult to underdetermined. Taking a deeper look into these specific cases, we manually inspected some randomly chosen timelines to see if these transfers are actually true positives or whether they are misclassifications. The objective of this comparison was to study the transfer of classifications across categories when using the new classification method. More specifically, we analyzed a random sample of 10% of the cases for each variation. Below we can see example tweets from three different users. With respect to @user2 and @user3, they show that our new method, as opposed to the old one, actually classifies correctly their timelines. Thus, by looking at their tweets it seems that the users are indeed young, based on the writing style but also because the tweets talk about exams, an activity usually related to young people. The case of @user1 is more contentious, as it seems too difficult to establish the life stage of the user based on those examples, which is why the underdetermined classification does not seem misplaced. • Adult to underdetermined variation example, @user1: tweet 1a: A zer nolako eguraldi kaxkarra ez al du gelditu behar edo. What a bad weather, shouldn't stop or what. -tweet 1b: Gu erakusteko prest, etorri daitezela lasai eskuzabalik hartuko ditugu eta. We are ready to show it, we will wait for them with open arms. • Adult to young variation example, @user2: tweet 2a: Horrelakoekin gustua ta guzti hartzen zaio ikasteari. With this, you take pleasure in learning. tweet 2b: Buenobueno ba ikasiko dut gehio jaja ta ikusikozu gaindituko dutt jaja. Weeeell weeeell, I'll learn more haha and you'll see if I can pass the exam haha. • Underdetermined to young variation example, @user3: tweet 3a: Ze txupi txatxi no me da la nota. Awesome I don't get to pass... tweet 3b: Ai naiz rayatzen pixkat asko con la mierda de la uni. Oh I'm going crazy a little bit with university shit. Figure 2 depicts the variations in the classification from the using the old method with respect to the new one . The manual inspection performed would indicate that such variations are in fact correct. In other words, it would seem that the method introduced in Sections 4 and 4.2 to develop classifiers to automatically annotate users in the Heldugazte corpus as young/adult/underdetermined produce better results. In this section we will study the relations that appear between Basque young Twitter users. The starting point will be the retweets of messages written in Basque by the 1,635 users classified as young in the previous section. We select the retweets because they are the type of interactions between users that can show correlations better than other interactions such as mentions . Specifically, from the 418,903 retweets of the 1,635 young users, we extracted 24,837 nodes and 148,304 edges or connections. The nodes correspond to the users doing the retweets but also different users receiving them . On the other hand, the edges represent if a source user has retweeted one or more times another target user, representing the connections in the graph. Once the retweets are gathered we proceed to transform the unstructured data into a readable graph. First, we created a giant graph using the data from each user. To build the graph, two features extracted from each retweet were used: the retweeter and the user retweeted. After extracting the data, the visualization of the graph was created using the gephi program . Second, we gave the network a spatial structure by using the ForceAtlas2 algorithm , ordering the nodes according to the established relations. This algorithm displays a spatialization process, giving a readable shape to a network with the aim of transforming the network into a map. This technique simulates a physical system in order to spatialize a network. As a result of this process, those nodes that are unrelated repulse each other, while related ones will attract each other. The algorithm can turn structural proximities into visual proximities, allowing the analysis of this particular type of data based on interactions. Thus, the relations can be displayed in a graph. After creating the graph, we focused on two different aspects. First we identified the most important nodes of the network, to establish which users are the most influential. In a second step we uncovered the implicit communities of Basque users, splitting the huge graph into more readable subgroups that allowed us to infer the communities of young people. --- Basque influencers among young users The most retweeted users of the graph can reveal important characteristics of the investigated sample. The most important nodes show which users are the leaders for our sample. Thus, in Table 6 we can see the top 15 most retweeted users, based on two different classifications. On the one hand, there are those users with most overall retweets . On the other hand, we have the users that have been retweeted by different young users, focusing on how many different users have retweeted these users . These two rankings illustrate which users are actually the most influential between young Basque Twitter users. By looking at the obtained rankings, we can see that at the top there are accounts related to Basque media: @berria , @argia , @naiz info , @topatu eus , @HamaikaTb , @euskaltelebista and @LeakoHitza ; and Basque journalists: @larbelaitz, @axierL, @boligorria and @iBROKI . We attribute this to the fact that those people perform important roles in the creation and distribution of Basque language content in the Web. After a manual analysis of the obtained influencers, it has to be said that only two of them are accounts related to young people: @ernaigazte and @topatu eus are both accounts related to organizations formed by young people. On the one hand, @ernaigazte is the account of the Basque nationalist left youth organization, named Ernai. On the other hand, @topatu eus is a digital media account related to young people, very related to the Basque nationalist left. The lack of influencers among young users, could be related to the characteristics of Twitter, which is mostly structured around to political issues. --- Basque speaking communities for young users Once the main network graph was created, we split it into subgroups to analyze how the sub-communities or subgroups inside each one are formed. We divided each graph into subgroups using the node2vec algorithm , which allows us to obtain consistent subgroups. The node2vec User Times retweeted @berria 8671 @argia 5646 @ernaigazte 4553 @topatu eus 4236 @enekogara 3274 @naiz info 3262 @ZuriHidalgo 2568 @AskeGunea 2561 @RealSociedadEUS 2531 @larbelaitz 2471 @ArnaldoOtegi 2188 @iBROKI 2031 @LeakoHitza 1893 @athletic eus 1818 @euskaltelebista 1744 Total RTs done to users. --- User Users retweeted @berria 998 @argia 844 @naiz info 710 @larbelaitz 585 @topatu eus 531 @ArnaldoOtegi 518 @ernaigazte 478 @enekogara 454 @HamaikaTb 442 @jpermach 427 @axierL 413 @ielortza 407 @MaddalenIriarte 404 @boligorria 398 @GureEskuDago 394 Young users retweeted accounts. Table 6: Most retweeted accounts by young users. algorithm can freely explore network neighborhoods which is useful to discover homophilic communities. Unlike modularity based algorithms , used in a previous analysis of Basque communities , node2vec gives the opportunity to choose the exact number of communities to be extracted. Besides, this algorithm can be tuned in order to give more importance to homophilia or to structural equivalence. Thus, Figure 3 shows that node2vec generates clearly distinguishable sub-communities, which in turn makes them more interpretable thereby facilitating the understanding of the existing relations between them. After splitting the graph into four communities, we had to infer the main characteristics of each subgroup. For this process we focus again on the most important nodes which are the ones used to define the community itself. Each of the subgroups displays a common characteristic, namely, all of them have a direct relation with topics or issues related to the Basque Country. Those topics are different in each of the subgroups in the graph, showing the characteristics or differences of each community. Thus, it can be seen that Basque language interactions are used to talk about various Basque current affairs and politics . Also, it can be seen that music and sports are also widely commented by young people. In other words, it seems that the main function of Twitter interaction is to spread content about politics and social issues but with a clear focus on the Basque community and language. In the following we describe the main characteristics of each of the four subgroups contained in the graph. • News : In this community, the nodes found at the top of the ranking are related to news media from the Basque Country , specific Basque journalists and also to other very active users that write about the most noteworthy news in Basque . • Nationalist left : The composition of this particular subgroup is characterized by nodes related, in different ways, with the nationalist/independentist Basque left. The nodes can refer to news media , political and social organizations and politicians from the main parties in this political movement . • Sports : In the Sports subgroup the most important nodes are actually journalists and news media specifically specialized in the sports domain. Thus, for this specific group the top accounts also refer to newspapers and television broadcasters. Other important nodes here are those related to sport teams, such as football teams and Basque ball clubs or their players, which are mostly footballers from professional teams or even well known cyclists . • Music : In the Music subgroup appear in prominent places music bands or singers which sing in Basque , although other accounts related to music seem to be also very active . Figures 3a and3b show that young Basque users generally interact with users related to social issues as well as with those related to leisure . Due to the new method applied for community detection, we are able to map the communities in a more consistent way, showing in a clear way where each community is located. The position of each community on the graph and the closeness between communities show how related the topics are between them. In this way we can see that communities related to social issues are next to each other, while the same occurs with the leisurerelated communities. The community related to politics is close to news and music, illustrating both the relation with current news and the political stance of some Basque music bands. Besides, in three of the four communities , media and journalists are referential, proving again that media is important at disseminating Basque content among young people, in spite of the main topic of the community. In this section we show that combining the community detection algorithm and the visualization of the spacial representation of the graph, humans can easily interpret the meaning and characteristics of the displayed data. Thus, any information based on user interactions could be displayed and interpreted using these techniques, helping us to transform unstructured information into knowledge. In this paper we have presented a new methodology to perform demographic analysis by processing a large amount of tweets in Basque language. We have applied machine learning and deep learning approaches to Natural Language Processing to extract structured knowledge from unstructured data. Our experimental results have shown that our new method produces good quality labeled data for training young/adult classifiers. This allows us to generates a new dataset of 80K tweets annotated at user level, namely, Heldugazte-Age. The analysis of the classifiers performance has shown that, when compared with manual annotations at tweet level, the annotations of our semi-automatically generated Heldugazte-Age dataset benefit from taking into account user-level information. Furthermore, we have experimented with modern deep learning techniques for NLP and for the detection and visualization of communities in Twitter. The use of these technologies has allowed us to get more consistent and readable results than in our previous approach , apart from a better understanding of communities and their interactions. As a result of our new methodology, we have seen that the the young Basque users can be grouped in four main communities. Furthermore, we have also seen that the most influential accounts among young users are related with Basque media, revealing the importance of this actor at disseminating content in Basque among the youngest. A general conclusion has been that Basque is mostly used in Twitter to speak about Basque-related topics, being that news, politics, sport or music. We believe that the methodology presented in this paper might be of interest for other NLP tasks and other types of social and demographic studies. Finally, we publicly distribute every resource to facilitate further research for low resourced languages such as Basque3 .

In this paper we take into account both social and linguistic aspects to perform demographic analysis by processing a large amount of tweets in Basque language. The study of demographic characteristics and social relationships are approached by applying machine learning and modern deep-learning Natural Language Processing (NLP) techniques, combining social sciences with automatic text processing. More specifically, our main objective is to combine demographic inference and social analysis in order to detect young Basque Twitter users and to identify the communities that arise from their relationships or shared content. This social and demographic analysis will be entirely based on the automatically collected tweets using NLP to convert unstructured textual information into interpretable knowledge.

Background Quality of life is an important measure for evaluating the health situation of individuals and populations and is strongly influenced by the constant environmental, technological, economic, and labor relation changes. In this sense, QoL is an important indicator of individual and collective health, provided that it is measured based on its complex conceptual framework [1][2][3]. Since the establishment of the World Health Organization Quality of Life group, the World Health Organization expanded the concept of QoL, adding to it the individual's understanding of their life condition in the context of cultural and social values in response to the expectations and concerns encountered in the development of their life plan [1,2]. Several factors interfere with QoL; these are related to the multidimensionality and subjectivity of its conceptual aspects. The worker's lifestyle can be a determining factor for health risk behavior, clinical condition, physical activity , and self-rated health. These elements, in turn, potentially affect QoL [1,2,4,5]. In Brazil, the industry and commerce sector accounts for approximately 40% of the country's formal jobs [6]. Therefore, it is important to measure the QoL levels of formal workers as the results could be used as occupational health indicators, thus contributing to the understanding of health conditions, the construction/implementation of public policies, and the planning of systematized actions for providing care [1][2][3]. QoL levels in workers are affected by their interaction with the labor market, which requires high productivity but offers inadequate working conditions and low pay most of the time. Some workers still live with occupational and chronic-degenerative diseases, which directly interfere with their QoL and promote presenteeism at work [7,8]. Empirical studies on QoL do not address formal workers; most of the data available on the topic refer to specific groups, such as women, the elderly, and people with chronic diseases. Thus, no information about young and healthy workers, which represents the majority of the workforce, is available. Investigating this topic may contribute to reducing the knowledge gap regarding this group's QoL levels and add to the theoretical and epidemiological framework of public policies on occupational health. Identifying the QoL levels of workers and recognizing it as a potential indicator of occupational health may allow the development of health promotion actions, which could make the work environment more compatible with an increasing market production without causing occupational pathologies in workers [3,9,10]. Therefore, thias study aimed to identify the factors associated with QoL in young workers from a municipality in Northeast Brazil. --- Methods --- Study Design and Population This was a cross-sectional study conducted in Brazil on users of the HealthRise programa program targeting the reduction of premature death from chronic non-communicable diseases, such as diabetes and hypertension-in Vitória da Conquista, aimed at improving users' access to primary health services, qualifying care, implementing medical records, and offering special tests to patients with chronic diseases [11]. The municipality of Vitória da Conquista is located in Northeastern Brazil, with an estimated population of 338,000 inhabitants in 2019 and a territorial area exceeding 3,700,000 km 2 [12]. Most of this population lives in urban areas, which corresponds to approximately 87% [12]. The municipality is located in an important road junction that services the outflow of production between Southeastern and Northeastern Brazil and has a service-based economy with a predominance of the health, education, commerce, and civil construction sectors. --- Data Collection Data were collected at Industry Social Service , a private, non-profit entity, whose mission is to qualify the workforce and promote workers' health [13]; thus, the participants were workers assisted by SESI. The inclusion criteria for participants were workers aged 18 years or above, living in the municipality, and having attended SESI for periodic consultations with the occupational doctor. The exclusion criteria were workers coming from other municipalities or those awaiting medical evaluation before dismissal. Data were collected between August 2017 and July 2018. We sought our study sample from the 3727 workers that scheduled consultations with an occupational physician during this period. However, 339 of them did not attend their consultations, 833 awaited medical evaluation after being dismissed, 516 were from other municipalities, 25 were under 18 years old, and 744 refused to participate in the study, all of which were consequently excluded from the study. Further information about the sample can be found in the "Results" section. --- Procedure Data were collected by trained interviewers using a digital questionnaire built using the KoBoToolbox platform, on tablets. The questionnaire was adapted from the Brazilian National Health Survey 2013 [14] and included information relevant to the outcomes of the project, such as evaluating the self-care of patients with chronic diseases , evaluating users' access to health services, and measuring workers' stress levels and QoL. Additionally, the objective measurements of weight and height were collected. A properly calibrated SECA 813 ® portable digital electronic scale was used for measuring weight, with the participants barefoot and wearing light clothing. To measure height, a portable NutriVida ® stadiometer was used, with the participants barefoot and in an upright position. --- Instruments and Measurement Variables QoL was considered the outcome variable and was measured using the EUROHIS-QOL eight-item index, an instrument created by WHO together with the WHOQOL group, aimed at developing research instruments that produce health indicators through an economic approach and have a possible application in different countries to facilitate comparison between Brazilian data and data from other countries [1,2]. The EUROHIS-QOL eight-item index instrument was validated and translated into Brazilian Portuguese. It consists of eight questions based on the four WHOQOL-BREF domains: physical health, psychological health, social relationships, and environment, with each item rated on a five-point Likert-type scale ranging from 1 to 5. The total QoL score is obtained by adding the scores on all items, and ranges from 8 to 40, with higher scores indicating higher QoL [1,2]. QoL was treated as a continuous variable and categorized into tertiles. Demographic, socioeconomic, behavioral, and clinical variables were considered explanatory variables. The Brazil Economic Classification Criterion of the Brazilian Association of Research Companies , which came into force from 2015, was used with an update of the class distribution in 2016 [15]. The demographic and socioeconomic variables of interest were gender, age, socioeconomic class, marital status, and work shift. Only two categories were considered for marital status: living or not living with a partner. Work shift was also divided into two levels: those working exclusively during daytime and other work shift modalities. The behavioral variables considered in this study were diet, smoking, alcohol consumption, and PA. Healthy eating was defined as the consumption of greens, vegetables, fruits, and fruit juices, with at least one portion of fruit or fruit juice and two servings of greens and vegetables at least five times a week [14,16]. Workers, who used tobacco in any quantity, even sporadically, were considered smokers . Risky alcohol consumption was described as consuming four or more doses of alcohol for women and five or more doses for men on the same occasion within the last 30 days [14]. PA was assessed using the International Physical Activity Questionnaire , and participants who engaged in more than 150 min of PA per week were considered to be physically active [17,18]. The clinical variables included were self-rated health , and nutritional status, which was classified by body mass index and grouped into two categories: non-obese workers and obese workers [19][20][21][22]. --- Statistical Analysis The descriptive analysis of the study variables was initially performed. Continuous variables were represented by means, and categorical variables were represented in simple frequencies and percentages. The homogeneity of variance of means was evaluated by Levene's test. ANOVA or Brown-Forsythe tests were used to determine the differences between means, and a Tukey's HSD test was used to show the differences. Two-way ANOVA was used to compare the mean QoL in men and women, taking into account socioeconomic and clinical characteristics. Cumulative odds ordinal logistic regression stepwise format with proportional odds was used to determine the effect of risky behaviors, clinical conditions, and demographic and socioeconomic determinants on QoL levels. A significance level of 20% was used in the model for the explanatory variables. The proportional odds were evaluated by a total likelihood ratio test, comparing the fitted models with models with variable localization parameters. The adjustment quality deviation test indicated that the model had a good fit and significantly predicted the dependent variable. Logistic regression produced estimates, which were calculated by points and intervals with a 95% confidence and a significance level of 5%. Statistical analyses were performed using IBM SPSS Statistics software version 27 . --- Ethical Considerations The research was approved by the Research Ethics Committee of the Federal University of Bahia/Multidisciplinary Institute in Health-Anísio Teixeira Campus, according to CAEE number 62259116.0.0000.5556. All the participants signed an informed consent form. --- Results The mean age of the 1270 workers who participated in the study was 33 years , and most of them were men . Of them, 49.5% belonged to social class C and 62.2% were married or lived with a partner. The predominant work shift was the daytime shift . A total of 86.6% reported having good or very good health. The prevalence of obesity was 14.8%, while that of unhealthy eating was 56%. Tobacco use was reported by only 8.4%, while risky alcohol consumption was reported by 28.7%. A total of 62.3% of the workers engaged in PA daily . The bivariate analysis of socioeconomic and clinical conditions showed a statistically significant difference in the QoL between men and women; men had a higher mean QoL than women . The groups with higher mean QoL were the ≤29 years and ≥50 years age groups with QoL scores of 31.0 and 31.2, respectively, followed by social classes A and B with a score of 31.2, and workers who reported very good health with a score of 33.1 . The bivariate analysis of QoL, including risky behavior, habits, and lifestyle, showed statistically significant differences with higher mean QoL for individuals classified as non-obese , non-smoker , and physically active . However, individuals who practiced risky drinking had a higher mean QoL level than those who did not . The comparison between the QoL in men and women using two-way ANOVA showed statistically significant differences for all demographic, socioeconomic, behavioral, and clinical variables . The ordinal logistic regression analysis showed that QoL was likely to be 30% higher in individuals who were physically active , and even after the odds ratio was adjusted for gender, age group, marital status, socioeconomic class, self-rated health, nutritional status, and risky alcohol consumption, it remained stable and statistically significant for active individuals . Males had a higher odds ratio value for QoL than females in all logistic regression models. It is worth noting that after adjusting the model by age group, marital status, and socioeconomic class , men were twice as likely to have a higher QoL as women were . Workers who reported having very good or good health were approximately six and seven times more likely to have a higher QoL, respectively, when compared to workers who reported having regular, poor, or very poor health . Model OR: odds ratio; CI: confidence interval; * These criteria evaluate individuals' socioeconomic level through a household assessment. Scores range from 0 to 100 points, with higher scores representing a higher economic stratum: A , B1 , B2 , C1 , C2 , and D/E . --- Discussion The main factors associated with QoL in young workers in the present study were the male gender and regular PA practice, even in the model adjusted for sociodemographic, economic, behavioral, and clinical variables. The stable QoL level in physically active individuals, even in the analysis model, reinforces the positive association of PA with all QoL domains. The studied population of formal workers comprises young individuals, predominantly males, which is similar to other studies conducted in Brazil, since 56% of formal jobs are occupied by men [23]. Although there is a movement in government and nongovernmental institutions in the contemporary world for valuing and inserting women in the labor market, these actions cannot guarantee gender equality in the productive sector. Most workers in the present sample were married, practiced PA, and reported having good or very good health. Moreover, some studies indicate a positive relationship between PA and self-rated health. This association may be due to the benefits generated by the practice of PA, such as reduced incidence of diseases, improved self-esteem, and cognitive ability, and promotion of social contact with people with healthy habits, which may favor a better self-rating of health conditions, which is a good health indicator for the population [24,25]. Self-rated health has good reliability and validity not only as a predictor of morbidity and mortality but also for identifying the health needs of the formal worker population and, more objectively, for stratifying their clinical health conditions and behavioral attitudes [25,26]. Most workers who reported having good or very good health also reported higher QoL levels, as identified in a Brazilian population-based study [26] in which self-rated health was 74.2%, a percentage that reduces as the number of morbidities increases, leading to worse self-rated health and reduced Q O L. The study showed better QoL levels among workers who practiced risky drinking. This result may be partially explained by the fact that the workers in this study were mostly young. Another explanation could be that as alcohol consumption is often associated with moments of leisure and partying, it may subjectively be perceived as spending quality time [25,27]. Gender is an important variable when we consider the historically consolidated differences between men and women, and the same was found in the results of this study. Men had a higher quality of life and occupied more job positions than women did. Several studies on the role of gender in health present possible causes for these differences, including double working hours for women , number of children, and the difficulty of entering the labor market faced by women [10,23]. In this sense, the power relations established by a gender bias have historically built the social division of labor, based on biological aspects associated with sexist social stereotypes and cultural norms of appreciation of men, consequently limiting women to unpaid activities and jobs considered to be of little administrative and economic relevance. This trend continues today, as can be seen in the income gap by gender, such that in 2019, Brazilian women had a mean income 11.63% lower than that of men [6, 28,29]. The inclusion of women in precarious job positions with low pay has a negative impact on QoL, as income is one of the determinants of lifestyle, access to goods and services, and, consequently, health condition, which is influenced by working life [28,30]. Physically active individuals showed higher QoL levels than those who were inactive. We constructed a model in which the main variables of the database were used to better understand their relationship with the outcome of the study . The main exposure variable in the model was PA, as the available literature shows that there is a relationship between PA and QoL level. The model derived from ordinal logistic regression showed a certain stability of QoL levels in physically active individuals [5,31,32]. This is because the benefits of PA go beyond the improvement of the clinical and biological condition since it promotes social interaction, the establishment of bonds of friendship, and emotional balance, which are subjective and integral elements of the multidimensional aspect of the QoL construct [32]. Our findings support that PA is a variable that directly and indirectly positively influences all domains of QoL [1]. The results of studies show that PA induces behavioral change, which is fundamental for disease control and prevention. Thus, the incorporation of PA in daily life becomes an important therapeutic alternative, capable of improving general health conditions, which necessarily results in better QoL levels for workers [31,32]. Most workers were considered active. This may be due to the benefits of PA practice, the desire for a better body image, and the availability of public equipment . Moreover, the population mostly comprised young people [33]. However, a significant proportion of workers were inactive; this may be a reflection of technological evolution that is providing comfort, increased productivity, reduction of time requirements and work, and at the same time, less physical effort. No statistically significant differences in QoL were found in terms of age and marital status in this study. Although age is an important factor in the labor market, it was not present in the regression and was associated with increased QoL levels. These results differ from those found in the literature [10]. However, when age is analyzed separately, workers over 50 years of age had the highest QoL means, not agreeing with the findings of other studies [9,10]. This contrast may be due to the greater financial resources and professional stability of these workers [34]. The higher socioeconomic classes had higher incomes, and consequently, greater purchasing power of goods and services, stability in work relationships, and more job satisfaction, which are determining factors for physical and mental health, essential elements for QoL [15,26,35]. Their economic and social position seems to be a determining factor of QoL levels, which is consistent with the results of our study: workers of higher socioeconomic classes had better QoL in all proposed models [26,35]. Self-rated health is influenced by subjective and objective criteria, and according to previous studies, it is a good predictor of mortality, being a reflection of biological, socioeconomic, and behavioral aspects [36]. We found that individuals who reported having better health had higher QoL levels. Nutritional status and alcohol consumption, in turn, were not associated with QoL since the results did not show statistical significance [25]. The findings of this study should be interpreted considering some methodological limitations. First, the presence of acute pathologies was not assessed when data were collected, which may have influenced QoL levels. Second, in terms of the study design, as cross-sectional studies are limited to identifying associations and causal relationships cannot be established, reverse causality can occur [37]. For example, although lower levels of QoL were associated with obesity, it can be argued that the problems caused by overweight obstruct QoL. The results of this study support the use of QoL as an epidemiological indicator for the planning of health-related arrangements for workers. Because QoL is an indicator of health conditions, it can be adopted into all levels of the healthcare system, especially primary care, which is usually a worker's first point of contact with the health system, to identify general health needs and diagnose possible occupational diseases. --- Conclusions Most of the participant workers engaged in regular PA, and demographic, socioeconomic, clinical, and some behavioral factors had little influence on QoL levels. The results of this study may contribute to guiding the implementation of public policies that systematically promote PA in the daily lives of workers. It also highlights the need for further studies to investigate factors associated with QoL in women to confirm and possibly deepen the understanding of the present results. --- Funding: The HealthRise Vitória da Conquista study was funded by the Medtronic Foundation through the HealthRise program, a global initiative led by the Abt Associates and Institute for Health Metrics and Evaluation designed to provide access to chronic illness care for individuals in underserved communities. This study also was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brazil -Finance Code 001. Medtronic Foundation and CAPES had no role in the study design, collection, analysis, or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication. --- Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Multidisciplinary Health Institute, Federal University of Bahia, . Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. --- Conflicts of

Background: This study aimed to identify the factors associated with the quality of life of young workers of a Social Work of Industry Unit. Methods: This was a cross-sectional study conducted on 1270 workers. Data were collected using a digital questionnaire built on the KoBoToolbox platform that included the EUROHIS-QOL eight-item index to assess quality of life. Demographic, socioeconomic, behavioral, and clinical variables were considered explanatory. The associations were analyzed using the ordinal logistic regression model at a 5% significance level. Results: Men and women had a mean quality of life of 31.1 and 29.4, respectively. Workers that rated their health as "very good" had an odds ratio of 7.4 (95% confidence interval (CI) = 5.17-10.81), and those who rated it as "good" had an odds ratio of 2.9 (95% CI = 2.31-3.77). Both these groups of workers were more likely to have higher levels of quality of life as compared to workers with "regular", "poor", or "very poor" self-rated health. Physically active individuals were 30% more likely to have higher levels of quality of life (odds ratio = 1.3; 95% CI = 1.08-1.65). After adjusting the model by gender, age group, marital status, socioeconomic class, self-rated health, nutritional status, and risky alcohol consumption, the odds ratio of active individuals remained stable (odds ratio = 1.3; 95% CI = 1.05-1.66). Conclusions: In the present study, self-rated health, physical activity, and gender were associated with young workers' quality of life.

Introduction Since December 2019, COVID-19 has been affecting societies and countries across the globe, posing important questions for individual and social lives amidst a hygienic crisis with multifaceted implications. While the virus was rapidly spreading around the world, conspiracy theories and other forms of fake news were finding a fertile ground to spread too. From the very beginning of the pandemic, the Director General of the World Health Organization claimed that beyond the battle of the virus itself, WHO is also battling the spread of conspiracy theories regarding the virus and any misinformation that undermines the importance of urgent and necessary measures to tackle the infection rates [1]. Indeed, several explanations of the origin of the virus were circulated during the beginning of the pandemic, based on conspiracies such as that COVID-19 had been intentionally engineered as a political mode of controlling the masses, or as a hoax which had been spread out, worldwide, by 5G networks. These conspiracy theories led in their turn to further hoaxes, creating vaccination hesitancy and denial threatening to become an immense public health problem [2]. --- Moral identity and morality-as-cooperation As it was previously mentioned, morality in the era of the pandemic has been prominent and is worth of further empirical investigation. In our multi-country study, we conceptualized morality as an individual trait, as well as a cooperative group-relevant variable. To begin with, moral identity, reflecting the individualized conceptualization of morality, refers to the subjective importance of morality to someone's personal identity, typically represented as a trait-based difference between people or as a trait in the form of self-schema [3]. Moral identity is composed by a private "having" component and a public "doing" component . Internalization concerns the degree of centrality of moral traits for someone's self-concept, while symbolization has more to do with the degree of reflection of these moral traits in public action and choices taking place in certain social settings [4]. From a motivational perspective, internalization relates to seeking for maintaining self-consistency between identity and behavior. On the other hand, symbolization is motivated by goals of selfpresentation and recognition of the self in terms of a social entity that seeks for self-affirming feedback from other people [5]. To date, research mainly focuses on the internalization aspect, calling for more empirical attention to the symbolization-"doing"-aspect of moral identity [5,6]. Concluding, symbolization and internalization act as synergetic in people's moral reasoning and decision making. Additionally, Colby and Damon [7] suggested that what differentiates moral people from other people is the extent to which they perceive their own sense of morality and personal goals as united. Extending this argument, Frimer and Walker [8] proposed a reconciliation model, which relies on moral identity and describes moral identity as entailing agency-related motives and communion-related motives . Such motives are supposed to act as synergetic and not as competing and, eventually, the integration of personal and communion moral goals provide a powerful source of commitment to action. Morality-as-cooperation, in this context, is exemplified in community-based and societal challenges, where moral actions are needed. Morality-as-cooperation, indeed, identifies problems and societal challenges, which can be solved via exhibiting cooperative behavior, such as reciprocation or help towards the group [9]. In light of mutualism as a prerequisite for morality-as-cooperation behaviors [10], there are situations, which beg for cooperative behaviors and where individuals benefit each other by jointly and cooperatively working with others. Coordinating to achieve a mutual advantageous outcome requests joint coalitions and efforts to compete with threats and uncertainties [11]. --- Conspiracy theories and moral concerns During the pandemic, adherence to health protective behaviors and compliance with public health policies has become a morally expected issue with relevant moral concerns, when public health protection rules are violated [12]. Prioritizing public health and responding to public health risks also raised multiple moral dilemmas that have become the subject of extensive debates [13]. To this end, we aimed to consider morality-relevant variables in empirical investigation of the topical debate on conspiracy theories that have flourished during the pandemic and the degree to which morality variables mediate the relationship between endorsement of COVID-19 conspiracy theories and adherence to public health behaviors and support for equivalent policies. Conspiracy theories share a common theme, in the way that they rely on negative and traumatic events conceptualized and explained as intentional and plotted plans from powerful elites against populations [14]. These "hidden plans" often entail a morality component. Indeed, Leone et al. [15] posit that conspiracy theories entail a moral evaluation component, on the grounds that traumatic and frightful events originate from malevolent elites [16] and, thus, trigger a moral reasoning relevant to the actual understanding of conspiracy theories [17]. Farinelli [18] has described conspiracy theories as "morality tales based on archetypal narratives about right versus wrong, good versus evil" . This conceptualization of conspiracy theories can reflect the degree to which people, who are attentive to morality issues may also believe in conspiracy theories which, in turn, appeal to people's moral perceptions, values and judgements. Moral concerns may be relevant to people's individual and also collective well-being, referring to individual and group-foundations respectively [19]. Moral foundations concerning people's group membership relate more strongly with endorsement of conspiracy beliefs than moral foundations concerning people's individual moral traits and judgements [15]. This claim lies on the nature of conspiracy beliefs per se as a reflection of an "us-versus-them" mentality. This mentality is often anchored in events against nations, religious, and other large groups as victims of such conspiracy theories [20] that are supposed to attack the populations. Leone et al. [15] have suggested that the group perspective of conspiracy theories activates a group identity perspective when people think in terms of their binding to the group and not in terms of individual binding. However, following the aforementioned conceptualizations of moral identity and morality-as-cooperation, especially in light of the Reconciliation Model, as dimensions that can complement each other, we hypothesize that there will be a significant positive association between both morality-as-cooperation and moral identity with conspiracy beliefs about COVID-19 . The threats that are entailed in conspiracy theories are supposed to trigger a defensive way of life on the part of the people, reflected in their moral principles of self-responsibility and cooperation to overcome the impact of such threats. This explains the expected positive direction of the expected association. --- Morality, beliefs and conspiracy theories and health protective behaviors While several conspiracy-based explanations of various events can fascinate people across time and be harmless [21], most conspiracy theories relevant to public health crises can be dangerous and harmful for individual and societal well-being and resilience [22,23]. These types of conspiracy theories include misinformation and disinformation related to COVID-19 and cast doubt on the existence of the virus per se, in turn making people who endorse such conspiracy theories less likely to comply with public health measures [24]. Concerning moral foundations, health choices, and practices of protection of public health during the pandemic exemplify a choice of individuals to depend on their moral attitude to protect themselves and others. Containing pandemic-related behaviors as actions that people need to take to tackle the spread of the virus, morality is one key factor that plays an important role in determining the extent to which people will engage, or not, in such actions. Indeed, individuals are supposed to be motivated to take actions, depending on their willingness and intention to defend and protect sacred values, such as their quality of life and health which, in turn, are threatened [25]. People, who endorse conspiracy theories rarely make health-related choices recommended by authorities. Pummerer et al. [26] confirm findings according to which believing in conspiracy theories decreases governmental support and adoption of health-related containing behaviors such as physical distancing. Uscinski and Parent [27] have showed that people high in conspiracy beliefs were less likely to engage in cooperative and altruistic behaviors or donation practices. More recent findings come from a study by Imhoff and Lamberty [28] on conspiracy worldviews of COVID-19 and engagement with pandemic preventive and hygienic behaviors. In this study, conspiracy theories that described the COVID-19 pandemic in terms of a hoax, were more strongly associated with reduced containment pandemic-related behavior . This finding was observed in the United Kingdom and the United States. Further, it has been suggested that when morality is central to people's self-concept, then moral judgements of actions, such as intention to adopt hygienic measures, are translated into actual behavior [29]; here, scholars have shown that moral judgement and empathy for significant others is a key predictor of engagement with preventive pandemic-related behaviors. Additionally, Pagliaro et al. [30] have evidenced that discretional COVID-19 related behaviors are exhibited when they are prescribed by their very moral essence to foster the collective wellbeing and welfare of the community amidst the pandemic crisis. We, thus, hypothesize that conspiracy beliefs will be negatively associated with pandemic-related behaviors and policy support, but this association will be positively mediated by moral identity and morality-ascooperation . Overall, with our study we aim to show how both political and non-political underpinnings of conspiracy theories about COVID-19 can lead to public health containment behaviors and support of measures. The contribution of our study lies on identifying the associations between conspiracy beliefs and health-related behaviors and policy support via two levels of morality: individual and group-based morality, examining how such associations occur across 67 countries that differ in the way they implemented measures and contained the spread of the virus. Up until now, very few studies take both the individual and group dimension of morality at the same into account [15,31]. --- Rationale of the study In the present study, we used morality-as-cooperation to measure the cooperative component of morality that ties individuals to their living community. Additionally, we used a trait-based measure of moral identity [32] to discover the traits that compose people's unique moral identity as part of individuals' self-concept rather than their cooperation with community group members. Moral identity is conceptualized as an individual trait-based tendency to consider morality as central to individual self-concept and sense of self-consistency on moral action and personal identity [5]. Formation of moral identity is, thus, linked with individual characteristics and contexts for moral actions [33]. When individuals integrate morality-relevant values into their self-concept, then a moral self-identity arises [34]. On the other hand, moralityas-cooperation conceptualizes morality in terms of a group-focused behavior and cultural solutions to issues of cooperation and conflict that may occur in people's social life. Moralityas-cooperation is based on interactions between groups and people, characterized by mutual coordination, social exchanges or division and disputes [9,35]. Morality-as-cooperation mechanism motivates people towards altruistic and cooperative behaviors, as well as opportunities to evaluate behaviors of others [36]. To the extent that moral identity can be conceptualized as a foundation for comprehending moral agency in groups and organizations [37], we suggest that moral identity and morality-as-cooperation complement each other. Moral identity, by definition, entails an interactive, cooperative component too, since individuals with moral identity develop a sense of personal responsibility, considering themselves responsible for their actions in particular contexts. Therefore, they act either proactively or reactively and always in congruence between their judgement of the context and their actions for the benefit of themselves and others, who are present in a situation at hand [38]. --- Materials and methods --- --- Measures We used three measures of public health support [31]. We used a Spatial Limiting Distancing 5-item scale , to measure people's maintenance of spatial distance and reduction of physical contact, with items such as "During the days of the coronavirus pandemic, I have been staying at home as much as practically possible". Furthermore, we used a Physical Hygiene 5-item scale , to measure people's adoption of health-related behaviors and improvement of their physical hygiene, with items such as "During the days of the coronavirus pandemic, I have been washing my hands longer than usual". Finally, we used a Policy Support 5-item scale , to measure people's support towards implementation of health-protective policies and measure as responses to the pandemic, with items such as "During the days of the coronavirus pandemic, I have been in favor of closing all schools and universities". We used an 11-point "slider scale with three labels: 0 = "strongly disagree", 50 = "neither agree nor disagree", 100 = "strongly agree". These labels were re-coded to a scale ranged from 0 to 10 . Additionally, we used a 4-item COVID-19 Conspiracy Beliefs scale , to measure people's endorsement of conspiracy theories about the origin and the causes of the pandemic, with items such as "The coronavirus is a bioweapon engineered by scientists." As it was mentioned in the introduction, in this scale, we included not only political-related items, but also items relevant to the scientific community, interest social groups not necessarily political, as well as global economic issues. A scale from 0 to 10 was used. Engaging government, science, interest social groups, global institutions and an explanation of the virus as a hoax in measurement of conspiracy beliefs helped us avoid effects of political attitudes and ideological preferences on the associations we examine. Lastly, we measured morality using two scales: First, we used a 7-item Morality-as-Cooperation scale [9] , to measure different cooperative moral behaviors such as helping groups. Participants were instructed to think when they decide whether something is right or wrong, to rate the extent to which a number of considerations is relevant to their thinking. Example consideration was "Whether or not someone worked to unite a community". A scale from 0 to 10 was used. Next, we used a 10-item Moral Identity scale [32], to measure people's self-identification based on moral prosocial attributes. This scale consisted of two subscales; moral identity symbolization and moral identity internalization. Participants first read 9 characteristics that might describe a person, who could be themselves or it could be someone else. Next, participants had to visualize in their mind the kind of person who has these characteristics. Participants were asked to imagine how that person would think, feel, and act. When they had a clear image of what this person would be like, participants were asked to rate statements such as "I am actively involved in activities that communicate to others that I have these characteristics" and "Being someone who has these characteristics is an important part of who I am" . A scale from 0 to 10 was used. Recent research which has also used this data and originated from the same international collaboration project have used advanced machine-learning algorithms, established measurement equivalence of the moral identity internalization and symbolization sub-scales across the 67 countries included in the data set [39]. The original paper of Van Bavel et al. [31] reported a two-factor model , with acceptable internal consistency. This two-factor structure was confirmed in our subsequent machine-learning analysis [39] after correlating residuals of items 8 and 9, and 4 and 7. . --- Covariates In all analyses, we controlled for: participants' age, gender, employment status and living area ; participants' levels of national identification, measured with two items ; participants' political ideology, measured in one item, where participants had to rate their political views in a scale from 0 to 10, where 0 indicated very left-leaning views and 10 indicated very right-leaning views. The full list of scales, along with the dataset in SPSS and a CSV form can be found here: https://osf.io/y7ckt/. --- Results We initially performed a Principal Components Analysis with varimax rotation on the four items measuring Belief in Conspiracy Theories, in order to check whether all items loaded on the same factor. The results of this analysis suggested that all items were captured by one dimension in 9 out of 13 estimation algorithms hence allowing us to aggregate the measure of Belief in Conspiracy Theories into one variable , used in the subsequent modelling. Summary statistics of all variables included in the analysis are shown in Table 1. Next, we fitted two Linear Mixed Effects models [40] to predict our three dependent variables of Spatial Limiting Distancing, Physical Hygiene and Policy Support. Specifically, in model 1 we predicted our measure of Spatial Limiting Distancing by Belief in Conspiracy Theories, Moral Identity , Moral Identity and Moral-as-Cooperation, with country as a random effect. In model 2 we predicted our measure of Physical Hygiene by Belief in Conspiracy Theories, Moral Identity , Moral Identity and Moral-as-Cooperation again with country as a random effect. Finally, in model 3 we predicted our measure of Policy Support by Belief in Conspiracy Theories, Moral Identity , Moral Identity and Morality-as-Cooperation with country as a random effect, as it was previously done. For all models, all standardized parameters were obtained by fitting the model on a standardized version of the dataset and 95% Confidence Intervals and p-values were computed using the Wald approximation. Results are reported in Table 2. For model 1, multi-level modelling showed that Spatial Limiting Distancing had a significant negative association with Belief in Conspiracy Theories, while having significant positive associations with Moral Identity , Moral Identity and Morality-as-Cooperation. The intercept model was found to be at 5.97 = 60.11, p < .001). The random effects Intraclass Correlation Coefficient for the model indicated that most of the variance in the model was explained by the fixed variables, while 12% of the variance could be explained by the random-effect of country, indicating moderate differences between the 67 countries in the data. In a similar vein, for model 2, multi-level modelling showed that abidance to Physical Hygiene had a significant negative association with Belief in Conspiracy Theories, while having significant positive associations with Moral Identity , Moral Identity and Morality-as-Cooperation. The intercept for the model was found to be at 6.13 = 61.97, p < .001). The random effects Intraclass Correlation Coefficient for the model indicated that most of the variance in the model was explained by the fixed variables, while 12% of the variance could be explained by the randomeffect of country, indicating moderate differences between the 67 countries in the data. For model 3, results showed that Policy Support aimed to combat the spread of COVID-19 had a significant negative association with Belief in Conspiracy Theories and significant Building on the results obtained in the linear mixed effects models, we subsequently fitted three multi-level mediation models in order to test if Moral Identity and Morality-as-Cooperation mediated the negative relationship between Belief in Conspiracy Theories and Spatial Limiting Distancing, Physical Hygiene or Policy support, respectively. The results of these three models are reported in Tables 3, 4 and5 3 below. Next, for multilevel mediation model 2, the model with Physical Hygiene as a dependent variable was significantly different from a baseline model = 27921.80, p < .001) and GFI , AGFI and RFI suggested a satisfactory fit. As in the linear mixed effects model, the results suggest that belief in conspiracy theories is negatively 4 below. Finally, for multilevel mediation model 3, the model with Policy Support as a dependent variable was significantly different from a baseline model = 27936.14, p < .001) and GFI , AGFI and RFI suggested a satisfactory fit. As with 5 below. --- Discussion The aim of this paper was to investigate how Beliefs in Conspiracy Theories led to adoption of health-related behaviors and support of public health policies; how these associations were mediated by Moral Identity and Morality-as-Cooperation. Our findings showed that Beliefs in Conspiracy Theories reduced Policy Support of governmental measures , as well as the abidance to health-related behaviors related to the pandemic. Importantly, this relationship was mediated by Moral Identity and Morality-as-Cooperation. While evidence suggests that conspiracy theories exemplify a kind of calling cards that signal the generation of collective action and other political actions [41,42] in the era of COVID-19, it has been shown that collectively adopting health-related behaviors as a form of action, as well as policy support toward the government measures, is reduced when people endorse conspiracy theories about COVID-19 [26,28]. Furthermore, our results confirmed the expected positive association between Beliefs in Conspiracy Theories and Moral Identity and Morality-as-Cooperation [15,20]. Mediation of Moral Identity and Morality-as-Cooperation in the relationship between Beliefs in Conspiracy Theories and containment of health-related behaviors, adds to current findings which have mainly focused on endorsement of moral principles of care and fairness that increase, in turn, people's inclination to trust the science and government demonstrating higher Policy Support and adoption of prescribed health-related behaviors [30]. Specifically, we have built on and expanded these findings by simultaneously testing Moral Identity as an individual trait and Morality-as-Cooperation as a variable related to collective life, identifying potential similarities or differences in their effects on the adoption of healthrelated behaviors and Policy Support. Although the theory suggested that Morality-as-Cooperation would have a stronger effect than Moral Identity [15], we found that both Moral Identity and Morality-as -Cooperation both significantly mediated the relationship between Beliefs in Conspiracy Theories and the dependent variables. People's moral judgements constitute key antecedents of the adoption of containment behaviors and compliance with and support of political decisions to tackle the spread of the virus, but when Beliefs in Conspiracy Theories are high, it is reflected in people's distrust toward science and government, exhibiting low adoption of relevant preventive hygienic behaviors [43]. conspiracy explanations of the virus can potentially raise a collective action dilemma. While cooperation between people and their moral foundations could potentially be increased amidst the pandemic, the dilemmatic nature of the underpinnings of the virus-whether the virus is a hoax or it is a fact-seems to determine people's actual hygienic behaviors and their moral judgements toward each other and themselves, evaluating their position in this context. Similar dilemmas for collective action have been identified in other cases, such as the antibiotic resistance [44]. Our study emphasizes the importance of individual and group moral foundations that can inhibit the effect of believing in conspiracy theories on exhibiting health-related behaviors and policy support for health-protective measures. The indirect association of Beliefs in Conspiracy Theories with individual and group-based moral foundations and Policy Support, as well as containment health-related behaviors open novel interpretations of conspiracy beliefs as perceptions of facts, which are morally infused. Policy-making programs that aim to reduce the consistently negative impact of conspiracy beliefs on societal factors [45] such as adopting health-related behaviors and complying with policy measures could take into account that beliefs in conspiracy explanations of the origin and nature of the virus do not simply constitute antecedents of cognitive biases or failures, nor maladaptive behaviors based on personality traits. Instead, conspiracy beliefs as well as adoption of and support of policy measures reveal a deeper moral stance regarding what is right and/or wrong and who is to blame for the situation . Beyond the contribution of our study, we should acknowledge some limitations of our study that could offer a fruitful avenue for future research. First, it is worth mentioning that while some of our effects were not very large, they might still have a big impact [46]. Specifically, as Abelson [47] has noted, seemingly smaller effect sizes can still matter in the long or shorter run. They can have the potential, when they are cumulated through longitudinal observations or observations across multiple countries or settings, to have important implications for psychological research. They can do so by offering explanations of life outcomes such as people's public health, institutional trust and compliance, overall life quality and resilience against misinformation. Second, the study was correlational, thus it prevented us from inferring causations. Future research can inform existing findings by setting up experimental designs and testing the effects of moral traits on different forms of conspiracy theories and support for public health and other similar policies. Furthermore, speaking of a period when conspiracy theories have become widely attractive, an interesting avenue for future research could be whether people would be also inclined to endorse conspiracy theories about situations unrelated to COVID-19 events. For example, people who adopt a specific conspiracy explanation about COVID-19 as a hoax, may tend to describe other events in terms of conspiracies and hoaxes too in a context of a general conspiracy mentality, i.e. a tendency to broadly believe in conspiracy theories [48]. Because other events, unrelated to COVID-19, beg for collective actions and adoption of prescribed behaviors too with moral underpinnings, it is of interest to identify similar relations between conspiracy beliefs, moral foundations and exhibition of actual prescribed behaviors. If we accept existing claims [49], suggesting a generalization of conspiracy beliefs about one specific event to other unrelated events, one could examine whether the respective effects of believing in conspiracy beliefs are also generalized to explain other events, as well. --- Policy-making impact of our research Practitioners and policy makers can benefit from our research by acquiring information about people's moral foundations that can mediate their inclination to endorse conspiracy beliefs with a subsequent effect on adopting necessary health-related behaviors to reduce the spread of the virus. In addition to this, national leaders can obtain useful information by this study, in order to get an overview of how they can effectively formulate and tailor political messages to appeal to people's moral principles. Once people perceive political messages and decisions as moral ones, then they would be more likely to willingly adopt containing health-related behaviors and reject any conspiracy theories that explain the origin and spread of the virus. Indeed, people's individual and collective health-related choices are put under moral constraints and raise important moral questions [50], potentially depending on what moral foundations politicians appeal to and whether they achieve to make people reject the conspiracy explanations of the origin and spread of the virus. Tackling conspiracy information-known also with the popular term 'infodemic'-during the COVID-19 pandemic, constitutes another one important task for governments and politicians. From a political communication perspective, conspiracy theories about COVID-19 often debunk political messages, distort public perception of the virus and, eventually, weaken the credibility of the source of such political information and guidelines [51]. Conspiracy theories about the pandemic have also reduced the perceived effectiveness of political messages and recommendations of health containment behaviors [52]. Thus, perceived morality constitutes an important dimension for policy-makers to take into account so that to reduce any distortion of information about the pandemic, enhancing public well-being and encouraging people's engagement with health recommendations [53]. By considering the mediating role of morality-either in an individual or a group, cooperative level-in the association between beliefs in conspiracy theories about COVID-19 and the actual engagement with health containment behaviors and policy support, politicians can tailor their political messages to communicate strategically structured narratives advancing public health, appealing to individual and collective welfare ensuring, at the same time, the communication of comprehensive and accessible health information. Overall, the COVID-19 pandemic constitutes a global challenge for governments and countries, as well as for the people whose lives have been significantly affected. In a digital era, when fake news, conspiracy theories and rumors circulate and spread impressively fast, the pandemic features an additional challenge: the identification of the actual effects of such conspiracies on important dimensions such as people's moral judgements and behavioral change, factors which are crucial to determine the spread of the virus and the future of public health. Appendix A: Symbolization and Internalization Items of the Moral Identity Scale . --- Symbolization Items • I often wear clothes that identify me as having these characteristics. • The types of things I do in my spare time clearly identify me as having these characteristics. • The kinds of books and magazines that I read identify me as having these characteristics. • The fact that I have these characteristics is communicated to others by my membership in certain organizations. • I am actively involved in activities that communicate to others that I have these characteristics. --- Internalization Items • It would make me feel good to be a person who has these characteristics. • Being someone who has these characteristics is an important part of who I am. • I would be ashamed to be a person who had these characteristics. • Having these characteristics is not really important to me. • I strongly desire to have these characteristics. --- The data can be found at the webpage of the international project: https://icsmp-covid19.netlify.app/survey.html. --- --- Data curation: Theofilos Gkinopoulos. Formal analysis: Theofilos Gkinopoulos, Christian Truelsen Elbaek. Methodology: Theofilos Gkinopoulos. --- Supervision: Theofilos Gkinopoulos. Writing -original draft: Theofilos Gkinopoulos. --- Writing -review & editing: Theofilos Gkinopoulos, Christian Truelsen Elbaek, Panagiotis Mitkidis.

Believing in conspiracy theories is a major problem, especially in the face of a pandemic, as these constitute a significant obstacle to public health policies, like the use of masks and vaccination. Indeed, during the COVID-19 pandemic, several ungrounded explanations regarding the origin of the virus or the effects of vaccinations have been rising, leading to vaccination hesitancy or refusal which poses as a threat to public health. Recent studies have shown that in the core of conspiracy theories lies a moral evaluation component; one that triggers a moral reasoning which reinforces the conspiracy itself. To gain a better understanding of how conspiracy beliefs about COVID-19 affect public health containment behaviors and policy support via morality-relevant variables, we analysed comprehensive data from the International Collaboration on the Social & Moral Psychology (ICSMP) of COVID-19, consisting of 49.965 participants across 67 countries. We particularly explored the mediating role of two levels of morality: individual and group-based morality. Results show that believing in conspiracy theories reduces adoption of containment health-related behaviors and policy support of public health measures, but moral identity and morality-as-cooperation significantly mediate this relationship. This means that beliefs in conspiracy theories do not simply constitute antecedents of cognitive biases or failures, nor maladaptive behaviors based on personality traits, but are morally infused and should be dealt as such. Based on our findings, we further discuss the psychological, moral, and political implications of endorsement of conspiracy theories in the era of the pandemic.

Using a comprehensive, proven tool like the Equity Matrix proposed by Dr Shainoor Ismail and colleagues is critical to avoid exacerbating inequities with the inequitable allocation of SARS-CoV-2 vaccines. Without such a roadmap to illuminate the diverse spectrum and sources of inequities to direct ethically just interventions, our passage out of this pandemic will inevitably be slow and bumpy. NAVIGATING INEQUITIES: A ROADMAP OUT OF THE PANDEMIC The COVID-19 pandemic has necessitated physical gaps between individuals with distancing, and exposed unjust gaps of disparities between populations. Without a thoughtful roadmap for the equitable allocation of vaccines, this pandemic will lead to the exacerbation of inequities between and within countries. Like SARS-CoV-2, the affliction of the 'social gradient'-where those at lower socioeconomic positions have inferior levels of health-infects all countries regardless of level of wealth. 1 The SARS-CoV-2 virus is novel, but disparities are not. It is no longer sufficient to don a 'single vision' equity lens to examine some disparities, some of the time. A comprehensive investigation of all health inequities and the complex reasons for these inequities is urgently needed for the implementation of fair, restorative interventions. Here, we present the first validated, successfully used equity framework applied to COVID-19 . Unlike anything in the literature to date, this framework illuminates the spectrum of health inequities, navigates their sources and intersections, and helps avoid deepening disparities through ethically just policies. Such a roadmap is critical to steer global leadership towards the mitigation --- Summary box ► The COVID-19 pandemic has magnified inequities and threatens to exacerbate them with the inequitable allocation of scarce resources like vaccines. ► Using a 'single vision' equity lens to consider some inequities some of the time without understanding their sources and intersectionality leads to policies and programmes that blindly deepen disparities and further stigmatise and discriminate. ► Diverse interventions that are inclusive and respectful of all populations will be paramount to stop transmission of this virus and inequities. ► We present the first validated equity tool applied to COVID-19 that provides evidence and guides research on the full spectrum of health inequities, analyses their sources and interactions, and directs diverse and ethically just interventions appropriate for pluralistic societies. ► This Equity Matrix fills a gap in discussions on managing the COVID-19 pandemic, and is urgently needed to steer global leadership away from deepening disparities and towards equitable programmes and policies. on February 16, 2024 ► Differential disease severity has been shown to have large independent associations with chronic medical conditions and obesity . 10 ► Differential disease severity has been shown to have moderate independent associations with obesity , and haematological malignancy. 10 ► No clear evidence of an independent association of differential disease severity in pregnancy currently exists 10 ; however, the possibility has been suggested in some studies. ► 49 of data collection processes in their communities). ► Consider these populations as key groups for vaccination. ► Facilitate rides to immunisation clinics or home visits for those who are immobile to improve access to vaccine and testing for infection. ► Offer vaccination at healthcare visits for pre-existing conditions . ► Enable IPC measures † to reduce exposure. --- Place of residence ► Differential exposures in institutions exist with evidence of a high number of outbreaks in long-term care facilities , hospitals, prisons, shelters. 50 ► Outbreaks involving large numbers of reported cases have occurred in rural, and remote communities. 50 ► Differential disease severity has been shown to have large independent associations with homelessness. 10 ► Homeless populations, and those living in shelters/group homes or in overcrowded neighbourhoods or homes , as well as rural, remote and Indigenous communities have differential exposure and challenges with physical distancing and other IPC measures †, as well as potentially decreased access to healthcare. --- ► Indigenous communities have been disproportionately impacted in previous pandemics . 4 Multiple possible intersections such as: pre-existing condition, SES, education/literacy level, social capital, racialisation ► Include these populations in clinical trials of interventions . ► Consider these populations as key groups for vaccination. ► Consider standing orders in institutions, and mobile clinics in hardly reached populations to improve access to immunisation. ► Enable IPC measures † to reduce exposure. ► Consider vaccinating all eligible individuals in remote areas facilitated by community members/leaders/advocates for efficient, effective use of resources. ► Support programmes and policies aimed to assist and empower systemically marginalised populations and improve access to healthcare. --- Racialisation ► Differential disease severity has been shown to have large independent associations with some racialised populations. 10 Evidence of higher rates of COVID-19 and differential disease severity in ethno-culturally diverse neighbourhoods exists. [17][18][19] ► Racialised populations have differential access to healthcare, and may experience stigmatisation and discrimination. Lower vaccination rates have been observed in immigrant children and seniors for other vaccine-preventable diseases . 41 Visible minorities and Indigenous Canadians appear to be less willing than non-visible minorities to get an effective recommended SARS-CoV-2 vaccine. 42 ► Immigrant/refugee populations or migrant workers may have differential exposure due to international travel. ► Racialised populations are disproportionately represented in precarious jobs and workplace settings such as in the food or healthcare sectors, and often reside in multigenerational living spaces, leading to differential exposure and transmission within communities. Multiple possible intersections such as: SES, place of residence, occupation, pre-existing conditions, social capital, education/literacy level ► Include populations from diverse racial, ethnic and cultural backgrounds in clinical trials . ► Address racialised barriers to accessing healthcare and support policies that target systemic racism and protect the rights of racialised groups. ► Consider these populations as key groups for vaccination. ► Offer alternate immunisation settings such as mobile, worksite, or after-hours immunisation clinics and testing. ► Enable improved IPC measures † to reduce exposure. Gender identity/sex ► Differential disease severity has been shown to have a large independent association with male sex, 10 that may be linked to immunological sex differences or gendered differences in behaviours such as differential access to healthcare before progression of disease. ► Gendered differences in caregiver roles, gender-based violence and socioeconomic instability may result in differential direct and indirect impacts of the pandemic. --- Multiple possible intersections such as: occupation, preexisting conditions, SES, social capital, risk behaviours ► Consider gender/sex-inclusive vaccination policies. ► Address gendered barriers to accessing healthcare and vaccination programmes . ► Support sex and gender-based analyses. ► Support programmes and policies aimed to assist and empower systemically marginalised populations and improve access to healthcare. ► Enable improved IPC measures † to reduce exposure. ► Support programmes and policies aimed to assist and empower systemically marginalised populations and improve access to healthcare. Education/literacy level ► Those with lower education or literacy levels potentially have decreased access to healthcare. Lower levels of education have been associated with lower vaccination rates in all age groups for various VPDs. 44 45 ► Those with lower education or literacy levels are less likely to be able to work from home, potentially leading to differential exposure. ► International students may have differential exposure if they travel internationally, and differential access to healthcare if not insured. Multiple possible intersections such as: place of residence, occupation, pre-existing conditions, smoking, SES, racialisation ► Offer alternate immunisation and testing settings to improve access . ► Provide educational materials at appropriate literacy levels. ► Have translators available in clinics. ► Enlist multilingual family/community members to assist in communication. --- Socioeconomic status ► Differential disease severity has been shown to have large independent associations with low socioeconomic status. 10 ► Populations with lower income status and inability to pay for IPC resources, higher risk occupations with limited IPC measures †, job insecurity and inability to work from home have differential exposure. ► Lack of healthcare insurance or inability to pay for healthcare interventions may result in differential access. Vaccination rates tend to be lower in lower socioeconomic groups for various VPDs even if vaccines are publicly funded. 43 --- BMJ Global Health of the unfair burden of this pandemic on marginalised populations who systemically experience greater obstacles to health. --- EQUITY MATRIX: A VALIDATED TOOL TO HELP FIX COMPLEX ISSUES The Equity Matrix is the first peer-reviewed, published tool that comprehensively outlines how to methodologically assess and respond to critical equity issues in a timely, practical way. It is part of a broader framework that also integrates ethics, feasibility and acceptability to ensure appropriate implementation of public health recommendations. 2 The resources and time taken to develop this overall framework through 5 years of environmental scans, stakeholder consultations, surveys and literature reviews have reduced the resources and time taken to implement it. The evidence-informed equity tool supporting the overall framework has been validated for its utility and comprehensiveness, and successfully used for vaccines in recent years. Traditionally, most vaccine guidelines rigorously assess clinical factors such as efficacy and safety. While equity is increasingly considered conceptually, it is not explored systematically or transparently. The Equity Matrix has been fundamental in guidance on COVID-19 immunisation by Canada's National Immunization Technical Advisory Group , the National Advisory Committee on Immunization, 3 4 and may be valuable to other NITAGs and organisations around the world hastening to develop recommendations on COVID-19 health products based on principles of equitable access and fair allocation. [5][6][7] Application of this matrix for the development of recommendations on the use of interventions such as SARS-CoV-2 vaccines will be critical to reduce, rather than exacerbate, inequities. The pursuit of health equity has become a worldwide public health objective. 1 The Equity Matrix, though initially developed for vaccine guidance, can be applied across a broad spectrum of policy and programme interventions to pursue this objective. While complex in nature, health inequities share common features: they are systematic, avoidable and unfair. 8 They arise because of the circumstances in which people grow, live, work, age, and the systems put in place to deal with illness. 1 They can be modified through altering institutions, policies and practices that cause inequitable distribution of power and resources. 9 The Equity Matrix provides a ► No large increased risk in hospitalisation in current or former smokers has been observed to date with limited data for associations with substance use disorders. 10 However, data continues to evolve. ► These populations may have differential access to healthcare. ► Indirect impacts of the pandemic could lead to increased substance use, with increased substance-related deaths and harms. --- Multiple possible intersections such as: SES, social capital, place of residence, pre-existing conditions, education/literacy level ► Include these populations in clinical trials (following REB guidelines and OCAP principles 49 --- BMJ Global Health roadmap to navigate health inequities with interventions that uphold the ethical principle of justice with the socially just distribution of limited resources. It can be applied within and between countries and regions so that scarce initial supplies of SARS-CoV-2 vaccines are effectively distributed globally based on a gradient of risk rather than a gradient of wealth. It is also an inclusive tool to direct surveillance and research and unveil epidemiological uncertainties and evidence of yet undiscovered inequities of novel diseases like COVID-19. Table 1 reflects the Equity Matrix we populated with emerging evidence for COVID-19 in consultation with other experts and stakeholders. It was used to inform guidance for the efficient, effective and equitable allocation of SARS-CoV-2 vaccines in the context of limited initial supplies in Canada 4 as well as research priorities for clinical trials. 3 The table incorporates results of a rapid review of risk factors for severe outcomes in Organisation for Economic Co-operation and Development member countries. 10 Technical experts and policymakers, in consultation with stakeholders across sectors, can populate the Equity Matrix with evidence relevant to the country or countries for which policies are being made. The completed tool can be adapted over time to succinctly and comprehensively summarise up-to-date evidence, and guide research and surveillance where evidence is lacking. The first column summarises biological and social factors that may contribute to health inequity. This allows the identification of groups at high risk, for whom reduced access to a public health intervention may further exacerbate the inequity. For each group identified, the sources of the inequity are explored to aid in the process of reviewing interventions that could address the inequity and improve access . While experts acknowledge that inequities must be taken into account when considering government interventions, there is no scientific consensus on how to effectively take action to reduce these inequities. 11 Evidence on interventions aimed at individual determinants of health and on the impact of intersectoral action on health equity is limited. 12 13 This should not be interpreted as a lack of effect, but rather insufficient rigour in evaluations of these interventions. 13 A systematic review conducted to inform the Equity Matrix found limited evidence on the effects of interventions on hospitalisation for or mortality from vaccine-preventable diseases, and the evidence from these trials was uncertain. 14 Randomised trials of immunisation aimed at improving health equity is unethical, and measuring the outcome of health inequities related to many vaccine-preventable diseases requires long-term follow-up. Well-conducted, scientifically rigorous, ethical studies of interventions evaluating the impacts on health equity would be valuable to populate this tool. However, a paucity of such evidence should not preclude the implementation of ethically just interventions. Therefore, the third column of table 1 was completed with stakeholder and expert input on interventions that could address inequities. The Equity Matrix supports the inclusion of multisectoral interventions for structural, systemic changes needed to address inequities. However, given the urgent need for equitable health policies in this pandemic, table 1 focuses on short-term actions within the health sector with examples for immunisation specifically. The information summarised in this matrix is critical to ensure equity issues are appropriately addressed and integrated into expert health guidance, including recommendations on the optimal use of SARS-CoV-2 vaccines. --- ILLUMINATING THE SPECTRUM OF HEALTH INEQUITIES Evidence linking severe COVID-19 disease to biological risk factors such as advanced age and pre-existing conditions is well-established. 15 16 Evidence continues to emerge separately about the links between COVID-19 and factors such as ethno-racial background. [17][18][19] The Equity Matrix comprehensively illuminates an array of overlapping factors that compound health inequity. The spectrum of factors included in the first column of this matrix was adapted from the PROGRESS-Plus model of health determinants and outcomes. 20 We expanded the categories captured in PROGRESS-Plus and added important health equity factors that were not explicitly included in the existing model, such as having preexisting conditions, racialisation , age, and behaviours such as drug or alcohol use and smoking. The categorical inclusion of these factors ensures that they are unequivocally considered. These additional factors are all critically relevant to COVID-19 epidemiology and infectious diseases in general, and were among the most commonly investigated factors in a systematic review on health inequities related to vaccination. 14 The resulting 'P 2 ROGRESS And Other Factors' framework was used to guide a rapid review on COVID-19 which included studies using multivariate analyses to adjust for potential confounders. 10 Using this framework, the review illuminated independent biological risk factors and independent social risk factors where evidence exists, as well as potential risk factors for severe outcomes of COVID-19 where evidence is sparse . Disregarding potential inequities due to the absence of disaggregated data, such as in migrant workers differentially exposed to SARS-CoV-2, could undermine an effective pandemic response. Using the Equity Matrix to recognise that inequities may exist where evidence does not exist is important for ethically just interventions where people are treated with equal concern and respect. --- BMJ Global Health --- NAVIGATING THE SOURCES AND INTERSECTIONS OF HEALTH INEQUITIES Public policies should unravel inequities to avoid perpetuating them. The second column of the Equity Matrix outlines reasons why an inequity may exist, including differential access to healthcare, or, as outlined in the Quinn and Kumar framework, 21 pathways through which various factors can influence the differential exposure, susceptibility, severity and consequences of infectious diseases. The matrix also illustrates how different factors converge to produce inequities. For example, inequities due to socioeconomic status intersect across multiple factors. Populations living in deprived areas, as well as visible minorities and immigrant populations, seem to be disproportionately affected by COVID-19 with differential severity in the USA, 17 England 18 and parts of Canada where data exists. 19 The rapid review of evidence summarised in table 1 found that low income and ethno-racial background were important independent risk factors for severe outcomes of COVID-19. 10 This review also found that diabetes is an important independent risk factor for severe outcomes of COVID-19. The association between low income, ethno-racial background and risk of type 2 diabetes is well-established in many countries. [22][23][24] Though found to be independently associated with increased risk of severe COVID-19, the intersection between these factors is important to acknowledge. The negative association between racism and physical health, mental health and health risk behaviours is well-established. 25 Even in countries with universal healthcare policies, evidence on differential access to quality healthcare among racialised populations exists. 25 Such intersections are especially important to note because the direct effects of genetic or biological host factors on disproportionately higher rates of severe COVID-19 in racialised groups remain unknown, and research into the contribution of policy, healthcare system and society-level determinants to the roots of such inequities is needed. 26 Male sex has been found to be an independent risk factor for severe outcomes of COVID-19, 10 even though cases of COVID-19 appear to be equally distributed between the sexes worldwide. 27 Biological or immunological differences may contribute to this inequity, 28 but gendered differences in occupations and behaviours, such as differential access to healthcare before progression of disease, may also explain observed differences. While direct evidence on differences in severe outcomes due to male sex exists, indirect and direct impacts of the pandemic due to gendered differences in caregiver roles, gender-based violence and socioeconomic instability cannot be ignored. Disproportionate rates of infections with devastating consequences have occurred in previous epidemics among Indigenous Canadians, 29 and are being observed in the current COVID-19 pandemic among Indigenous Peoples in other countries. 30 31 Differential disease exposure and severity in these populations are not only linked to differential access to quality healthcare, but also systemic racism in society and the healthcare system, inequities in the social determinants of health such as adequate housing, 32 and an underlying history of colonialism. 33 Similarly, migrant and ethnic minority populations have been disproportionately affected by infectious diseases and epidemics in the midst of economic crises in the past due to lack of access to healthcare and poor living conditions. 34 In the current pandemic, these populations are also subject to differential exposure to SARS-CoV-2 due to disproportionate representation and precarious occupations in sectors such as food and healthcare, deemed essential and not subject to lockdowns. 35 Reduced access to paid time off and personal protective equipment or other infection prevention and control measures in congregate living and work settings perpetuates transmission of infection within communities. Recognising the intersectionality of the social determinants of health in predisposing systemically marginalised populations to inequitable health outcomes is critical to changing the course of this pandemic. A focus solely on biological explanations of disease is narrow and has been damaging to science and people. Labelling populations as 'vulnerable' without understanding the reasons for the increased risk in outcomes causes further stigmatisation and discrimination. The Equity Matrix allows for holistic reflections on biological and social inequities and their origins and intersections, rather than fragmented assessments that could perpetuate harmful misunderstandings, injustices and ineffective public health solutions. --- IDENTIFYING HOW TO CLOSE, RATHER THAN WIDEN, THE GAP OF HEALTH INEQUITIES Applying an 'equity lens' to bring inequities into focus without a vision to alleviate them is short-sighted. In order to be effective, public health actions in a pluralistic society cannot ignore diversity in disparities. Therefore, the Equity Matrix includes a third column, identifying practical interventions to reduce varied inequities. In the face of limited interventional resources such as vaccines and cost-prohibitive population-level programmes or product characteristics where vaccine effectiveness cannot support herd immunity, population-based risk assessments must extend beyond traditional biological risk factors to be successful and fair in immunisation strategies where some groups may get vaccine earlier than others. To reduce morbidity and mortality of COVID-19 and minimise disruption to society, healthcare systems and the economy, specific strategies to increase access to interventions and respectfully engage systemically marginalised populations will be required, as summarised in table 1. Pluralism is a prerequisite for success and respect for the diversity of communities and traditions is the key to open doors rather than a reason to put up walls. This has been evident in the success of Indigenous communities in Canada who, through Public health interventions should navigate inequities in order to reduce them through inclusion, rather than exacerbate them with policies of neglect. For example, real efforts should be made to include populations from a variety of socioeconomic, gendered and racialised backgrounds in clinical trials. Unfortunately, an examination of a subset of published studies investigating the effects of COVID-19 treatment found that a third of the studies did not report ethno-racial data, and Black patients were under-represented in all studies relative to the burden of disease among Black communities in which the studies took place. 37 Such exclusions in vaccine clinical trials contribute to distrust and vaccine hesitancy. 38 While there is a race to develop SARS-CoV-2 vaccines, public surveys indicate that there may not be a race to get one. [39][40][41][42] Vaccine hesitancy has been deemed one of the top ten global heath threats by the WHO in 2019 43 and could limit the success of a COVID-19 immunisation programme. In Canada, an examination of acceptability of SARS-CoV-2 vaccines, and coverage rates of existing vaccines, reveals additional issues contributing to inequities in groups already at high risk of severe COVID-19. Visible minorities and Indigenous Canadians appear to be less willing than non-visible minorities to get a SARS-CoV-2 vaccine that is effective and recommended. 42 Though significant differences in willingness to get a SARS-CoV-2 vaccine have not been observed by sex or socioeconomic status in Canada, 41 42 vaccination coverage rates in general have tended to be lower among men and those in lower socioeconomic groups, 44 45 suggesting issues related to decreased access to vaccines. Even among healthcare workers, who tend to have higher vaccine coverage rates, uptake varies by ethnicity. 46 Studies have demonstrated protection against infection with the appropriate use of PPE in healthcare workers. 47 48 Unfortunately, differential access to, and training in the use of PPE and other IPC measures intensifies inequities. Key reasons for vaccine hesitancy include complacency, inconvenience in accessing vaccines, and lack of confidence. 43 Population differences in access to and trust in vaccines and healthcare, as well as intent to get a safe, effective SARS-CoV-2 vaccine, should influence the direction of effective interventions, as outlined in the Equity Matrix. For example, offering publicly funded vaccines in key high-risk groups, ensuring vaccines are safe and effective in all high-risk groups, and providing mobile immunisation clinics with recall and reminder systems could increase access to vaccines for many populations. Engaging social influencers and leaders of cultural and faith-based groups with community-driven efforts for coordinated public health approaches and immunisation programme planning, and providing culturally sensitive educational materials in appropriate languages, literacy levels and media channels could combat misinformation and mistrust about vaccination in general, and SARS-CoV-2 vaccines specifically. Receiving a recommendation from a healthcare provider is linked to increased acceptance for vaccination in general, 2 and trust in doctors is a key factor in willingness to get a SARS-CoV-2 vaccine. 41 Empowering healthcare providers and the public with transparent evidence-informed recommendations based on systematic analyses of critical equity issues, in addition to factors such as the safety and efficacy of interventions, is vital. Diverse strategies will be required for diverse inequities . --- CONCLUSIONS Navigating health inequities using this roadmap can assist in the development of evidence-informed, pluralistic, socially just, effective policies that respect diversity rather than deepen disparities. Our systematic framework examining the spectrum of inequities and their sources enables policy makers around the world to direct public policies toward interventions and programmes that will reduce avoidable, unjust, unnecessary inequities in this pandemic and beyond. The Equity Matrix also provides an inclusive tool to direct surveillance and research in order to unveil epidemiological uncertainties of novel diseases like COVID-19. Data disaggregation by variables such as racialised groups is critical to avoid assumptions or the exclusion of populations from the COVID-19 response. Without seeing these data, we turn a blind eye to inequities. While our matrix for COVID-19 may be limited by evidence available and will require input and experience of experts and stakeholders, its utility is enhanced as it draws on evidence and experience from other vaccine-preventable diseases. The Equity Matrix provides a comprehensive map identifying where additional research and surveillance into inequities, their intersections and evaluation of interventions to address inequities, is required. In a journey where the desired destination is clear, but the path to success is less certain, a roadmap to navigate the unknowns and illuminate inequities that exist along the way has been absent, and is urgently needed. Our application of the Equity Matrix to COVID-19 sheds light on the ever-growing spectrum of inequities magnified by this pandemic, and leverages previous successes of its use to develop effective, comprehensive and timely guidance on track with the race for COVID-19 interventions. We urge policy makers in countries across the social gradient to use a tool like the Equity Matrix as a compass to direct ethically just policies towards interventions and programmes that will reduce disparities. Policies and research that ignore inequities, including their sources and intersections, will perpetuate them and inevitably slow our passage out of this pandemic and into a more equitable and just future. --- Data availability statement There are no data in this work.

The COVID-19 pandemic has exposed social inequities that rival biological inequities in disease exposure and severity. Merely identifying some inequities without understanding all of them can lead to harmful misrepresentations and deepening disparities. Applying an 'equity lens' to bring inequities into focus without a vision to extinguish them is short-sighted. Interventions to address inequities should be as diverse as the pluralistic populations experiencing them. We present the first validated equity framework applied to COVID-19 that sheds light on the full spectrum of health inequities, navigates their sources and intersections, and directs ethically just interventions. The Equity Matrix also provides a comprehensive map to guide surveillance and research in order to unveil epidemiological uncertainties of novel diseases like COVID-19, recognising that inequities may exist where evidence is currently insufficient. Successfully applied to vaccines in recent years, this tool has resulted in the development of clear, timely and transparent guidance with positive stakeholder feedback on its comprehensiveness, relevance and appropriateness. Informed by evidence and experience from other vaccinepreventable diseases, this Equity Matrix could be valuable to countries across the social gradient to slow the spread of SARS-CoV-2 by abating the spread of inequities. In the race to SARS-CoV-2 vaccines, this urgently needed roadmap can effectively and efficiently steer global leadership towards equitable allocation with diverse strategies for diverse inequities. Such a roadmap has been absent from discussions on managing the COVID-19 pandemic, and is critical for our passage out of it.

Introduction. Hierarchy is a central organizing principle of complex systems, manifesting itself in various forms in biological, social, and technological systems [1]. Therefore to understand complex systems, it is crucial to quantitatively describe hierarchies [2][3][4][5] and to identify their origins and benefits [6,7]. Among the various forms of hierarchy, here we are concerned with social hierarchies emerging through competition, including dominance and status hierarchies or socioeconomic stratification [8,9]. Ultimately, such hierarchy represents a ranking of individuals based on social consensus: a high ranking individual is expected to win a conflict against a low ranking one. This type of organization is present in societies ranging from insects to primates and humans [3,[10][11][12], and has been linked to resource allocation, individual health, collective decisions, and social stability [7,[13][14][15]. The prevalence of social hierarchies motivated a long history of theoretical research in statistical physics and mathematical biology [6,[16][17][18][19]. The unifying theme in explaining the emergence of hierarchies is positive reinforcement of differences known as the winner effect: initially equally ranked individuals repeatedly participate in pairwise competitions, and after an individual wins, the probability that they win later competitions increases. Conditions for hierarchies to emerge under this mechanism and their structure has been thoroughly investigated [9,11,[18][19][20]]. Yet, from experiments focusing on animal groups, we known that in addition to social reinforcement, intrinsic attributes also play a critical role in hierarchy formation [9,11]. The relative strength of the two effects depends on context; however, it was observed that they both affect hierarchies ranging from species with relatively simple social interactions, such as cichlid fish [21], to species that form highly complex societies, such as primates [13,22]. Despite the clear indication from experiments that both talent and reinforcement matter, we are lacking general theoretical understanding of their synergistic impact [23,24]. Here, we develop a rigorous model incorporating both and show that this captures a much richer landscape. For pairwise interactions, we show a trade-off between relationship stability and having more talented individuals be the high-ranked leaders. We then extend the model to open populations, where individuals enter and leave the group, and we characterize both the global and the local structure of hierarchies. Another pressing issue is to understand the response of hierarchical structure to perturba-tion, e.g., the effect of removing an individual. In particular, animal behavior experts must often make strategic decisions to remove individuals from captive societies due to health issues or in attempt to promote social stability, which sometimes lead to unanticipated reorganization of hierarchy and even societal collapse [14,25]. We show herein that if either talent or social reinforcement dominate hierarchy formation, the associated models predict smooth response and no rearrangement. It is only if their effects are equally important, that removal of an individual can lead to a non-trivial series of rank reversals. Model. Our starting point is a classic model by Bonabeau et al. that considers only social reinforcement [6]. It describes a group with N members, where the rank of each member is determined by its ability to defeat others in pairwise competitions. This ability is quantified by a score x i , where the subscript indexes the individuals. The scores are initially identical ≡ 0) and they change through two discrete-time processes. First, through positive feedback: In each time step, participants are randomly paired to compete with each other, and the winner increases its score by δ. Individual i wins against j with probability Q ij = 1 1 + exp[-β -x j )] , where β is an inverse temperature-like parameter: for large β the outcome of the fight is deterministic, for β = 0 both parties have equal chance to win. The second process is forgetting: The effect of a fight wears off exponentially, i.e., x i is reduced by µx i in each time step. Describing the full process with the deterministic equation x i = x i + δ N -1 j =i Q ij , it was shown that, depending on the relative strength of reinforcement and decay, the model supports either egalitarian or hierarchical steady state solutions [6,26]. To introduce intrinsic attributes, we offset the score of each participant in Eq. by base abilities b i and b j : Q ij = 1 1 + exp[-β + b i -x j -b j )] . Parameter b quantifies talents that are independent of social processes, yet are relevant to conflict outcomes, such as strength or intelligence. This modification, although formally simple, requires new mathematical description and leads to series of non-trivial behaviors and unanticipated emergent properties. Two individuals. To understand the consequences of intrinsic differences, it is insightful to first investigate a population of N = 2. The deterministic equation describing the steady state is 0 = -µ∆x + δ 2 1 + exp [-β] -1 , where ∆x = x 1 -x 2 and ∆b = b 1 -b 2 ≥ 0. Introducing dimensionless quantities ∆x = β∆x, ∆ b = β∆b and = µ/ leads to 0 = -∆x + 2 1 + exp -∆x -∆ b -1, meaning that the steady state is determined by the talent difference and a single parameter measuring the relative strength of decay to social reinforcement [27]. supporting the opposite order, which is "unfair", meaning that the less talented outrank the more talented. In other words, social reinforcement can outpace intrinsic abilities. We call the ∆x > 0 solution "fair" and the ∆x < 0 one "unfair", since high-ranked individuals tend to have better access to resources, more impact on collective decisions, and higher chance to foster offspring. Figure 1c shows the dependence of c on ∆ b. In general, no closed-form solution is available; limiting cases, however, can be worked out analytically: for small differences we find To answer this question, we quantify the stability of a dominant-subordinate relationship with Q, the probability that the dominant wins a conflict, Q ≈ 1/2 indicates an unstable relationship and Q ≈ 1 a well-defined relationship. Stable relationships reduce overall aggression and are positively associated with individual health [15]. Figure 1b shows that strong social reinforcement increases Q, revealing a fundamental tradeoff between stability and fairness: stable relationships require strong social reinforcement; however, strong reinforcement allows for unfair hierarchical states. Similar trade-off was experimentally observed in rankings of products in a marketplace competing for the attention of consumers: strong social reinforcement led to less accuracy in selecting the highest quality product, and to larger differences in market share [28]. Open populations. So far we focused on the relationship of two individuals, now we turn our attention to larger, changing populations. We study groups of N individuals where the talent of each individual is drawn randomly from a distribution p. We initially allow the population to reach a stable ranking. Then in each step, we remove a random individual and add a new member i to the bottom of the society, i.e., x i = 0, and again allow the population to reach a stable ranking. For simplicity we restrict our investigation to the β → ∞ limit, in which case Q ij becomes a step function. This allows us to explicitly formulate the condition for two consecutively ordered individuals to reverse ranks during the evolution of the hierarchy [29]: Eq. is satisfied, we reverse their order; we repeat this until no more pairs are reversed. b -b > ∆x ≡ δ µ , In the Supplemental Material, we derive various properties of the hierarchy through exact combinatorics and meanfield-like approximations [29]. The talent b of an individual represents an intrinsic ability or a combination of abilities that influence the outcome of a fight. In our analysis we derive a number of properties of social hierarchies for general continuous talent distribution p, including heavy-tailed distributions. Whenever specific p is necessary for calculations or simulations, we focus on the standard normal distribution. Indeed, body size, intelligence, and other relevant abilities are often normally distributed. We now systematically investigate the structure of the emergent hierarchy as a function of ∆x, the additional talent difference needed to overcome rank difference. We measure correlation between rank and talent and between rank and experience using Kendall's tau coefficient, where experience is the amount of time an individual has spent in the population. For example, τ tal = 1 indicates talent completely determines rank and τ tal = 0 indicates no correlation. Analytical calculations and simulations show that for large ∆x, rank is dominated by experience, meaning that the only way to advance in the hierarchy is if a higher ranking individual is removed; and for small ∆x rank is dominated by talent . These two limiting cases are separated by a regime where both talent and experience matter, theory predicts that the crossover point, where τ tal = τ exp = 1/2, is ∆x c ≈ 0.36 for N = 100. Experimental measurement of τ tal is challenging since it requires exact identification of the relevant talents; determining τ exp , however, is straight forward. Indeed, Tung et al. established small captive groups of macaques by introducing animals one-by-one into an enclosure and found that the Spearman's correlation between rank and experience is ρ exp = 0.61, demonstrating that some real systems are in fact near the crossover point [13]. In addition to global correlations, we also quantify local orderedness by calculating τ tal , the talent-rank correlation averaged over a sliding window of length w. Counter-intuitively, Fig. 2a shows that in the crossover regime τ tal is negative, meaning that locally rank and talent are anti-correlated. Figure 2b provides an additional aspect of this paradox situation: The expected talent b of an individual ranked kth at a random time step monotonically increases with rank; yet the probability that the th individual, the one immediately outranking the kth, is less talented than the kth is greater than 1/2. To understand the mechanism producing the local anti-correlation, first consider two consecutive individuals forming an ordered pair with respect to talent, i. individual passing the kth will necessarily pass the th too. Therefore an unordered pair will remain unordered until one of the pair is removed. This asymmetry in creating ordered and unordered pairs is responsible for the local anti-correlation. Finally, we also investigate the effect of removing an individual. We find that in the talent or experience dominated limiting cases the system's response is trivial and no reorganization happens. However, Figure 3 shows that p rr , the probability that removal of an individual induces rank reversals, is non-zero in the crossover regime. For N = 100, Understanding the response of hierarchies to external perturbation is an important issue. Particularly, removal of animals from primate groups can sometimes lead to large shifts in hierarchy and instabilities endangering the group [14,25]. Here we demonstrated that traditional models of hierarchy formation, those only considering either intrinsic abilities or social feedback, predict trivial response to removal, and that both effects have to be present simultaneously to observe rank reversals. rr for large ∆x; and are independent of N for small ∆x, i.e., τ tal = τ rr . The location of their extreme value is at the crossover of these two regimes and in case of normal talent distribution scales as ∼ 1/6 /N 1/3 . We find universal bounds τ tal ≥ -2 ln 2 + 1, p rr ≤ 0.294 . . . , for any continuous unbounded talent distribution, and these bounds are reached in the large population limit. For global talent correlation, on the other hand, we find that or if individuals do not randomly select opponents, but selectively compete with similarly ranked ones. The properties we observed for finite hierarchies may become more pronounced in the large population limit, for example, if N ∆x → ∞ but √ N ∆x → 0, global correlation τ tal converges to one, while local correlation approaches its theoretical minimum. In Table I, we provide detailed enumeration of possible behavior in the large population limit assuming ∆xN α = C, where C > 0 is constant. τ tal → 1 if √ N ∆x → 0 and τ tal → 0 if √ N ∆x → ∞. Discussion. We studied the synergistic effect of talent and social reinforcement on the structure of competitive social hierarchies, and we identified behaviors that cannot be observed if either effect dominates. Although we derived our model assuming pairwise conflicts and a winner effect, we believe that the results can be interpreted more generally: The mechanism behind both local talent-rank anti-correlation and removal-induced rank reversals is that to pass someone in rank it is not enough to be more talented, but the talent difference has to be sufficient to compensate for the advantage of being higher ranked -a In the limit of N → ∞ τ tal τ exp τ tal p rr N diff /N 1 < α 1 0 1 0 0 α = 1 1 0 τ tal p Finally, our results prompt many research questions, both experimental and theoretical. For example, local anti-correlation and removal-induced rank reversals are predictions that are testable through experiments. Future theoretical work may investigate sources of complexity not captured by our model, for example, the role of aging or slow deterioration of talent, or non-linear hierarchies, where social tiers are occupied by multiple individuals.

Hierarchy of social organization is a ubiquitous property of animal and human groups, linked to resource allocation, collective decisions, individual health, and even to social instability. Experimental evidence shows that both intrinsic abilities of individuals and social reinforcement processes impact hierarchies; existing mathematical models, however, focus on the latter. Here, we develop a rigorous model that incorporates both features and explore their synergistic effect on stability and the structure of hierarchy. For pairwise interactions, we show that there is a trade-off between relationship stability and having the most talented individuals in the highest ranks. Extending this to open societies, where individuals enter and leave the population, we show that important societal effects arise from the interaction between talent and social processes: (i) despite positive global correlation between talent and rank, paradoxically, local correlation is negative, and (ii) the removal of an individual can induce a series of rank reversals. We show that the mechanism underlying the latter is the removal of an older individual of limited talent, who nonetheless was able to suppress the rise of younger, more talented individuals.

Introduction Nowadays, the issue of masculinity is the subject of various social debates as well as of interest and discussions among representatives of various scientific disciplines, especially the cultural studies popular today. The situation is different when we refer to past epochs when these topics were rarely addressed, and even when they were addressed, the category of masculinity was generally not used explicitly. How this state of affairs was possible in a male-dominated world is interestingly explained in the Introduction to the encyclopaedia dedicated to men and masculinities, co-edited by Michael Kimmel and Amy Aronson, where it is pointed out that: There are libraries filled with volumes about men. And University course catalogues are overflowing with courses about men. But rarely are these volumes or these courses about men as men -that is, rarely are they also about gender. Instead, they are called 'history' or 'literature' or 'political science'. It's probably fair to say that if the course doesn't have the word 'women' in the title, it's more than likely about men -except that it's rare for 'masculinity' to be discussed at all 1 . Thanks to Raewyn Connell, in contemporary studies on men, masculinity is increasingly referred to in the plural form. Based on her research, R. Connell argues that there are different masculinities existing simultaneously in every culture and epoch, with one that is dominant and can be called hegemonic 2 . This type of masculinity is also the subject of this research. However, since the understanding of hegemonic masculinity according to R. Connell 3 , based on power relations , only partly corresponds to the specificity of my research, below I will not only use the above-mentioned category but also the notion of traditionally understood masculinity. In my opinion, this concept better reflects the cultural background of the constructing of the dominant model of the ideal man in the area I have analysed. Starting from the perspective of social constructionism 4 , this article focusses on the cultural conditions of structuring traditionally understood masculinity in the mid-19 th century on the example of the Polish community of West Prussia. At this point -especially for the non-Polish reader -it is worth to devote at least a few sentences to the introduction of West Prussia. The region in question was part of the Polish state in the second half of the 18 th centu-1 Introduction, in: Men and Masculinities A Social, Cultural, and Historical Encyclopedia, vol. 1: A-J, by M. Kimmel, A. Aronson, Santa Barbara -Denver -Oxford 2004, XV; See K. Kłosiński, Dekonstrukcja męskości [Deconstruction of masculinity], "Teksty drugie" 2015, 2, p. 13. 2 R.W. Connell, A Whole New World: Remaking Masculinity in the Context of the Environmental Movement, "Gender and Society" 1990, 4, p. 454. 3 R.W. Connell, J.W. Messerschmidt, Hegemonic Masculinity: Rethinking the Concept, "Gender and Society" 2005, 6, s. 832; Ł. Skoczylas, Hegemoniczna męskość i dydwidenda patriarchatu. O społecznej teorii płci kulturowej Raewyn Connell [Hegemonic masculinity and the dividend of patriarchy. On Raewyn Connell's social theory of cultural gender], "Refleksje" [Reflections] 2011, 4, p. 12-14. 4 The understanding of social constructionism was adopted after P. Berger and T. Luckmann. Among other things, these authors point out that: "Reality is socially defined. But the definitions are always embodied, that is, concrete individuals and groups of individuals serve as definers of reality. To understand the state of the socially constructed universe at any given time, or its change over time, one must understand the social organization that permits the definers to do their defining". P.L. Berger, T. Luckmann, The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Penguin Books, London 1991, p. 134. ry. Then, as a result of the partitions of Poland, from 1772 , it became an integral part of Prussia. At different times, West Prussia formed a separate province with a capital in Gdańsk . In 1920, the Treaty of Versailles returned most of the region to Poland, and the Free City of Danzig was established on the territory of Danzig itself and the surrounding area. Since 1945, the entire area is located within the borders of Poland and is now known as the Vistula Pomerania. Since I have highlighted the Polish community as the subject of my research, it is also worth noting the nationality relations that existed in the area under study. According to various estimates, in the second half of the 19 th century and the beginning of the 20 th century, the population of West Prussia was dominated by speakers of German. The Polish community was a minority . The third national group was a small Jewish population 5 . It is worth noting that due to the multicultural character of West Prussia, the nationality choices of its individual inhabitants were not always unambiguous. Even in the first half of the 19 th century, identity was determined more by religion than nationality, i.e., Catholicism was clearly dominant among the Polish population and Protestantism among the German population 6 7 . As a result, the views represented by the Polish community were influenced more by the Polish Catholic clergy and representatives of the Polish national movement than by the Prussian administration that was associated with Protestantism. My decision to limit the analysis to the Polish population was therefore guided both by the cultural and religious differences between Germans and Jews and by the influence exerted on the community under study by the Polish intelligentsia . Warszawa 1909, p. 553; J. Borzyszkowski, Inteligencja polska w Prusach Zachodnich 1848-1920 [Polish intelligentsia in West Prussia 1848-1920], Gdańsk 1986, p. 18; M. Wojciechowski, Stosunki narodowościowe w Prusach Zachodnich w początkach XX wieku [Ethnic relations in West Prussia at the beginning of the 20 th century ], in: Regiony pograniczne Europy Środkowo-Wschodniej w XVI-XX wieku. Społeczeństwo -gospodarka -polityka [Border regions of Central and Eastern Europe in the 16 th -20 th century. Society -economy -politics], eds M. Wojciechowski, R. Schattkowsky, Toruń 1996, p. 72-73. 6 Cf. K. Wajda, Świadomość narodowa chłopów polskich na ziemiach pod panowaniem pruskim: przesłanki integracji narodowej chłopów i jej przebieg [National awareness of the Polish peasantry under Prussian rule: the prerequisites for national integration of the peasantry and its process ], "Acta Universitatis Nicolai Copernici. Historia" 1997, 322, p. 141. --- Paweł Śpica --- Methods and source basis of research Conducting research on the issue of masculinity in relation to the 19 th century poses a certain challenge when we consider that this category was not at the centre of attention of the Polish society at the time, and also when the researcher's area of interest is limited to a relatively small region, perceived on a macro-historical scale as peripheral. This is the case with the aim of the present research, i.e., investigating cultural factors constructing the traditionally understood masculinity in the Polish community in West Prussia in the middle of the 19 th century. The realisation of this objective was only possible thanks to a contextual analysis of the Polish journalistic writing of the studied region published in the period in question. In the Polish environment of West Prussia, the position of men in social life, and even more so the subject of masculinity closer to our current considerations, was not the subject of wider analysis. In the area studied, the functions and upbringing of women were given far more prominence as a result of the special role which was attributed to the mother in the process of maintaining and developing Polish culture at a time when Poland was deprived of its own statehood, and because of the emerging process of women's emancipation in Europe. For the 19 th -century society of West Prussia, however, the roles and position of men in the family and society were so obvious, unambiguous and unquestionable that they were barely discussed. Yet, due to the fact that the publicists of the time often condemned the vices and shortcomings of the Polish population, we are in possession of source material on the expectations placed not only on women, but also on men, which allows us to formulate conclusions on the image of the model man of the period in question. The research results presented in this article have been developed based on the method of critical analysis of sources, characteristic for historical-educational research. In this particular case, the Polish journals of West Prussia from the mid-nineteenth century were analysed. All Polish newspapers published in the region at that time were selected to carry it out. However, their journalists very rarely wrote about men as men. Therefore, this publication uses only those press titles that directly and indirectly refer to the role of men in the family and society. Therefore, the conducted research covered the following newspapers from the mid-nineteenth century: "Biedaczek" , "Katolik Dyecezyi Chełmnińskiej" , "Nadwiślanin" and "Gospodarz" . In addition to the above-mentioned press, the journal "Pielgrzym" published since 1869 was also used, due to the references to the analysed issues clearly appearing in this newspaper. --- Cultural factors conditioning the construction of male identity in the light of an analysis of the Polish West Prussian journalism publications Preserved publications allow for the identification of several fundamental factors constructing male identity in the mid-19 th -century West Prussia, to which I would like to devote particular attention in this article, i.e., the dominant narrative on the role of the male in the family, the place of the male in social structure, the role of insurrectionary and organic work ethoses, the influence of professed faith and views proclaimed by Catholic moralists on the role of men in the family and society, as well as beliefs resulting from these views on male sexuality. In addition to these factors, it is also worth mentioning the cultural transmission of attitudes and behaviour patterns, quite often based on traditional customs and habits originating in the nobility and folk culture, as well as the importance of individual family relationships. All the above-mentioned determinants, apart from the last two, which are more or less correlated with each other, constituted the hegemonic ideal of masculinity of Poles in the studied region at that time. At the same time, some of these factors shaped different types of identity . --- Dominant narrative on the male role in the family In the 19 th -century Polish journalism of West Prussia we are confronted with a uniform picture of the role of men in the family and society, resulting from the discourse of gender polarization dominant in the region in question. This discourse apparently delineated the basic characteristics of masculinity. It was usually promoted by anonymous authors of articles published in the most influential opinion-forming Polish periodicals published in West Prussia8 . The nature of the views presented by the aforementioned authors was generally imitative, as the idea of gender polarisation appears in the views of many philosophers, including Jean-Jacques Rousseau, Immanuel Kant, Johann Gottlieb Fichte, Wilhelm von Humboldt or Joachim Heinrich Campe. This idea was based on the belief that there are natural differences between men and women, who complement each other but they are determined to perform different social roles 9 . The concept of gender polarisation was therefore strictly essentialist. In the traditional view, the tasks of a man, as a husband and father, primarily included looking after the family's material well-being, providing it with a sense of security and representing it externally 10 . Translating these tasks into the language of modern family pedagogy, it can be said that the man was required to perform, above all, the economic, protective and representative functions. As the undisputed "head of the family", he also acted as the main decision-maker. Analysing various press articles from the area of West Prussia, we can confidently conclude that in the studied area men were mainly expected Paweł Śpica to be hard working, thrifty, frugal, sober and honest 11 . These expectations were a consequence of the difficult economic situation of many Polish families in the 19 th century. The cultural image of the man of West Prussia did not differ much from the model presented both in other Polish lands and in the other provinces of the Prussian state. This state of affairs can be explained by the patriarchal narrative which prevailed throughout Europe and the provincial character of the region in question which, in the 19 th century was by no means original in terms of the development of original views on social life. Nevertheless, the dominating discourse of masculinity in West Prussia also had its specific foundations. This is because the concept of gender polarisation played an important role in the process of maintenance and development of the Polish national identity. Specific roles assigned clearly to men and women, based on the traditional division of duties, also had particular patriotic connotations. The model of gender polarisation was therefore part of the process of transmission of the Polish culture. This model not only preserved the existing social and intra-family relations, but also a specifically understood Polishness, which in the absence of Polish statehood was based primarily on the traditional family 12 . However, it is essential to note that in the conditions of West Prussia this discourse had a "noble" character, i.e., it was strongly associated with the views proclaimed by the Polish national elites, whose individual members, even if they did not have noble lineage, nevertheless adopted the noble culture as their own, superior and most significant culture. The Polish journalism publications in West Prussia strengthened conservative social relations; moreover, family relations of the time, despite the modernisation of society and the economy, most often still reflected a feudal mentality in which the position of man in the family reflected the role of the monarch in the state 13 . It is worth adding, however, that in the light of Polish publications, this role was definitely not to be despotic in character, but enlightened, based on respect for the subordinate household members such as wife and children. Although a man had the right to physically punish his children -which was in line with the widespread belief that the use of corporal punishment had a generally positive effect on upbringing 14 -moralists of the time strongly condemned the abuse of offspring and wife15 . In the Polish lands, as late as the mid-19 th century, the image of the strict parent was widely accepted. Excessive leniency and indulgence towards children were perceived as detrimental 16 . What is significant is the coherence of the narratives of the various Polish authors writing texts on social and family issues in the studied region. It resulted from the essentially monolithic structure of the Polish intelligentsia of West Prussia, dominated by the Catholic clergy17 . The dominance of the clergy among the Polish intelligentsia of the region was the result of complicated national relations, but also of the views of the local population. In the second half of the 19 th century and at the beginning of the 20 th century, clergyman was one of the few professions of the intelligentsia that allowed the preservation of Polish identity in a relatively straightforward way. Moreover, the peasant community dominant among the Polish inhabitants of West Prussia saw no need nor had financial means to educate its sons in other professions requiring more than a basic education, namely that of a priest. Meanwhile, candidates for the priesthood were often supported by scholarship assistance. The views presented by the Catholic clergy strongly influenced the entire Polish movement, including the publishers and editors of the Polish press. It is therefore not surprising that the Polish journalism of West Prussia ignored alternative views on the role of men and it expressed strong criticism of various socialist, revolutionary, emancipatory, etc., movements developing in Western Europe18 , which could contribute to upsetting the established social relations in West Prussia, and thus also lead to a redefinition of the current position and role of men. Meanwhile, the uniform, coherent, conservative "model of masculinity" created by Polish publicists was a serious factor in shaping the local population's ideas about the place of men in the world and provided the men themselves with concrete guidelines on how to structure their own identity. --- The place of men in social structure The coherence of the dominant discourse of traditionally understood masculinity in West Prussia was also rooted in the social structure of the region in question. It was a peripheral, poorly industrialised region, the largest city of which had in the Paweł Śpica mid-19 th century a permanent population of no more than 60,000 19 . As I have already mentioned, the Polish population was dominated by peasants living in rural areas who, despite internal stratification, generally held an unequivocally conservative world view. As there was no middle-income or wealthy Polish bourgeoisie in the area under study, the leading role in Polish society during the studied period was played by the nobility. This was a relatively small but influential social layer, with a conservative outlook on male-female relations similar to the one that characterised the peasants. In the 19 th and early 20 th centuries, belonging to a particular social stratum determined men's professional activity, their functions in the public space, and significantly influenced their mentality. In the rural environment, the more austere farmers, fishermen or forest workers differed in their manners from the representatives of the nobility. While the former were mostly focussed on how to make ends meet and provide for their families through hard work, the latter could devote themselves to politics, the economy, the national cause, etc. Men from the privileged strata stepped into the role of statesmen, "fathers of the nation". Their vocation, or even their national duty, resulting from the ideals of organic work, was to educate the people and try to raise them to a higher economic and cultural level 20 . Their conviction of their unique role played in the society somehow legitimised the views they proclaimed. In its "noble" version, traditional masculinity was characterised by broadly defined fatherhood, which meant taking moral responsibility for the local community. Polish men, representatives of nobility, were supposed to be guides, leaders, models of the "masculine" virtues. However, in their relations with the representatives of the peasant strata -despite the shared faith and language -a certain distance and distrust was apparent, resulting from the different lifestyle and economic position 21 . The stratifying context of the structuring of masculine identity was particularly visible in men's attitudes to education. In the mid-19 th century, concern for a good intellectual education for boys was a clearly discernible trend among the nobility, which reflected the change that had taken place in the approach of the elite to the upbringing of their offspring. For, at the beginning of the 19 th century, "noble" fathers did not yet attach such an important role to the education of their children. Many fathers belittled the education of their sons, claiming that the wealth they would inherit in the future would be far more useful to them than studying. Meanwhile, more or less from the 1840s onwards, a belief in the moral and social value of education and upbringing prevailed among the local nobility 22 . These views significantly differentiated the men of nobility from those of peasant origin, who did not see in academic education any chance of improving their situation, as they usually could not afford to send their children to secondary and higher schools. In addition, peasants often regarded their children's compulsory schooling as a burdensome necessity reducing the number of hands needed to work the land23 . --- Organic work ethos versus insurrectionary ethos The "model of masculinity" propagated by the Polish journalists of West Prussia consisted of various components. It was an ideal based both on Catholic morality inscribed in the Polish national context and on elements of Prussian culture, which somewhat differentiated the Polish inhabitants of Prussia from their compatriots from Russia or Austria. In this model, however, we do not find all the features constituting the Prussian mentality , but only those perceived as valuable from the perspective of shaping and developing the Polish identity. These qualities included above all the values constituting the Prussian work ethos, including industriousness, thrift and frugality, and stemming directly from the German bourgeois mentality and the capitalist mentality 24 . This ethos, in turn, was one of the factors creating the ideals of organic work and groundwork popular especially among Poles in the Prussian partition starting from the period of the Spring of Nations 25 . The foundations of these ideals included the above-mentioned characteristics and were a pattern particularly socially desirable in the last decades of the 19 th century and in the early 20 th century due to the economic struggle between Poles and Germans. These ideals were therefore strongly inscribed in the context of national antagonisms. Nevertheless, the promotion of this peculiarly organic work ethos also involved the creation of a rational masculinity model. Although for West Prussia this was undoubtedly a hegemonic model, the local Polish journalism used to refer also to the opposite ideal, which we can call the sentimental-emotional masculinity. When I write about the sentimental-emotional masculinity, I mean the model based on the insurrectionary ethos, characteristic especially of the Polish nobility of the Russian partition, clearly dominant from the end of the 18 th century until the fall of the January Uprising in 1863, but also present in later years, especially at the beginning of the 20 th century . The model of sentimental-emotional masculinity, contrary to rational masculinity, was based on the insurrectionary ethos and consisted of: a longing for lost independence and the former greatness of the Polish state, a sense of national injustice and pride ex-Paweł Śpica pressed in a kind of moral superiority over the partitioners, and the need to undertake armed struggle for independence. The insurrectionary ethos assumed that irrespective of previous experience [i.e., defeats -author's note] [...] the moral duty of a Pole, a duty from which one cannot evade regardless of one or another rational argumentation, is to fight for independence, to shed blood for Poland, if necessary -to give one's life for it26 . It was therefore a martyrological discourse, showing the defeats suffered in terms of ethical victory. Due to the conditions of West Prussia in practice this model never found fertile ground for growth. However, it was present in the local press. For example, in the quote from A Short Excerpt from the Heroic Deeds of Hetman Stefan Czarnecki, referring to the glory of the Polish army from the times of the First Republic and published by the "Biedaczek" magazine in 1849, it was pointed out that in the past it used to be "more glorious to fall in battle than to flee". While creating the insurrectionary ethos, specific "masculine" features were pointed out, including bravery, courage, resignation of oneself for the sake of the homeland, soldier's hardship, honour, and above all trust in God. The possession of the above-mentioned qualities was supposed to inspire pride in the eyes of other people, especially women and relatives. "Happy is the mother who gives birth to such sons"27 -it was written. Despite the martyrological motifs clearly present in the Polish journalism of West Prussia, the key component of hegemonic masculinity in the studied region remained "masculine" rationalism, which directly resulted in a down-to-earth attitude and anticipation of the far-reaching consequences of decisions and actions, especially in the areas of economics and education. This ideal was also strongly associated with the criticism of all selfishness, iniquity and immoral behaviour. They were opposed with the idea of a "man of character", i.e., capable of acting properly according to reason and will. As explained on the pages of "Farmer" in 1850: This truth, acquired by reason, constitutes the scope of man's duty in social and moral life, and this virtue, which is the realisation of duty by the will, constitutes the man's character. [...] Where there is [...] a correspondence between duty and will, between conviction and execution, there is the power of character, and the weaker the will in working towards the execution of one's conviction, the weaker the character28 . --- The influence of faith and the views proclaimed by Catholic moralists on the role of men in the family and society Irrespective of the place occupied by men in the social structure, the perception of masculinity and the role of the man in the Polish environment of West Prussia was shaped to a large extent by the Roman Catholic religion and more specifically the interpretation of the truths of faith by the Catholic moralists of the time in relation to the family, in particular to the role played by the father. As I have already mentioned, the religious context as a factor constructing male identity was extremely important in the case of the Polish community of West Prussia, as the principles transferred by the Catholic Church were more important to the Poles than the views proclaimed by the Prussian ruling elite perceived as foreign and Protestant. As the parish community was the centre of social life for the local community, the views expressed by the clergy were widely heard, which does not mean, however, that they were accepted by everyone and implemented as an everyday practice. Interesting light on the role of a man in the family and society is shed by moralistic articles published in the mid-19 th century on the pages of "Katolik Dyecezyi Chełmińskiej". They referred primarily to the ideal of a good husband and father. In the light of the texts published in "Katolik..." it can be concluded that, in the perspective of authors, a "masculine" man is above all a person characterised by chivalry, a cheerful and lively manner and a "fiery" mind. Moreover, he should possess the qualities required of a model woman, i.e., physical and mental health, strength and firmness, diligence, thrift and resourcefulness in various situations 29 . In the light of the common understanding of masculinity, the behaviour of men in the mid-19 th century, especially those recruited from the noble class, was guided by the principle of honour. Nevertheless, it was often understood in a different way than the Catholic publicists would have wished. For, the common understanding of honour quite often boiled down to... extravagance and living beyond one's means. This way of thinking was strongly criticised by the clergy, who condemned the mismanagement and profligacy often seen in the Polish environment. According to the editors of "Katolik...", this problem affected the nobility, peasants and craftsmen alike, and in each case it had negative consequences in the form of poverty, marital conflicts, etc. 30 . However, according to both Catholic and secular moralists, the greatest bane of Polish society was drunkenness, downplayed by many and associated primarily with men, which was seen as a defiance of 29 P. Śpica, Relacje małżeńskie, seksualność i wychowanie do życia w rodzinie w połowie XIX wieku w świetle wydawanego w Prusach Zachodnich "Katolika Dyecezyi Chełmińskiej" [Marital relations, sexuality and education for family life in the mid-19 th century in the light of 'Katolik Dyecezyi Chełmińskiej' published in West Prussia], in: Rodzina i dziecko w zmieniającym się świecie: perspektywa historyczna i pedagogiczna [Family and child in a changing world: historical and pedagogical perspectives], eds K. Jakubiak, R. Grzybowski, Toruń 2020, p. divine destiny, including the conjugal vocation 31 . Excessive consumption of alcohol was a common phenomenon, a kind of "ritual" accompanying various occasions, holidays, funerals, etc. Some people celebrated this "ritual" every Sunday, immediately after the Holy Mass, by heading to the local inn 32 . While drunkenness was regarded as a typically male trait, "living beyond one's means" was a fault for which the Catholic moralists condemned the entire Polish community. However, the phenomena cited above were part of a social diagnosis but were excluded from the hegemonic discourse that unequivocally assessed them as resulting from flawed attitudes 33 . As I have already mentioned, despite the fact that Catholic publicists did not use the category of "masculinity" directly, it can be concluded that they most closely linked it with the performance of the functions of husband and father, which on the one hand -in accordance with the model of gender polarisation -included maintaining and ensuring the material existence of the family 34 , and on the other: faithfulness and fulfilment of the marriage vow, associated with the attitude based on faith, love, respect and the indissolubility of marriage 35 . According to 19 th -century Catholic moralists publishing in West Prussia, masculinity therefore meant above all the maturity to build a lasting marital relationship, maturity in faith , taking responsibility for the family, the ability to go beyond the zone of one's own comfort, to forgive, to endure any inconvenience and to persevere despite various adversities. Despite such an idealistic vision, Catholic publicists also pointed to human imperfection and the need to forgive each other for their mistakes 36 . --- Beliefs about male sexuality The ideal indicated above was also associated with a strict vision of the male sexuality. In the Polish journalism of West Prussia, sexuality was not the subject of extensive analyses. It was tackled mainly by Catholic moralists, who, however, very rarely referred to it, and when they did, they wrote about it in a denunciatory way, i.e., in terms of danger, the possibility of committing a sin; their content was most often limited to listing the negative consequences of unwanted and extra-marital pregnancies . Hence, more attention was paid to the virtuous conduct of girls than of boys 37 . The Polish population of West Prussia at that time was prudish and sexual issues constituted a taboo. The views of local Poles on human sexuality, including male sexuality, were determined by the Catholic ethics. This ethics required men to shun all adultery , as well as all other external and internal impurities and cruelty to their wives. Acting in accordance with the above-mentioned qualities was at the same time identified with paternal care for the upbringing of children. The Catholic moralists thus tried to make men aware of the importance of influencing their offspring by their own example, as well as of the long-term and social consequences of the selfish behaviour of some men who focus mainly on satisfying their own needs 38 . The Catholic morality dominant in the Polish community of West Prussia forbade men to have any extramarital relations. However, this ethic was not always followed in practice, which is evidenced by the number of pre-and extra-marital pregnancies. It was also forbidden to relieve sexual tension through masturbation, which -according to the views of the Swiss doctor Samuel Tissot, popular in many countries and in West Prussia 39 -was supposed to have disastrous consequences for physical and mental health 40 . Male sexuality was to serve mainly procreative purposes. The dominant discourse was thus marked by the idea of a controlled and restrained masculinity. It is worth mentioning that in the conservative Polish community of West Prussia divorce was considered unacceptable according to the principle: "What God has put together let no man break apart" 41 . The subject of homosexuality remained almost completely tabooed, almost never appearing in the Polish journalism of the studied region, and when raised, it was presented in a very laconic way and reduced exclusively to the sin of sodomy, which aroused disgust among Catholic moralists 42 . --- Paweł Śpica --- Conclusion In the Polish journalism of West Prussia in the middle of the 19 th century, men were very rarely written about "as men", and the category of masculinity, typical for contemporary cultural studies, was not used at all. In contrast, a little more was written about men as husbands and fathers, but with little space devoted to their parenting function. This was because the Polish environment in West Prussia was strongly influenced by the discourse of gender polarisation dominant in the 19 th -century Central and Eastern Europe, which proclaimed that men and women had different roles in society, and the tasks of the man in the family were primarily focussed on its material stability, sense of security and external representation. Although a man as such played a decisive role in deciding on the education and future of his children, as well as in administering physical punishments to his offspring, the vast majority of the daily care and upbringing of the youngest members of the family was delegated to the wife-mother. This state of affairs was, by the way, seen as "natural". The dominant discourse of masculinity in the Polish environment in West Prussia of the mid-19 th century was thus clearly in line with essentialist concepts. At the same time, it was a discourse that preserved traditional social relations, and treated secular currents coming over time from Western Europe increasingly strongly as a threat not only to the existing social order, but also to the existence of the Polish nation. Hegemonic masculinity in the Polish community of West Prussia was constituted by many cultural factors, in particular by the national ideas and Catholic morality. Although it was a post-noble masculinity, it was dominated by a bourgeois and capitalist ethos. It was therefore at the same time a rational masculinity, however not devoid of the sentimental and emotional component characteristic of the noble insurrectionary ethos. Despite the fact that in the 19 th -century West Prussia the position and role of the male in the family and society was clearly defined, determined, among other things, by the patriarchal standards dominant in the European cultural, it must be emphasised that in the practice of family and social life this patriarchalism took on different shades, and in individual cases was only a formal, legal and ideological construction, not fully reflecting the everyday reality. Nevertheless, the traditional understanding of masculinity, typical for the 19 th -century West Prussia, was so firmly founded that it was not shaken by the turbulent history of the first half of the 20 th century. More pronounced changes in the understanding of the male role in Polish society and the family as a whole began to take place as late as from the 1970s onwards and have become particularly evident only recently. --- Bibliografia

The article presents cultural factors contributing to hegemonic masculinity in the Polish community of West Prussia in the mid-19 th century. Starting from the perspective of social constructionism, Raewyn Connell's theory of hegemonic masculinity was used in the research process to reveal the dominant discourse of masculinity in the studied milieu. The source base for the research was the Polish journalism of the region in question (predominantly the local press), mainly from the late 1840s and 1850s, less frequently from later periods. The contextual analysis of the research material made it possible to identify five basic, cultural determinants cultivating hegemonic masculinity in the Polish community of West Prussia in the period studied, analysed in detail (in separate parts) in the present article. These determinants include: the dominant narrative on the role of the male in the family, the position of the male in social structure, the role of the insurrectionary and organic work ethos, the influence of professed faith, including the views proclaimed by Catholic moralists, and the resulting beliefs about male sexuality. The research has shown the significant influence of Catholic ethics and national ideas derived from the nobility on the creation of a hegemonic image of masculinity in the Polish community of the studied region. This discourse was explicitly conservative and essentialist in nature. It also played an important role in maintaining and developing Polish identity in the absence of the Polish statehood.

Introduction Non-biomedical research, especially that using methodologies typical of the humanities and social sciences , has been chiefly grounded on the epistemological pillars of qualitative research. As described by Diniz and Guerriero , this type of research involves " participant observation, ordinary observation, open or closed interviews, ethnography, selfethnography and focus groups" and other methodologies. In the field of research involving human beings, the differentiation between biomedical and non-biomedical research occurs even in the definition of terms-research "in human beings" in biomedical areas, designating research that may be more invasive , and research "on human beings," which has a more social character . Exploring the nature of these different approaches to research involving humans, we note that they reflect distinct scientific cultures and readings of ethical research, including readings and application of documents supporting the mainstream ethics regulation of human-subject research such as the Nuremberg Code , the Declaration of Helsinki , in its versions endorsed in Brazil, and the Universal Declaration on Bioethics and Human Rights . For Guerriero , the approach in the biomedical field is supported by practices and methods aligned with these guidelines and regulatory standards for research in humans in Brazil, based on the Resolutions of the National Health Council , such as Resolution CNS 466/2012 . Thus, the norms resulting from this regulatory framework have a solid biomedical nature and are aligned with the particularities and research practices in this field of knowledge. Regarding the ethical regulation of studies in non-biomedical fields that adopt methods typical of those in HSS, it is grounded on Resolution CNS 510/2016 . It assumes that "...the researcher-participant relationship is continuously built up in the process of the research, and can be redefined at any time in the dialog between subjectivities, implying reflexivity and the construction of non-hierarchical relationships...", highlighting the Universal Declaration of Human Rights of 1948 and the Inter-American Declaration of Human Rights and Duties as pillars that support the dignity, freedom, and autonomy of humans . These distinct perspectives on the ethical regulation of research involving humans configure different research cultures with peculiarities in the epistemological and scientific domains. For "research culture" we mean "… the behaviours, values, expectations, attitudes and norms of our research communities. It influences researchers' career paths and determines the way that research is conducted and communicated." . In the context of research involving humans, there are cultural and epistemic "clashes" in the ethics review and research practices between biomedical and non-biomedical fields , especially those associated with HSS. This conflictual situation reminds us of Charles Snow )'s "Two Cultures" , which has been applied to different epistemological and disciplinary battlefields. --- The clash of two cultures in the conduct of human-subject research in Brazil In "The Two Cultures", Snow demonstrated the misunderstanding between two large groups, the literati and the scientists. As a physicist and a writer, he traversed these two environments and noted how different the ways of thinking were among the two groups; he even stated that they almost did not communicate at all, although he noted no significant differences among them regarding factors such as social origin or economic status . Regarding the distorted image of one group concerning the other, Snow describes the following scenario: Literary intellectuals at one pole-at the other scientists, and as the most representative, the physical scientists. Between the two a gulf of mutual incomprehension-sometimes hostility and dislike, but most of all lack of understanding. They have a curious distorted image of each other. Their attitudes are so different that, even on the level of emotion, they can't find much common ground. The appropriation of the "two cultures" embedded in Snow's view has been made in different academic contexts . In human-subject research, this perception of conflict is also evident when comparing research involving humans in the biomedical sciences and non-biomedical sciences, especially in HSS. Udo Krautwurst, in his "Culturing Bioscience; A Case Study in the Anthropology of Science" , discusses the assumptions that inform the ethics review processes in the biosciences. The author argues that there is a tacit assumption that "there is only one way to be ethical", and that as "subjects of research", humans are "equally powerless over what the researcher says or does. It is presumed that the research is necessarily on people rather than about people...". When it comes to research involving humans it is usually assumed that there is a power imbalance between the researcher and the participants, i.e., as the researcher would hold a position of power in this relationship, ethical concerns over this asymmetry becomes a sensitive issue . Addressing this power dynamic is among the elements underpinning the work of research ethics committees, which are usually expected to protect the rights and dignity of research participants. Yet, some controversy over this role exists, and this idea of protection of research participants is another issue that Krautwurst explores. In ethnographic studies, for example, the assumption that research participants necessarily need to be protected by research ethics committees is among the sources of conflict . Despite the lack of mutual understanding over the ethics of human-subject research in biomedical and non-biomedical fields, discussions on criteria for ethics review of HSS research protocols have put research ethics in the spotlight in Brazil. --- Bridging "two cultures" of research on humans through a heated debate on ethics review Notwithstanding the conflicts underlying the ethics review of human-subject research in biomedical and non-biomedical sciences , these two cultures somewhat mingle in the CEP/CONEP System. Since its inception, this regulatory system has the premise of encompassing all areas of research involving humanswhether biomedical or non-biomedical. In the HSS academic community, researchers and scholars in education have been among the most sensitive to the issue of ethics in human-subject research . Mainardes writes that "in addition to concerns with the ethics review standards and procedures, it is considered essential to conceive research ethics as a training issue, which involves the study and discussion of research ethics at undergraduate and graduate levels…". We believe the perceived importance that the role of ethics in research plays in the educational context has contributed to reflexive stances such as that by Mainardes . The launch of a document entitled "Ethics and Research in Education: Subsidies Volume 1" by the National Association of Graduate Studies and Research in Education reflects the attention on the big picture of research ethics. In this document, ANPED "reaffirms its commitment to the constant improvement of research in education and to the issue of research ethics" , with contributions by various relevant authors addressing risks, consent, confidentiality, vulnerability, plagiarism, research with vulnerable populations, etc. Guilhem & Novaes observe that notwithstanding the differences between the fields of social and biomedical research, there are similarities; for example, respect for the autonomy and dignity of people is a common central issue. Participants' free and informed consent is essential for research in both domains, differing only in the form of registration and, in some cases, time of obtaining consent. Aligned with this observation, the view that research ethics is not and should not be restricted to concerns in the regulatory field echoes several authors. Grisotti cites the limitations of establishing this ethical debate based on submitting projects to the Research Ethics Committee and thus reducing it to filling out forms on Plataforma Brasil. In fact, there needs to be more than this process to ensure responsible conduct in research. This thought is corroborated by authors such as Sobottka , who argues that the type of social control laid out by the resolutions is that mediated by the CEPs. "As an ex ante procedural control , it is limited in its scope and, above all, in its reach" . Brooks, Te Riele, & Maguire , in "Ethics and Education Research", point to the relevance of ethical discussions in education based on real situations experienced by researchers. Despite this sensitivity in the field of education, a study by Nunes has revealed the limited inclusion of research ethics in graduate programs in Brazil. This study analyzed 8,892 subjects in 37 education programs and found that only 0.78% of them include research ethics in their course syllabus. Nunes points to "the urgent need to guarantee disciplinary training and systematic on the topic of research ethics in graduate programs in Education". Mainardes highlights this gap as he emphasizes the need for more studies in the field of research ethics, an increase in the number of publications on procedures related to research ethics in HSS, translations of texts in foreign languages, and the need for a document with general guidelines on research ethics in conduct of research in education. Given this context, the goal of this study was to shed light on perceptions of coordinators of graduate programs in HSS on the ethics review of human-subject research in Brazil. Specifically, we investigated perceptions about the ethics regulation of research involving humans through a survey instrument sent to 148 coordinators of graduate programs in HSS in public universities. We sought to explore the relationship they had established with the Brazilian national CEP/CONEP system, which has been implementing changes in recent years, impacting research in these areas. As part of a previous study , the main hypothesis was that research practices for the ethical conduct of human-subject research in the biosciences in Brazil have established conversations with those in HSS studies. By engaging in the heated debate over ethics review, researchers in HSS have facilitated a reduction in the gap between these two cultures of conducting research involving humans. --- Methods Our research was based on a survey instrument sent to the coordinators of 148 graduate programs in HSS . They were distributed among six universities -among the top 20 public universities in the national ranking -in southeast Brazil, the region with the highest share of research funding and graduate programs in the country : Federal University of Rio de Janeiro , Fluminense Federal University , State University of Rio de Janeiro , University of São Paulo , State University of Campinas , and Federal University of São Paulo . The protocol associated with this step was approved by the CEP at the Clementino Fraga Filho University Hospital , UFRJ -CAAE 93926818.6.0000.5257. The sampling was intentional and non-probabilistic. This type of sampling seeks to collect information from a particular group, intentionally selected and of vital importance to the research, with the potential of understanding the problem. Drawing upon previous studies , Palinkas et al describe that purposeful sampling involves the identification and selection of people "especially knowledgeable about or experienced with a phenomenon of interest" and the importance to have "the ability to communicate experiences and opinions in an articulate, expressive, and reflective manner". The survey instrument, consisting of a demographic section with five questions and a content section with four questions, was prepared on the SurveyMonkey platform and then sent to the e-mail of the invited participants. The survey began on December 3, 2018, and the last response was received on January 7, 2019. Among the issues of interest in this study were those exploring how the relationship with the CEP/CONEP System had occurred during the transition process, still underway, regarding Resolution CNS 510/2016. After the independent analysis of the corpus compiling the content of the responses, we drafted thematic categories, which we then refined in an iterative analytical process that led to our defining of four categories. As described by Duarte , thematic categories are "articulated to the central objectives of the research" as well as "to the theoretical/conceptual references that guide the view of the researcher". --- Results and Discussion The survey of coordinators of graduate programs in HSS about the CEP/CONEP System explored their perceptions on aspects of the submission of research protocols for ethics review in their fields. With this survey, we also sought to understand the relationship of Resolution CNS 510/2016 with the setting investigated among participants. --- Quantitative findings from the survey across HSS fields -interacting with and reading the CEP/CONEP System For an overview of participants' institutions represented through their graduate programs, Figure 1 shows the percentages for each university, and Figure 2 shows the distribution of graduate programs' main fields represented by the coordinators , according to the assessment areas of CAPES. Most coordinators who participated in our research were serving their programs in this capacity for one to five years . Half of respondents declared that they had already served as coordinators or vice-coordinators before taking that position at the time of the survey. Among respondents to question P7 [whether he/she or his/her supervisee had already had a research protocol submitted to the CEP/CONEP System] , 67% did not have submitted their research for ethics review by the national regulatory system. One speculation could be that neither these participants nor their supervisees conducted human-subject research. In our sample, only one coordinator alleged that no member of the graduate program had conducted humansubject research up to the time of the survey. Yet, it is unlikely that it was the case for most HSS programs in our study, given the nature of most research in HSS studies, which encompass human behavior, agency, mindsets, social and cultural processes, among other phenomena. Instead, this result is aligned with the status of human-subject research in HSS in Brazil, in which the ethical framework has not been traditionally relying on the ethics review by the CEP/CONEP System . This observation echoes results shown in Figure 3. When asked about the submission to the CEP/CONEP System of research projects involving humans in their graduate programs, including those using interviews and/or questionnaires , only 32% declared that it was mandatory. However, 21% of the respondents indicated "Mandatory, but not included in any document of the graduate program". Based on the self-reports of the coordinators surveyed, the absence of normative guidance in almost 90% of these HSS programs is prevalent. Despite this lack, it is interesting to note that 60% of coordinators indicate existing recommendations or requests in their Programs for submission to the System , which gives insight into the changes in the landscape for ethics in human-subject research in Brazil. --- Qualitative findings from self-reports of participants The qualitative results from the textual corpus based on the self-reports collected from the survey were organized into four thematic categories, as described in the Methods section: On aspects related to the Plataforma Brasil and bureaucracy, the bureaucratic nature of the process of ethics review is an issue in many countries . In Brazil, researchers submitting research protocols to the CEP/CONEP System are also voices of complaints over the long material and/ or time involved in the submission process. In fact, such criticism is a common topic in biomedical and nonbiomedical fields . Among our respondents, excessive bureaucracy was mentioned by six coordinators . One comment that illustrates this concern is from Respondent #21: "[T]he system is super important and certainly prevents abuse. But it is also excessively bureaucratic. It gives importance to side issues and asks for an excessive number of documents". Overall, however, despite their criticism of several aspects, and based on their final comments , the coordinators surveyed seemed to be more supportive of the System than unsupportive. Box 1 ** -Final comments of 38 respondents [coordinators of graduate programs in humanities and social sciences in Brazil] on the role of national regulatory system in Brazil for human-subject research. --- How do you see the role of the ethics regulatory system in human research in Brazil? --- Supportive "I consider it an important safeguard for all parties involved . However, it is always important to have very clear risk typification, in order to streamline and reduce bureaucracy in research processes with minimal or zero risk to participants." "Consistent and necessary but needs adjustments in non-health areas." "It matters." "It gave rise to an important debate, in which it is important to take into account the characteristics of ethnographic research, such as those that predominate in the field of anthropology, and also the particularities of each context." "As fundamental to the development of our research." ** After the translation of the self-reports into English, minor edits related to punctuation, misspelling, and/or capitalization were made, for clarity. Práxis Educativa, Ponta Grossa, v. 18, e21955, p. 1-15, 2023 Disponível em: <https://revistas2.uepg.br/index.php/praxiseducativa> 9 "Important, although we have the risk of excessive bureaucratic regulation. The most interesting thing would be self-regulation, based on general normative principles, by the University and Research Institutions themselves, with due regard for broader legal implications." "Important and guiding." "Good. And I believe that researchers should be aware of the importance of respecting research collaborators, , as well as giving them feedback as soon as the research is completed." "As a necessity." "Positive but generalizes procedures from one area to another." "I consider it important, but it should be done based on a broad survey with researchers in the area." "I think it's important." "Indispensable." "I believe it to be extremely important and fundamental both in terms of preserving those involved in the research and the researchers themselves and avoiding possible abuses. In the case of the Humanities, however, especially in situations involving only oral or written language data collection, the intense bureaucratization can generate major obstacles for research." "[E]ssential." "[T]he system is super important and certainly prevents abuse. But it is also excessively bureaucratic. It gives importance to side issues and asks for an excessive number of documents." "Important but needs to be flexible to cover different areas of knowledge." "Fundamental. It should be extended to fields other than the biomedical sciences." "This is a relevant subject that deserves attention from researchers, supervisors, and program coordinators since it may involve interviews and experiments with humans. On the one hand, it is essential to establish some kind of regulatory system based on discussions with the scientific community. On the other hand, it is important to pay attention to the specificity of each area and the uniqueness of each research so that there is no excessive bureaucratization and a 'stiffening' of procedures." "Positive, but we have to be careful not to make research unfeasible, especially those that use the interview and photography procedures." "I think it's adequate, but perhaps too bureaucratic, especially regarding the functioning of Plataforma Brasil, which could be more user-friendly, and require less extensive use of handwriting. I also think that it is neither possible nor advisable for research in social sciences and humanities to be so tied to the formats and protocols of medical research. There is also a lack, but this is the case worldwide, of a specific resolution for searches on the Internet or with data collected via the Internet, in social media, for example, in addition to clearer delimitations on what public access, public domain, and intellectual property are in Internet content, and clarity for researchers about the differences between business ethics on the Internet and research ethics in that context, for example, regarding the construction of applications for generating research data." "Fundamental." "Adequate. Every time I needed ethics committees for my research or for my supervisees, we found possibilities." "Of extreme relevance and importance." "Fundamental. Research with so-called vulnerable populations, such as indigenous people, has to be regulated and looked at very carefully. It is a pity that in social sciences, we are treated as if we were health professionals collecting blood samples from the people we ask, in general, questions that do not involve any risk to their integrity." --- Unsupportive "The system sometimes becomes more demanding and rigorous than necessary, mainly with research with humans and not on these humans. In many cases, the human being is not the object of study, but the practices, regulations, or systems in private and public organizations. Even so, having to submit the research to the system becomes a time-consuming and unnecessary obstacle because it does not bring risks or harm to individuals. Often, the system is cunningly used as an argument to deny access to organizational data, even if there is no contact with humans, or there is no risk of exposure to them." "Complicated and controversial." "[C]omplex and confusing." "Excessive, with regard to conducting research that require interviews and application of questionnaires" Práxis Educativa, Ponta Grossa, v. 18, e21955, p. 1-15, 2023 Disponível em: <https://revistas2.uepg.br/index.php/praxiseducativa> 10 "Pretty confusing. It does not help the researcher much." --- Neutral "I feel that the discussions are very much based on references from the so-called "hard" sciences, addressing our methodologies as based on excessively objective aspects." "I think it has not yet been made clear that research in Human Sciences must follow this regulation. The Faculty of Philosophy at [university] was discussing the creation of [a] Research Ethics Committee, but I do not know its outcome." "Not aware." "Still under development and could be improved" "Centralizer. We would need to have a more diffused system, created by the specificities of each university institution and with its guaranteed autonomy for the management of ethical-academic processes." "I think it has improved." "[U]nknown." "In the case of human sciences, I believe it is still very little publicized. I cannot say that there has been a more direct and effective concern to frame research according to this system." Note that while bureaucracy is an issue raised by several coordinators, Respondent #47 reasoned that: [ethics review by the System is] "… an important safeguard for all parties involved . However, it is always important to have very clear risk typification, in order to streamline and reduce bureaucracy in research processes with minimal or zero risk to participants." One factor that is consistent with a relatively new ethics regulatory framework for research involving humans in HSS in Brazil is that the views of the CEP/CONEP System among the coordinators surveyed are not mature. Some of these respondents demonstrate that learning about the ethics review System is a gap to be filled. Respondent #39 states that "I think it has not yet been made clear that research in Human Sciences must follow this regulation. The Faculty of [field omitted] was discussing the creation of a Research Ethics Committee, but I don't know about the outcome.". This self-report is aligned with that from Respondent #2: "… Research with so-called vulnerable populations, such as indigenous people, has to be regulated and looked at very carefully. It is a pity that in social sciences, we are treated as if we were health professionals collecting blood samples from the people we ask, in general, questions that do not involve any risk to their integrity.". Accordingly, only some of the coordinators surveyed showed complete familiarity with or expressed a clear position on the CEP/CONEP System. Of the 14 respondents with comments associated with thematic category 3, "familiarity with and/or position regarding the ethical regulation of HSS", eight indicated that they had begun to approach the CEP/CONEP System. This observation is demonstrated in the following statement by Respondent #38: "Although I was aware of the Resolution through indirect references, I had never read its content carefully, given the specificities of research in History. However, in practical terms, there was always a concern of the supervisors of the area with the ethical values in research involving living human beings or with direct descendants up to the third generation, such as Oral History and research involving written documents or audiovisual records under public protection about recent History. However, as a program coordinator, this research caught my attention [motivating me] to more clearly formalize and standardize the specific issues in the field of History in terms of Research Ethics with and in humans. Thank you.". This recent encounter is also shown when, for example, it is declared that the CEP/CONEP System, though not unknown to researchers in HSS, is still distant from the reality of the research culture in their areas, as indicated by Respondent #19: "I've already heard of it, but I never studied it in depth". In fact, approximately 30% of the respondents were unaware of Resolution CNS 510/16, suggesting that despite Resolution CNS 510/16 putting "the ethical debate on another level", it takes time to consolidate a new resolution . Other coordinators alleged that although the regulation was known in HSS, the research they developed had not approached " more burning issues to be governed by the Resolution", as Respondent #14 comments. For Respondent #10, knowledge of Resolution CNS 510 was eventually gained, once the coordinator became a member of the CEP. Otherwise, " I would not have known, nor would I have sought to know." For other coordinators, their knowledge of Resolution CNS 510/16 is " [o]nly in general terms" ; this lack of specific knowledge is also reflected in the comment of Respondent #39: "I was aware of the one [Resolution] for health research. I didn't know about the [normative] demand for the humanities". Respondent #42 wrote that "[o]nly now, in 2018/19, we will have the first research project submitted to the Ethics Committee in Human Sciences.". These comments echo the results in Figure 3. Previous data have shown a timid number of graduate programs in HSS recommending or requiring submission of research protocols to the ethics review by the CEP/CONEP System . The changing landscape, with growing awareness of the ethics review system among HSS researchers, may be understood in light of an analysis by Diniz & Guerriero . The authors reasoned that "the imposition of the review system through research funding agencies, health institutions where data is collected or journals at the interface between biomedicine and the humanities…" motivated social researchers to "seriously address" research ethics . Finally, on the regulatory aspects of the graduate programs, consistent with Figure 3, there were different situations about the submission of research protocols to the CEP/CONEP System. Some of the coordinators reported at the time of the survey that their universities were implementing requirements or beginning to consider developing them. Respondent #8 claimed that "We haven't included it as part of some normative guidance, but we consider this inclusion a mandatory step". When it comes to normative guidance, Respondent #16 wrote that "[t]he rules at the university are changing and [submission to the System] will become mandatory and documented". In line with these times of transition, Respondent #22 explained that when students enrolled in their graduate program, they had to acknowledge awareness of the obligation to submit human-subject research to ethics review, by signing a specific document. Other coordinators reported that there was no recommendation, as can be seen in Figure 3. Respondent #26 is among the 21% of respondents reporting that submission to the System was "Mandatory, but not included in any document of the graduate program", as also shown in Figure 3. This respondent reported that the program he/she coordinated did not deem it necessary to make such a recommendation in any specific document, as it was assumed that researchers were expected to follow the principles of the national regulatory framework [CEP/CONEP System] for the ethical conduct of human-subject research. Overall, as can be seen in Box 1, a considerable number of respondents seem to be willing to get acquainted with, explore the possibilities of, and discuss research ethics in light of the ethics review process underway in Brazil. Whereas the challenges to address specificities and demands related to the ethics review of research involving humans in HSS remain, these coordinators surveyed suggest there is a promising space in HSS graduate programs to strengthen the role of research ethics in the design and conduct of research relying on human participation. --- Final considerations Our results indicate that some reluctance to interact with the Brazilian national regulatory framework for the ethics review of human-subject research in HSS is noted as a factor in our dataset. Nevertheless, according to our results, the majority of the HSS coordinators of the graduate programs surveyed is willing either to get acquainted with this regulatory system and/or exploring the ethics of conducting human-subject research in HSS. This finding comes from both the quantitative and qualitative data collected. Whether this finding has been influenced by the affiliation of these HSS coordinators, with public universities in southeast Brazil, which accounts for the highest share of research ethics committees is an open question. Irrespective of this possible source of bias, the data suggests that these coordinators, representatives of their programs, seem to appreciate this regulatory framework at the reflexive and normative levels. Although bureaucracy is an issue that may discourage interaction with the CEP/CONEP System, most of the comments in the corpus pointed to willingness to take a step further and address the growing demand for addressing the ethics regulation of human-subject research in HSS fields. Yet, these are the views of coordinators and do not necessarily reflect those from the faculty members in their graduate programs. Notwithstanding this caveat, according to these results, it seems reasonable to suggest that "two cultures" with their own scientific practices shaping research involving humans have been bridged by an ongoing debate over research ethics, which goes beyond the normative aspects of ethics review . We believe the perceptions of these coordinators in HSS reinforce the need for broadening the look at the role of research ethics in conducting human-subject studies in the humanities and social sciences. --- Limitations This study is not immune to different kinds of biases, including social desirability bias, which is "the tendency to underreport socially undesirable attitudes and behaviors and to over report more desirable attributes" , leading to the possibility of inaccurate self-reports and thus unreliable conclusions in survey-based research. Looking "ethical" in research is scientifically and socially desirable, which makes investigations on perceptions of the ethics review challenging, irrespective of being related to human-subject research. Additionally, the hierarchical position of these coordinators in their graduate programs and burden of responsibility to take a position toward a highly sensitive issue in the HSS community might have led to more conservative responses. However, we sought to minimize these potential biases by the type of survey questions, more related to factual information than behavior, organized into a semistructured instrument that offered space for participants to elaborate on their responses.

The ethics regulation of human-subject research in Brazil is by the Research Ethics Committee (CEP)/National Commission of Research Ethics (CONEP) System. While primarily focusing on biomedical research, this System also oversees the ethics of non-biomedical research. However, a distinct clash between these two research cultures is evident in Brazil. Despite existing guidelines, the cultural clash * This manuscript is an adaptation of part of the first author's master's dissertation, supervised by the second, with editions, modifications, and adjustments made in collaboration between the co-authors. The dissertation is listed in the references (Rocha, 2020). ** Doctoral student in the area of science education, management and dissemination in biosciences of the Graduate Program in Biological Chemistry of the Institute of Medical Biochemistry Leopoldo de Meis (IBqM) at the Federal University of Rio de Janeiro (UFRJ), with a scholarship from the Coordination for the Improvement of Higher Education Personnel (CAPES).

Introduction Several studies have demonstrated an increased prevalence of mental illness among medical students relative to the general population. [1][2][3][4] For example, Gaspersz et al found that the incidence of common mental disorders among medical students was 54% in the pre-clinical phase, and 48% in the clinical phase. 5 A meta-analysis of 59 studies providing a comprehensive summary of the mental health of medical students in Brazil found the prevalence of depression, stress, common mental disorders, and burnout to be 30.6%, 49.9%, 31.5%, and 13.1% respectively. 6 These and other studies recommend acknowledging and making attempts to alleviate the mental health consequences of studying medicine. 7,8 Experiencing mental illness during medical studies has implications for students' ability to apply cognitive, social, and clinical skills, ultimately leading to reduced competency in providing future patient care. 9 Several researchers in the Arab region have addressed the high prevalence of mental illness among medical students. In a systematic literature review, Elzubeir et al asserted that studies on Arab medical students' mental health report a high prevalence of perceived stress, depression, and anxiety. 10 A study conducted in Jordan reported a high level of exhaustion , disengagement , and "minor" psychiatric illness among medical students. 11 The same study also used the CAGE questionnaire to evaluate alcohol use among students, and found that 8% of students qualified as possibly alcoholic. 11 Another study conducted among medical students in Oman, a Gulf state with a socioeconomic context similar to Qatar's, found that the prevalence of Burnout Syndrome and depressive symptoms were 7.4% and 24.5% respectively; these figures were higher among preclinical stage undergraduates than among clinical stage students. 12 These and other studies highlight the importance of addressing medical students' mental health problems and call for a deeper understanding of the factors behind them. Mental health stigma is a major barrier to seeking help for mental health problems among medical students, leading to the exacerbation of the symptoms of mental health disorders. 1,13 Erving Goffman described stigma as a process by which an individual with an attribute that is deeply discredited by their society is rejected as a result of that attribute. 14 Based on this definition, the World Health Organization refers to stigma as "a mark of shame, disgrace or disapproval which results in an individual being rejected, discriminated against, and excluded from participating in a number of different areas of society". 15 In 2001, the WHO identified MHS as a key barrier to effective treatment of mental illness due to its negative impact on individuals' willingness to seek treatment. 16 Studies conducted in Western contexts show a correlation between MHS, hesitancy to seek help, and increasing symptom burden. 1,[17][18][19] For example, a cross sectional study of medical students at Michigan University identified stigma as an explicit barrier to the use of mental health services for 30% of first-and second-year medical students experiencing depression. 1 In another study conducted in Poland, both medical students and psychiatrists reported negative implicit attitudes towards mental illness. 18 Maalouf et al point out that stigma is scarcely addressed in mental health reports coming from the Arab region. 19 Similarly, Dardas and Simmons argue that little is known about how MHS manifests within the Arab community, making it difficult to design and test interventions. 20 Despite this gap in the literature, a few studies have reported on the impact of sociocultural values and religious beliefs on MHS. For instance, Ghuloum et al found that over a third of respondents in Qatar's general population believed that people with mental health disorders are intellectually disabled , and 48.3% believed that mental illness is a punishment from God. 21 Shehata et al's study of stigmatizing beliefs about mental illness among health sciences students and medical students in Egypt found that the majority of students held stigmatizing beliefs. 22 The positive correlation between MHS and medical students' reluctance to seek help has been evidenced by several researchers. 1,13 The prevalence of MHS among medical students can impact whether and how they seek help when they face mental health problems, which may lead to poor academic performance, substance abuse, damaged self-image, and eventually a decline in clinical practice and fitness to work with patients in the future. In this study we aim to determine the prevalence of mental illnesses among medical students in Qatar, taking the students in College of Medicine at Qatar University as a case study to represent medical students in similar Middle Eastern contexts. This is the first study of its type in Qatar, and as such, it contributes valuable knowledge for building tailored mental health preventive and intervention programs for medical students. This study is also the first to investigate these issues in the context of a problem-based learning curriculum in the Arab region; Elzubeir et al 10 noted that literature on medical students' mental health in this specific type of academic programme was lacking in the region. --- Methods --- Study Design A cross-sectional survey was conducted using the online survey platform Google Forms in August and September 2019, targeting all 250 medical students in the second to fifth years of the programme. Participants were required to complete a consent form before accessing the survey. The survey consisted of three parts. The first gathered sociodemographic data on each student; the second identified the prevalence of mental health problems among the students using standardized screening tools; and the third examined students' views on stigma. Since the instruction language in the college is English and students are fully proficient in the English language, all materials were distributed using the original English versions. --- Setting The College of Medicine at Qatar University was established in 2015, and is one of the only two medical schools in Qatar. The medical program in CMED at Qatar University consists of six years. The first year is transition phase, in which students study the basic core subjects that prepare them for the medical curriculum. This is followed by two and a half years, which constitute the pre-clerkship phase, in which students study the basic medical sciences using a problem-based learning pedagogical approach. Finally, in the last two and a half years, known as the clerkship phase, students continue their studies in hospitals and other clinical environments. The students are of different nationalities, with a majority of Arab-Muslims and a minority of Muslims non-Arabs from Asian countries such as Pakistan, Bangladesh, and the Philippines. All non-Qatari students are children of expatriates and were born and raised in Qatar. --- Inclusion and Exclusion Criteria The inclusion criteria included respondents who were in at least their second year and were actively enrolled in medical studies. The study excluded Phase I students, who were not yet studying basic medical sciences. The study also excluded students who had dropped out. At the time of the study, there were no sixth-year students. --- Survey Instruments The survey consisted of a total of 64 questions divided among the three sections. --- Sociodemographic Data This section was based on the sociodemographic section of the WHO's Study of Global Aging and Adult Health . SAGE is a longitudinal study implemented in several countries with the primary goal of generating valid, reliable, and comparable information on a range of health and well-being outcomes that hold importance to public health. 22 It is a standardized survey instrument which collects basic sociodemographic data such as age, gender, nationality, marital status, current accommodation status, family income, and parents' level of education and employment history. In addition we added two questions about year of study, and a year repetition. --- Symptoms of Depression, Anxiety, and Psychological Distress The second section collected data on the students' mental health using three standardized screening tools: the Patient Health Questionnaire , Generalized Anxiety Disorder , and Kessler Psychological Distress Scale . These tools have been previously validated for screening purposes globally and regionally, and they are commonly used in Qatar in clinical settings and for research and screening purposes with the general population. [23][24][25] The American Psychological Association reports, based on previous studies assessing the PHQ-9, that the tool exhibits good diagnostic validity; scores greater than 10 had a sensitivity and specificity of 88% for Major Depressive Disorder. 26 It also has high internal consistency and well-constructed psychometric properties. 26 Accordingly, the Diagnostic and Statistical Manual of Mental Disorders -5 recommends its use as a tool to evaluate the severity of depression. 27 This study utilized the 9-item PHQ-9 to assess for self-reported symptoms of depression. A score of 5-9 indicates mild symptoms, a score of 10-14 moderate symptoms, a score of 15-19 moderately severe symptoms, and a score of 20-27 severe symptoms. Several international studies conducted in Bangladesh, Korea, Portugal, United States, Iran, Germany and Peru have examined the validity and reliability of the GAD-7 tool across many different populations and found it to have high internal consistency as well as good convergent validity. 28 Spitzer et al 29 confirm that the GAD-7 has good reliability, in addition to criterion, construct, factorial, and procedural validity, indicating that it is a valid and efficient tool for the assessment GAD in research and in clinical practice. A score of 6-10 on the GAD-7 indicates mild anxiety, a score of 11-15 moderate anxiety, and a score of 16-21 severe anxiety. Country-specific validation studies assessing the K-6 tool in different languages, including English, have come to the consensus that this tool has good internal consistency reliability and factor-based and other construct validity across various populations. Furthermore, multiple studies have demonstrated its effectiveness in measuring non-specific psychological distress in non-clinical populations, including Arab populations and student populations. 30,31 A score of 1-5 on the K-6 indicates mild distress, a score of 6-10 moderate distress, and a score of 11-24 severe distress. --- Mental Health Stigma The third and final section consisted of 27 questions adopted from Schwenk et al 1 which were used to evaluate respondents' levels of stigmatizing beliefs about mental health and their attitudes toward seeking help. The items in this section addressed points identified in previous studies and implemented both positive and negative framing to allow for an internal consistency check. Section three also included validated instruments on stigma used in studies of the general population. Focus group discussions and pilot-testing were performed to modify the questions. These items were specifically tailored to fit medical students. 1 Cronbach's alpha was used as a measure of internal consistency reliability for this section and was found to be 0.789, which represents acceptable internal consistency. --- Study Size and Bias The population of our study was well-defined and small, since the medical student community in Qatar is limited; therefore, total population sampling was used to minimize potential bias occurring through sampling technique. Five responses were eliminated due to possible response bias, as these participants had selected extreme values on the Likert scales for all items. --- Statistical Data Analysis Statistical analyses of the survey responses were performed using the STATA 16.0 software package. --- Determination of Depression, Generalized Anxiety, and Psychological Distress To evaluate the prevalence of depression, generalized anxiety, and psychological distress, the scores for the three screening tools in the second section of the survey were calculated. The established standardized cut-off scores for each screening tool were then used to classify respondents as exhibiting "no", "mild", "moderate", "moderately severe", or "severe" symptoms of depression and "no", "mild", "moderate", or "severe" symptoms of anxiety and psychological distress . To further facilitate statistical analysis, students were then divided into "low-risk" and "high-risk" groups for each disorder using standardized cut-off scores that indicate a highly likely requirement for medical intervention . Separate stepwise logistic regressions were run to test the correlation between the risk of depression, anxiety, and psychological distress and each of the sociodemographic factors to assess the impact of these factors on the students' mental health. --- Determination of Mental Health Stigma The questions in the third section of the survey were classified as positive or negative with regard to the respondents' viewpoints on stigma. Positive questions were those for which a response of "agree" demonstrated stigmatizing view; therefore, "agree" responses were assigned a score of 1 and "disagree" a score of 0. Negative questions were those for which an "agree" response revealed no evidence of stigma; therefore, "agree" responses for such questions were assigned a score of 0 and "disagree" a score of 1. An overall stigma score was calculated for each respondent. The MHS scores were then divided into three quantiles, with the two lower quantiles defined as students holding lower levels of stigma and the upper quantile as students with high stigma. Stepwise regression using a p-value <0.2 was used to determine which sociodemographic factors were associated with high stigma. --- Ethics and Informed Consent This study complies with the Declaration of Helsinki, and received ethical approval from Qatar University's Institutional Review Board . Participants gave informed consent using an online form, which they were required to complete prior to accessing the survey questions. As this study was anonymous, participants who were found to have severe depression, anxiety, or psychological stress could not be referred for care; however, the survey included a section that advised the students to call a hotline operated by Hamad Medical Corporation if they answered certain questions positively. --- Results --- Characteristics of the Participants The overall survey response rate was 74.8%. The response rates for second, fourth, and fifth-year students were 83%, 85%, and 76% respectively. The response rate for third-year students was lower, at 48%. Five responses were eliminated due to response bias, as these participants had selected extreme values on the Likert scales for all items. Of the remaining sample of students , 68.1% were female and 31.9% were male. 36.3% of respondents were Qatari, 45.1% were from other GCC and Middle Eastern countries, and 18.9% were from other countries. The largest age groups represented in the sample were 17-20 and 21-25 . The mean age of the students was 21, with a standard deviation of 9.1 . --- Prevalence of Self-Reported Symptoms of Mental Health Disorders Depression The prevalence was 31% for mild depressive symptoms and 30% for moderate depressive symptoms. Eight students exhibited severe depressive symptoms and 32 had moderately severe depressive symptoms, representing a prevalence of 4.4% and 17.6% respectively. When categorized into low-and high-risk groups, 52.2% of the students were classified as having a high risk of depression. --- Generalized Anxiety The rates of mild and moderate symptoms of generalized anxiety were 29% and 27.5% respectively. Additionally, 10.4% of the students exhibited severe symptoms of generalized anxiety. When classified into low-and high-risk groups, 37.9% of the students were found to have a high risk of anxiety. --- Psychological Distress A total of 60 students, or 33% of participants , were found to demonstrate mild symptoms of psychological distress, while 44 students, or 24% , showed moderate symptoms. The prevalence of severe symptoms of psychological distress was 39.6% . When categorized into low-and high-risk groups, 22% exhibited a high risk of psychological distress. Stepwise logistic regression indicated that students whose fathers had graduate or postgraduate degrees had lower odds of developing symptoms of depression than those whose fathers had only a primary or secondary education, as indicated by an odds ratio of 0.14 and an OR of 0.15 respectively . --- Association Between Sociodemographic Factors and Risk of Symptoms of Anxiety Students whose fathers held graduate or postgraduate degrees also had lower odds of developing symptoms of anxiety compared to students whose fathers had only completed primary or secondary school. This was shown by an OR of 0.26 and an OR of 0.25 , respectively . --- Association Between Sociodemographic Factors and Risk of Symptoms of Psychological Distress Students whose mothers were never employed had higher odds of developing symptoms of psychological distress compared to students with working mothers, as shown by the OR of 3.65 . --- Stigma Analysis The calculation and categorization of students' stigma scores revealed that 31.9% of the students exhibited signs of high stigma against mental illness. Stepwise logistic regression revealed several associated factors, including nationality, whether the student had repeated an academic year, the year of studies the student was currently in, and maternal and paternal levels of education . --- Discussion --- Prevalence of Mental Illness Symptoms In this study, we used self-report screening tools to measure mental illness symptoms. Our findings indicate a higher prevalence of depression among medical students compared to the general population of Qatar. 20,32,33 Sharkey argues that "the reported numbers [in the general population] are likely to be underestimated because of the limited diagnostic abilities of primary care doctors" [p20] 34 . The gap between our findings and past studies on the general population may be explained by the excellent access to and preexisting relationships with research subjects that our context and sample afforded, and/or to our use of screening tools that depend on self-reporting-and thus reflect experience of symptomsrather than official diagnosis of mental illness. However, the discrepancy remains alarming. 52.2% of the students in our study had moderate to severe levels of depression. This is in line with the results of other studies of medical students in the Arab region, including studies in Saudi Arabia, Jordan, Egypt, and Oman. [8][9][10][11]35 For instance, a study conducted recently in Oman reported high prevalence of depressive symptoms among medical students. 11 The similar environmental and cultural backgrounds in these countries could explain these comparable levels of depression among medical students. These findings are also supported by a meta-analysis conducted by Puthran et al, who found that the global prevalence of depression among medical students is 28%, and medical students in the Middle East are more likely than non-Middle Eastern students to be depressed. 36 The prevalence of generalized anxiety disorder among medical students in our study, based on the GAD-7 scale, was found to be 67%. This is also substantially higher than the prevalence among the general population, which has been measured at 10.3%. 20,32 It is also higher than the global figure reported in a meta-analysis of 69 studies conducted in February 2019, which reported the prevalence of anxiety among medical students as 33.3%. 3 Our study found the prevalence of psychological distress among participants to be 96.71%, with 63.83% reporting moderate to severe symptoms. This is higher than prevalence of psychological distress among medical students in Saudi Arabia and Egypt. 8,38,39 Our analysis suggests that these high levels of stress may be attributable to both academic and non-academic factors. The medical curriculum and workload do contribute to the reported stress levels, but MHS and sociodemographic factors also help to explain the numbers amongst medical students in Qatar and other nearby countries. 37 --- Association Between Sociodemographic Factors and Reported Mental Health Problems We found correlations between mental health and several sociodemographic factors, including parents' education, academic year, year repetition, gender, and nationality. Our study found that students whose fathers had higher education --- 975 were less likely to develop depression and anxiety. This is in line with Park et al, who found that the children of mothers with less than a secondary school education had higher odds of experiencing a major depressive episode . 40 They concluded that a low level of maternal education was associated with MDE in early adulthood, independent of paternal education and other childhood and early-adulthood risk factors. This may be because parents with higher education are more aware of the challenges and stressors of academic life and are therefore better able to support their children during their journeys. Students in their fifth year reported a higher prevalence of depression and anxiety symptoms than students in other years. This finding is in line with other studies conducted in the Middle East, 41 but differs from studies conducted elsewhere. 35 As a result, there is no consensus in the impact of the year of studies on medical students' mental health. The differences could be explained by the different medical curricula employed in each college, and the different sociocultural factors between the Middle East, north Africa and the Gulf. We found that Arab students were at higher odds of developing anxiety compared to their peers from other nationalities and cultural backgrounds. This is consistent with Quek et al, who found that medical students of Middle Eastern and Asian origin are at higher odds of anxiety than students from other parts of the world. 3 This could be attributed to the fact that MHS in Middle Eastern societies is considerably higher than in other regions, which could discourage students experiencing mental health challenges from seeking help. This stigmatization leads them to reject being labeled as mentally ill, exacerbating their mental health symptoms due to delays in seeking help, suboptimal treatment, and poor outcomes. 39,40 Hankir et al found a positive correlation between MHS and reluctance to seek help among medical students. They posited that medical students hesitated to seek help due to fear of being perceived as less capable and discredited by their colleagues and mentors. 12 MHS in general is also a barrier for applying for jobs, renting homes, finding partners, and more. 44 Our findings reveal that male students were at twice the odds of developing psychological distress compared to female students. We believe that these findings are due to the patriarchal beliefs common in Middle Eastern cultures, in which men are shamed for revealing mental health problems and those who do are perceived as weak and unworthy, or "not a man". 42 Another possible explanation is the higher mental health literacy among women. 45 Our study also found that students whose mothers had never been employed were at greater odds of experiencing psychological distress. This differs from the results reported by others, such as Ibrahim et al, who found lower anxiety risk among students with unemployed mothers, but a higher risk for those with unemployed fathers. 37 We argue that the interplay between the medical curriculum and social and sociodemographic factors explains these diverging results in different contexts. --- Association Between Sociodemographic Factors and Stigma Middle Eastern cultures value family honor and the concealing of emotions. These values contribute greatly to stigma regarding mental health services, as those who seek such services could be viewed as weak and thus bring shame to their families. 42 Mental health stigma leads to delays in seeking help, suboptimal treatment, and poor outcomes. 43 Hence, preventive programs in medical schools should be seen as an essential component of the curriculum. In a systematic review of studies addressing MHS in the general population in Qatar, Zolezzi et al concluded that "most people have poor mental health literacy, insufficient information about the causes and symptoms of mental illness, and stigmatized perceptions toward people with mental illness" [p1324] 47 Zolezzi et al also found a wide range of stigmatizing beliefs, actions, and attitudes regarding the treatment of mental illness. 48 This explains our finding that students in Qatar are likely to have stigmatized views of mental health problems, influenced by several sociodemographic factors such as parents' education, year of studies, and repetition of an academic year. Past studies have shown that students who have repeated a year of their studies develop low self-esteem and perceive the cause of repeating the year as personal inability to perform, which leads to higher risks of mental health problems and stronger MHS. 45,46 We also found a correlation between living with family and higher mental health stigma. Living with family can affect a person with mental illness either positively or negatively depending on whether or not their family responds supportively to the difficulties they face. Families' attitudes in general often reflect the social stigma toward mental illness. Past studies have indicated that family members of persons with mental illness often express stigma in relation to https://doi.org/10.2147/AMEP.S371053 --- DovePress Advances in Medical Education and Practice 2022:13 976 their relatives' illnesses. 47 This stigma is transferred to the people experiencing mental illness-in this case the medical students-who feel increasingly stigmatized and avoid seeking help. --- Implications for Practice The results of this study demonstrate the need for more effective and numerous awareness campaigns, workshops, and lectures designed to break the stigma surrounding mental illness in the general population and among medical students in particular in Qatar. This will increase the chance that students will seek help in earlier stages of their illnesses and alleviate academic, social, and other complications. We believe that incorporating wellbeing programs in medical schools curricula, tailored to the sociocultural context and the needs of students, can be a preventive action that will encourage students to seek help when they need it. Such programs can help students in manage their time, stress, and energy in a way that helps them find time to care for themselves and boosts their mental health. We also recommend that medical educators act as role models to their students and discuss these issues openly, sharing their personal stories of mental illness and how they sought help and recovered. We recommend including activities like empowerment and life skills workshops within the curriculum, which can indirectly help to prevent and treat mental illness among students. --- Conclusion and Limitations Based on the data we collected in this study, we conclude that the medical curriculum is not the sole cause of mental health problems among medical students; rather, a combination of several factors and structures is responsible. The heavy academic demands on students interact with sociodemographic factors and with social attitudes and stigma towards mental illness. Such stigma prevents them from seeking help and leads to a deterioration that could be prevented with earlier intervention. Although this study provides evidence for the association of several factors with stigma, it is important to consider further factors that could contribute to students' hesitancy to seek help. Among these factors are level of religiosity, self-esteem, family support, and reasons for choosing to pursue medicine. This study invites deeper analysis and further qualitative investigation to better explain its findings. In addition, it should be noted that the COVID-19 pandemic, which began a few months after our survey was conducted, has caused significant impact on medical students' mental health. It is important to examine this impact, and any decrease in MHS and subsequent changes in students' help-seeking behaviors which may have resulted from the widespread mental health crisis which accompanied the pandemic. The main limitation of this study was its inability to include medical students from the other college of medicine in Qatar due to bureaucratic obstacles. Of the two colleges of medicine in Qatar, the one in which this study was conducted is larger and more diverse. However, future studies that include both medical schools in Qatar may be able to determine whether the type of curriculum influences students' mental health in the same context. --- Disclosure The authors declare no conflicts of interest in relation to this work.

Past studies have shown high prevalence of mental illness among medical students. This is often linked to the demands of the medical curriculum, and to mental health stigma that prevents students from seeking help. This study aimed to examine experiences of mental health problems among medical students of different nationalities in Qatar and to uncover sociodemographic factors related to their prevalence and associated stigma. Methods: A cross-sectional online survey was conducted with medical students in their second through fifth years at the College of Medicine at Qatar University. The survey began with a consent form, and those agreed to take the survey were directed to the questionnaire. The survey comprised 64 items across three sections. The first section collected sociodemographic data. The second section screened depressive symptoms using the PHQ-9; anxiety symptoms using GAD-7; and psychological distress symptoms using Kessler-6. The third section included 27 questions adopted from Schwenk et al, which evaluate students' perceptions of stigma and their attitudes toward seeking help with their mental health. Results: One hundred and eighty-two students participated in the study. The prevalence of self-reported symptoms of severe depression, anxiety, and psychological distress was 4.4% (95% CI 2-9), 10.4% (95% CI 7-16), and 39.6% (95% CI 33-47), respectively; the prevalence of high stigma was 31.9% (95% CI 25-39). Parental education, repetition of an academic year, progress in medical studies, gender, and nationality had statistically significant correlations with mental health problems and stigma. Conclusion: In addition to the impact of the requirements of medical study, the high prevalence of reported mental illness among medical students is impacted by sociodemographic factors and the mental health stigma that constitutes a barrier to seeking help. Preventive wellbeing programs should be an essential component of medical curricula.

Introduction "We are in the same storm, but we are not all in the same boat" Barr , cited in Zhou and Kan The arrival of COVID-19 brought with it unprecedented global change. At the time of writing, the USA had almost 47 million cases , and Britain was on its way to 10 million , with numbers rising as the winter months approached. The most effective strategy to minimise transmission of the virus has been to restrict movement of people, including quarantine, isolation and a series of national lockdowns in the UK . In a global first, this has included the closure of schools and universities, who instead opened their virtual doors to students for much of 2020. Though the vaccine has enabled greater movement, including the return of face-to-face higher education, the pandemic, associated restrictions and their collective impact persist. Those with direct experience of COVID-19 are more likely to be depressed, anxious and experience post-traumatic stress disorder , while as a wider group, feelings of stress and loneliness are at their highest in over a decade . The title of this paper highlights how different ethnic groups within the USA and UK are not all in the same boat but actually, all in the same storm. As Damian Barr explains while 'some of us are on super-yachts, some have just the one oar'. When a storm like COVID-19 hits, it became visible that people racialised as Black are some of the most vulnerable in society; caught in the eye of the storm with very few material tools to survive it. While the impact of COVID-19 has been far-reaching, evidence points to specific groups being at greater risk of its effects. In terms of mental health impacts, for example, those at highest risk are young adults , women , ethnic minority communities , people experiencing socioeconomic adversity , unpaid carers and clinically vulnerable groups such as those with disabilities . The Doreen Lawrence review captures the systemic nature of the problem, recognising that "Covid-19 has thrived on inequalities that have long scarred British society" . --- The impact of COVID-19 on Black, Asian and ethnic minority communities The pandemic has further illuminated the wide-reaching consequences of structural racism: Black men are 4.2 times more likely to die from a COVID-19-related death than White men , while the risk of infection is 56% higher within Black, Asian and ethnic minority communities compared to White British . The first ten healthcare professionals in the UK to die from COVID-19 were from the Black, Asian and ethnic minority communities , as were one-third of the COVID-19 patients admitted to critical care units in 2020-of which the highest percentage were Black . Accompanying the aggregate of research in the UK , a similar pattern is observed in the USA , with death rates for Indigenous , Pacific Islander , Latinx and Black higher than White and Asian Americans and rates of hospitalisation higher across ethnic minority communities than White . With each COVID-19 death leaving behind bereaved family members , there is a predicted surge in post-pandemic mental health disorders disproportionately affecting ethnic minorities . The current mental health impacts of COVID-19 on ethnic minorities are notable , with ethnic minorities in the USA reporting disproportionately worse mental health outcomes and in the UK reporting higher levels of depression, anxiety, abuse, self-harm and thoughts of suicide since before the pandemic . Findings from Proto and Quintana-Domeque indicate that Bangladeshi and Indian men experienced a higher risk of mental health problems than White men , though research from the UK Household Longitudinal Study found no interethnic differences in mental health problems when comparing pre-and during-pandemic data . Other studies suggest that it is unlikely ethnicity itself causes differences in mental health outcomes, but rather correlations with other factors such as occupation, low income, higher likelihood of infection or death and racism . The clear and disproportionate effect of COVID-19 on ethnic minorities across the USA and UK thus highlights the impact of racism on physical and mental health. As Razai et al. , racism both shapes social determinants of health and has its own effect on the health of ethnic minorities. These social determinants include socioeconomic status, living in urban areas, poor and overcrowded housing, high risk occupations, comorbidities and cultural barriers, all of which ethnic minority groups are more likely to be adversely impacted by Razai et al. . Over time these have, and continue to, lead to worse health outcomes for minoritised communities-as J. Kēhaulani Kauanui states, 'racism is a structure, not an event' . It is unsurprising, then, that Public Health England have suggested racism and discrimination may have contributed to increased risk of exposure to and death from COVID-19 among ethnic minority groups and that discrimination, cultural and institutional racism are empirically associated with ethnic disparities in COVID-19 outcomes . 'White supremacy', much like COVID-19, is a virus . When speaking about the impact of COVID-19 on ethnic minorities, therefore, we must never stray far from the topic of racism. --- Black Lives Matter While the purpose of this article was not to explain why Black Lives Matter, 1 it has done so by virtue of the data. As the evidence reflects, Black people are among the most disproportionately impacted by COVID-19 both across the USA and UK contexts. Considering the current and historic experiences of Black people in a multitude of areas such as housing , employment ) , income , poverty rates , criminal justice and health , the impact of COVID-19 on the Black community is devastating but unsurprising. And yet, it was the public murder of George Floyd by Minnesotan police that marked a 'rest-of-the-world' awakening to the lived reality of Black communities. By summer 2020, the Black Lives Matter movement rose to global prominence which, whilst reigniting a collective desire to fight for Black human rights, also 'reopened psychosocial wounds' . Together with arrival of COVID-19, itself considered a traumatic event , for Black communities, 2020 and 2021 have been a 'pandemic within a pandemic' . The ongoing traumatisation of Black communities thus persists, and 1 Rather, it is assumed. with it a need to tend to and centre Black mental health. This is echoed in the evidence base, which calls for more attention to be given to Black lives , particularly Black women , and especially Black women in academia who are centred in the current study. --- Black in academia One of the hotspots where inequality thrives is the Academy, which has long been associated with structural racism . Black students experience greater mental health problems and difficulties transitioning into the Academy , having navigated a schooling system with a 'whitewashed curriculum' , disproportionate rates of exclusions and anti-Black racism . For example, Bird and Pitman examined the lack of diversity on reading lists at two research intensive UK universities and found only 7% of the Social Sciences authors reviewed were from "BAME" backgrounds-despite there being a 39% UK domiciled student population. Also, in Mwangi et al. study 'Black elephant in the room', Black students contextualising campus racial climate within US racial climate, Black student participants 'demonstrated how the systemic anti-Black racism being resisted through larger movements such as Black Lives Matter is also reflected and reproduced in US universities ' . The effects of these show anti-Black racism is fluid and relentless , reflected in attainment and outcomes of young Black students at key educational stages including school and university . Among Black staff, institutional racism yields different but equally immobilising effects, such as higher numbers of precarious contracts , lower retention rates and restrictions to development and progression . As Bowden and Buie note, the way these manifest are often covert and insidious: One does not need explicitly racist policies to yield racial disparities. For example, providing grant supplements only to those with active awards is not openly racist; however, when you consider that Black principal investigators are funded at a much lower rate than their white colleagues, the result is less access for Black researchers to these funding opportunities. These challenges are reflected in the mental health experiences of Black students and staff , perpetuated by stigmatisation, lack of belonging, isolation and marginalisation. For Black women in particular, the lack of support beyond that which is self-and communally cultivated creates pressures that can lead to mental health implications . With staff in particular, these pressures are exacerbated when their White female counterparts are active participators in their exclusion in the workplace, and thus strategies are taken for self-care and survival . Compounding these experiences are the differential healthcare outcomes for Black staff and students who do attempt to access support, which include overdiagnosis and undertreatment , as well as a lack of culturally responsive services . Unsurprisingly when considering the evidence, the sudden imposition of online teaching, cancellation of lectures, use of alternative assessments and adjustments to new technology have had differential impacts on students and staff of colour. Research shows that over the pandemic, ethnic minority groups, including those from the Black community, have had reduced access to research resources and have been more adversely affected in terms of mental wellbeing, social isolation and academic support . Undoubtedly, there are also incoming Black university students impacted by the use of algorithms and teacher assessments, which together produced disproportionate downgrading of GCSE and A-Level results2 . In the USA, there are higher levels of attrition in higher education among ethnic minorities and steep declines in university and college enrolment since the pandemic began, with a 14.3% drop among Black men . Some of the largest barriers reported were acclimation to the online curriculum, financial constraints and prioritising care-giving responsibilities; though without disaggregation of ethnic minority data, it is unclear whether these experiences represented those of Black students. Similarly in the UK, while young carers are 1.5 times as likely to be from Black, Asian and minority ethnic communities , it is not clear what proportion of these are Black young people. This exemplifies, and further reinforces, the need to disaggregate ethnic minority data if we are to better understand the impact of COVID-19 on Black students. As Pennant notes, there is a need to support Black students, especially Black girls and women, who continue to endure the sharper end of both racism and the pandemic . For instance, prior to the pandemic, young Black girls have been rendered invisible in secondary education regarding academic success . Therefore, the racialisation of gender is multi-dimensional and important to consider amongst young Black pupils in the academy. Meanwhile, university staff are feeling the weight of responsibility to manage the impact COVID-19 has had on their students-and in particular, Black academic staff who do so with an additional 'Black Tax'3 . Black women bear the weight of this even more so, describing the pressure to 'save our students from the stress and weight of the pandemic' and 'save our departments from the increased scrutiny around how we view, assess and address diversity, equity and inclusion' , particularly in line with the concurrent upsurge of the Black Lives Matter movement. Just as Bowden and Cullen write: 'truth be told, having to work harder to obtain similar opportunities to our colleagues takes an emotional toll. We are tired' , Walton and colleagues say: 'We are sick and tired, of being sick and tired!' . The impact of COVID-19 on the next generation of Black graduates and the Black staff in Higher Education is thus significant. However, its long-term effects on Black staff and students are unknown , and there are a number of calls for further research that explores the impact of COVID-19 on Black communities. Indeed, many of the recommendations arising from such research are around engaging in further study , such Proto and Quintana-Domeque who call for additional research on the potential differential effects of the COVID-19 pandemic by ethnicity. Similarly, the UK Centre for Mental Health highlights a need for research that goes beyond outdated and homogenised comparisons of 'White' and 'BAME' to focus on specific ethnic groups, without which we cannot truly understand the variances in health and outcomes that emerge between different communities . Given the paucity of such research within the Academy, it is the aim of the current study to explore the experiences of Black students and staff within the context of the COVID-19 pandemic. Recognising that within the US and British education systems, Black girls and women are often ignored , this research places particular emphasis on these voices. --- Critical race theory One way in which to uplift the voices of the Black community is through critical race theory , a theoretical framework that examines the role of race, racism and power . With its roots in legal studies and Black feminism , a founding tenet of CRT is in its commitment to eliminate all forms of oppression through an interdisciplinary approach . Solórzano and Yosso define critical race theory in education as: a framework or set of basic insights, perspectives, methods, and pedagogy that seeks to identify, analyse and transform those structural and cultural aspects of education that maintain subordinate and dominant racial positions. Theoretic elements of CRT include the centrality and intersectionality of race and racism, race as a social construct , whiteness as property , the challenge to dominant ideology and the value of experiential knowledge . The first, centrality of race, is about the permanence of racism and how it is so embedded within society that it is difficult to recognise and hard to address . It should therefore be foregrounded in research . This includes intersectionality Crenshaw, 1991), acknowledging how the centrality of racism is impacted by other interrelated identities and social structures, such as gender, sexuality and class . The second element, social construction, means race is not inherent within itself or a person, but is a complex and longstanding product of society . Third, whiteness as property refers to the value whiteness holds in relation to other races . Historically, this was in the context of property and rights, but over time has become associated with other abstract concepts such as time, creativity or education . The fourth, challenge, means countering dominant narratives around neutrality, objectivity, colour-blindness and meritocracy in society, all of which maintain the value of whiteness. One way in which to challenge these 'majoritarian stories'4 is through a fifth element, experiential knowledge. CRT research values the lived experiences and narratives of minoritised communities, i.e. counter-storytelling , hence its use in the current study. Utilising a critical race theory framework, this paper attempts to explore the disproportionate impact of the COVID-19 pandemic on Black students and academic staff in UK and US higher education. Importantly, this paper will argue that systemic racism within the Academy plays an integral part with regard to reproducing social and health inequalities, particularly in the midst of a global pandemic. This ran concurrently alongside the murder of George Floyd and subsequent global outcry, which was a key topic for discussion. The study emanates from the marginalised voices of forty-three Black students and academic staff within the UK and US higher education sector. Exploring the impact of the pandemic against of backdrop of structural and institutional racism serves as an important vehicle not only for promoting counter-storytelling narratives but also advocating to redistribute the labour of anti-racist endeavour away from Black members of the Academy. --- Methodology --- --- Procedure Participants engaged in individual, semi-structured interviews and focus groups to explore their lived experiences of navigating the pandemic whilst in higher education. To inform the focus groups, all participants completed a reflective journal to chart and document their experiences of COVID-19. These became reference points and points of departure within the facilitated group discussions. Participants were contacted through purposeful sampling aided by both researchers institutional and personal affiliation to UK and US institutions. The objectives of the study were explained to all the participants, and informed consent was ethically obtained. Participants were drawn from a combination of staff and students from UK and US higher education institutions. These ranged from the Russell Group 5 and Post-1992 6 universities in the UK; and Ivy League, 7 Historical Black Universities 8 and R1 universities in the USA. 9 The sample comprised of eighteen students and twenty-four members of staff at mid-career level. In the total sample of forty-two participants, there were twenty-six Black women and sixteen Black men . All were interviewed on multiple occasions between 2019 and 2021 through a mixture of in-person and online interviews. Discussions were facilitated by the researchers who had experience in crosscultural working and qualitative methods. All focus group sessions and interviews were audio-recorded and transcribed verbatim, before being coded in NVivo and data analysis applied. Across the 3-year period, a total of eight different focus groups containing a minimum of three participants during each session were conducted. --- Design The analytic process was driven by a CRT conceptual framework, which informed the development of knowledge constructed through interactions between the researcher and participants. In terms of positionality, the researcher had lived experience both as a Black academic and students within higher education. It is recognised that associated insider/outsider statuses can be both a benefit and a limitation of the research process , and thus the role of the reflexive researcher becomes especially important here. For example, both researchers' sharing lived experiences of racism could display bias and/or racialised flashbacks, affecting the procedure stage illustrated in the previous section. It is undeniable researcher bias affects data validity, and thus to limit our biases, the researchers kept their subjectivity to not conform to one truth supporting different perceptions. Additionally, countering racialised flashbacks was considered, i.e. the researchers personal experiences of diverse forms of racism coinciding with participants perceptions during the interviews. Thus, self-reflective notes and joint-researcher communication was used to protect mental health, preventing any responsive emotions during the interview processes. The interview design was informed by a critical race-grounded methodology process that captured the experiences of Black staff and students during the pandemic, including the concurrent Black Lives Matter movement. A grounded theory rejects the notion of positivist paradigms and methods taking a deductive approach to research creating an 'apartheid of knowledge' , which results in reoccurring narratives of deficit approaches towards people racialised as Black. This is congruent to CRT, where a CRT methodology centres the voices 6 New University, synonymous with Post-1992 University or modern university, is a former polytechnic or central institution in the UK that was given university status through the Further and Higher Education Act 1992 or an institution that has been granted university status since 1992 without receiving a royal charter. . 7 1 The Ivy League consists of eight private research-intensive universities in the North East of the USA. These universities regularly score highly in league tables based on research and teaching. The Ivy League are the most selective universities in the world . 8 Historically black colleges and universities are institutions of higher education in the USA that were established before the Civil Rights Act of 1964 with the intention of primarily serving the African-American community. 9 There are 146 institutions that are classified as 'R1: Doctoral Universities -Very high research activity' in the Carnegie Classification of Institutions of Higher Education as of the 2021 update. of groups who are racially marginalised to challenge these constructions from traditional orthodoxies aka Eurocentric epistemologies co-constructing racism in the form of White supremacy. In sum, this means while CRT was used conceptually to guide the structure of the interview schedule, the amalgamation of grounded theory was flexible enough to allow participants to divulge their stories. Another methodological point of significant importance was the fact that both researchers were Black men and some of the key findings in the forthcoming discussions concerned the gendered implications of racism and COVID-19 for Black women in particular. As well as contending with the ways in which lived experiences of racism, and especially anti-Black racism, would be an unavoidable aspect of both the way we conducted the research and our interpretation of the data, it was essential for our CRT framework to reflexively contend with race, gender and power . This required the researchers to, firstly, consider how they engaged with the participants and, secondly, how they interpreted the Black women's reflections concerning how institutions had disproportionality treated them in comparison to some of the Black men's narratives. While this paper contributes to scholarly interventions on the particularities of Blackness, racism and COVID-19, the researchers remain cognisant of how the intersection of gender should not be understood as supplementary to the findings, but rather, as an always already symptom of what bell Hooks' contends as a key pillar of white supremacist capitalist patriarchy which infiltrates all forms of racism . In this way, as custodians of anti-racist endeavour, it becomes our scholarly duty, as well as being integral to our position as Black men to be conscious that when discussing and researching the treatment of Black women in the workplace, we keep hooks' note on how 'interlocking systems of classism, racism, and sexism work to keep women exploited and oppressed' . These intersectional and gendered structures poignantly highlighted by hooks remained that at the forefront of how we engaged with the participants and the data we interpreted on the connection between COVID-19 and racism in universities. --- Findings and discussion Three themes were identified through analysis. The first theme, pandemic within a pandemic, had two sub-themes around racial traumatisation and loneliness and isolation. A second theme emerged, job precarity, alongside a third theme, exploitation of Black labour. The key tenet of CRT incorporated in the aforementioned themes is concerned with asserting that racism as 'ordinary, not aberrational'. Thus though the discussions are located in the context of universities, the analytical framework demonstrates that racism is 'normal' and the living everyday experience . It is our contention through CRT-and connecting individual narrative with the aforementioned global racist structuring of society-that COVID-19 exacerbated all forms of racisms and, in particular, anti-black racism. --- Pandemic within a pandemic There was a strong theme around racism and COVID-19 representing a 'pandemic within a pandemic' for Black people . For many participants, it was impossible to separate the impact of anti-Black racism, which was especially illuminated by the Black Lives Matter movement, and the impact of COVID-19, which was disproportionate among Black communities: The impact has been difficult, as a Black member of faculty my experience has obviously been compounded by the events of George Floyd and BLM and a President10 that had no care for Black lives. For participants three and thirteen, both Black US staff, experiences of racism during the pandemic were not only magnified through COVID-19 and the murder of George Floyd, but also within the White House. President Trump compounded the US racial divide, with both participants experiencing an acute disregard for Black lives. The pandemic was hard. In the US, we were up against it on several fronts: we were being failed terribly by a xenophobic President 5 who was whipping up as many racial divisions as possible. All of this in the midst of the George Floyd murder and the disregard by the President for his life. Former President Trump was a percolating theme for US participants as it was continuously perceived that his actions played an important role in their experiences during the pandemic. For the US participants, the pandemic was exacerbated by the lack of involvement from the government which clearly had demonstrable effects on their sense of self . As the governmental system reverted from a true engagement with anti-racist policies, it instead constructed racist policies . This highlights the systems connectivity with an original CRT doctrine 'white dominance' where 'white supremacy', i.e. 'everyday mundane actions and policies that shape the world in the interests of White people' impact communities who are racially marginalised. Moreover, participant perceptions also illustrate how White dominance creates discrepancies in the Black and White experience during the pandemic . For example, as reflected in the comments from participant nine, for Black people, living in a world designed by and for White communities requires constant negotiation and navigation. While their White counterparts were dealing with one pandemic, Black people were dealing with two: Negotiating the pandemic has been very difficult. As a Black person it's highlighted even more the blatant disregard for Black Lives. Yosso identifies navigational capital as ethnic minority students' skills and abilities to traverse social and educational institutions, which are often unsupportive or hostile environments that neglect and were not created with communities of colour in mind. This additional resource, which beneficiaries of 'white supremacy' are neither required to develop nor value , speaks to the 'Black tax' and burdens described both by Walton et al. and participant eighteen: During the pandemic seeing the rise of the Black Lives Matter movement globally, it makes you sit and think about the inequality we suffer as people of colour -includ-ing Covid. It weighs heavy on you after a while. This comment reflects wider research naming racism as a co-factor to COVID-19 and the acute awareness that the Black communities' experiences have been worse as a result. As well as pointing to the centrality of racism, which for many of the Black participants underscored their experiences of the pandemic, the participant also speaks to a sub-theme regarding the reinforcement of Black suffering, i.e. structured emotional consequences at the hands of racism. --- Racial traumatisation This sub-theme theme was around the impact of racism on mental health during the pandemic, and specifically how the murder of George Floyd and subsequent Black Lives Matter movement was linked to feelings of trauma and suffering, particularly among student participants. For young Black men, death was represented not just in rising and disproportionate COVID-19 figures , but in policing, which was the leading cause of death in young Black men in America : A big thing for me during Covid was trying to access mental health support through my university and when I tried to explain the impact that George Floyd murder had on me, like, being a triggering incident for my own personal experiences of racism, they [the counsellors] dismissed it as me being hypersensitive'. This participant's experience is consistent with the findings of a large US study demonstrating that police killings of unarmed Black Americans have adverse effects on mental health among Black Americans . Their account also reflects existing research highlighting the challenges Black students have accessing mental health services . Black men are seen as stronger , more impervious to pain , and as such are undertreated across both mental and physical health spheres. Both are indicators of racism and patriarchal masculinity and point to a need for reform within current mental health and pastoral services. Given research into collective trauma , which the Black community experience both through ongoing racism and the COVID-19 pandemic , the experiences of student participants are perhaps unsurprising: The above and previous comments suggest the intersections of race and location illustrate the role George Floyd and Black Lives Matter had within the student sample during the pandemic. Despite participants residing in the UK, the George Floyd murder-in the USA-had an impact, and this coincides with P18 above mentioning the role and rise of Black Lives Matter 'globally'. Thus, racism can be experienced vicariously for Black participants also known as 'secondhand racism' , whereby Black people on the witnessing end of racial abuse are impacted by it too. This reference to the 'everyday' and 'daily' nature of racism, often referred to as 'death by a thousand cuts' , sits alongside how the participants understanding of the murder of George Floyd was a more 'deeply felt wound in Black communities that extends beyond the individuals that directly experience it…this type of collective trauma must be understood as an urgent public health crisis' . These repeated lacerations are linked in the wider literature to racial stress , racial trauma and racial battle fatigue , especially among African American students in higher education institutions . Ragland Woods and colleagues highlight how these experiences both compound and are compounded by feelings of isolation and disconnectivity, which also emerged as a subtheme in the current study. --- Loneliness and isolation A second sub-theme emerged around feelings of loneliness and isolation. This is consistent with wider research that student and ethnic minority groups were at higher risk of being lonely during the UK lockdown, although the latter were already considered at greater risk prior to the pandemic . Participant five describes the challenge these feelings brought with them: It was really difficult, I felt a sense of loneliness that was really difficult…and disconnectivity. For participant nine, this difficulty was directly linked to disconnection from the Black student community: The pandemic was a very difficult period. I missed the sense of community that being around other Black people. I entered University as a mature student, I'd say this really impacted in terms of my mental health and what I was achieving. We are communal beings, and a central tenet of Black resilience is in what Yosso describes as community cultural wealth . The lack of access to support and community highlighted by participant nine, a Black woman, reaffirms the question asked by Pennant: 'who's checkin' for Black girls and women in the pandemic within a pandemic?' . This is important when considering ways in which to increase support for Black students, particularly Black women who typically receive less support from the educational systems around them . The intersections between race and gender play a pivotal role in the pandemic within a pandemic as the system is especially susceptible to neglecting the lives of Black women. Having access to supportive relationships not only underpinned participant nine's mental wellbeing, but also her productivity. This is also echoed in the experience of a Black female staff member: It gets a bit lonely sometimes. It feels like a lot is on you, you know. It's a lot of pressure to keep it up…sometimes you feel like you can't. But you have to, especially in these times, where you don't know if you're coming or going. Participant ten associates her loneliness with pressure, and a sense that she must keep going regardless. For Black staff, who may not know if they are 'coming or going', perhaps due to job precarity or fear of reprisal or exclusion, stopping or slowing down might be seen as privilege not afforded to them. This speaks to the type of loneliness experienced through a lack of care from others, as well as an emergent theme around job precarity and how this may be compounded in 'these uncertain times'. The participant testimonies highlight that the pandemic within a pandemic is conditioned and maintained by the structures of whiteness that marginalise the lives of Black people both physically and psychologically. Through the neglect of the health and wellbeing of Black students and staff in higher education, the ongoing backlash at both governmental and institutional level has created a perfect storm of marginality underpinned by a global emergency and the ongoing generation of anti-Black racism . Therefore, the theme 'pandemic within a pandemic' is underpinned by the normalisation of white dominance and anti-Black racism, where the government and higher education institutions reinforce a racist system that creates negative emotional consequences for Black lives . --- Precarious employment Black staff in higher education institutions relayed their anxieties around job insecurity, including the uncertainty around what this could mean for them: The pandemic has been very difficult. I was on a temporary contract, so my thoughts also gravitated towards the economic impact of Covid and where this would leave me professionally. Black people in the UK are twice as likely to be employed on zero-hour contracts . The associated uncertainty, as described by participant four, feels even more compounded when factoring in national economic crisis and global recession . It is unsurprising that job insecurity interrelated with wellbeing, as was the case with a US participant: I found the pandemic really debilitating, not only did I lose my job but my mental health was also severely impacted. This for me was extremely difficult given how long it had taken me to gain employment in academia. My redundancy was because of the pandemic and I guessed what I noticed at my institution and the sector my generally was that a lot of Black people, people of colour were the ones being made redundant. Importantly, this suggests that redundancy practices are experienced as a form of institutional racism. This speaks to both the centrality of racism and whiteness as property , rendering Black staff feeling more vulnerable and more disposable. In more juxtaposing narratives, participant blamed the pandemic for their adverse circumstances but are absent in blaming their racist institution. Again, an interpretation could be the camouflage of institutional racism and how it can use the pandemic as an excuse to conduct racist practices by threatening their job security. Additionally, and similarly to students, gender discrepancies amongst Black staff were evident. Participant four fears losing his job, while participant eleven was made redundant. Here, we play close attention to the tenants of CRT and intersectionality in foregrounding the gendered implications of race and power for Black women in particular . The race-gender element has been shown to have diverse impact amongst Black staff supporting studies and statistics emphasising the difficulties for the roles of Black women in academia . These patterns of oppression reinforce Black people's having to work 'twice as hard to get half as far' , paving the way for the exploitative practices that are central to Black staffs' experiences within the UK and US Academy . --- Exploitation of Black labour A strong theme that emerged related to participants' experiences of exploitation, specifically staff: The impact has been increased workload and for most Black members of staff a feeling of exploited labour during these very difficult times. Participant sixteen revealed exploitative practices as a shared experience among the Black community, which replicates the findings of a recent UK study . Experiences of exploitation were well-documented in the literature prior to the pandemic. Tuitt and Stewart argue that US higher education institutions are designed on the premise of plantation politics and resemble entrenched colonial systems and practices of exploitation and domination. The exploitation of Black people in the academy is not only violent, it's normal, to the point that none even questions it'. This comment reflects the permanence and normalcy of racism in the US education system, and how little remedial or reparative action is taken. Participant fifteen's use of the term 'violence' also speaks to the significant harm caused by these practices, the cuts, wounds and lacerations discussed earlier taking yet another structural form. For participants in the present study, their experiences of exploitation are confounded by the pandemic and concurrent Black Lives Matter movement. Where the spotlight on racism following the murder of George Floyd brought with it an increased global awareness of the plight of Black communities, the additional labour required to elevate and respond was outsourced to Black staff in the Academy: I became the person of colour that was designated to mobilise and complete anything to do with race equality. My labour was not acknowledged and I was positioned as being the person that should lead this without any resource, time or support. This comment subtly illuminates an entrenched pattern of white fragility , which sits on the other side of the proverbial coin to the notions of Black strength, invulnerability and invincibility described earlier . From these myths arise the expectation that Black people must care for, soothe, and protect White people ; historically, this was most visible on the plantations but continues systemically today, as reflected in participant experiences and the wider literature. In essence, to uphold this ideology , the burden of racism, and the labour of antiracism, must fall to Black staff: As a Black academic member of staff, I felt that the burden of the university's objective to become anti-racist fell squarely upon me. This was not only physically crushing but also emotionally crushing. There was this exploitative expectation that as a Black member of staff I would relish the opportunity to have my workload added to without any concession'. I found this period [pandemic] deeply stressful; the George Floyd murder actually created more work for me as my institution placed a lot of pressure on me as the only Black member of faculty to deliver anti-racist interventions and decolonising things… it was a lot and very stressful. The experiences of participants thirteen and two highlight the emotional and physical toll of racism, harking back once again to the literature around racial stress and trauma . Not mentioned in participants' comments are the silent beneficiaries, their White colleagues, who remain invisible and thus weightless . Within the previous two themes, there is a familiarity amongst Black staff and students which suggest that both precarious employment and the exploitation of Black labour are underpinned by a very particular form of anti-black racism dependant on the structures that allow for race or the racialisation of Blackness to be hidden by Whiteness . It seems pertinent to conceptualise that Whiteness is not solely used to normalise Eurocentric culture and its predominant representation in literature, which camouflages processes of Black exploitation and White passivity. It is therefore important that, as well as counter-storytelling, dominant discourses are challenged in action, including the prioritisation of Black care with a shift in labour away from Black staff and students in the Academy . --- Conclusion Utilising a CRT framework, this study qualitatively explored the impact of the pandemic on Black students and academic staff in UK and US higher education institutions. In doing so, it provides a counter-narrative to dominant racial discourse that platforms the voices of Black people as a distinct group. Firstly, as was evident from the dominant theme, Black communities experienced COVID-19 as a pandemic within a pandemic. Racism, 'the other pandemic' , underscored all participants' experiences, which for US staff and students in particular were compounded by the politics and presidency at the time. The use of students and staff together in the overall discussion and arguments presented a move within sociologies of education in higher education which takes seriously the importance of the academic lifecycle in unison. Significantly, the findings show the correlations and continuation of racism experienced across the academic landscape demonstrating the need for more connections to be made between service users and providers especially concerning global emergencies like COVID-19 and systemic racism. The concurrent murder of George Floyd, and subsequent Black Lives Matter movement, added to a collective trauma experienced by both US and UK participants, many of whom also described feelings of loneliness and isolation. Importantly, the results replicate existing findings from before the pandemic surrounding the exploitation of Black labour and precarious employment . These represent longstanding systemic issues within the Academy that require urgent redress. These themes and subthemes, which are discussed against the existing research landscape, serve to illuminate some of the structural and societal problems that Black students and academic staff experience unceasingly. Ultimately, as every racial group had adverse experiences under the 'Same Storm', i.e. COVID-19, those experiences are diverse requiring different levels of bespoke attention. The juxtaposition of experiences is evident as people racialised as Black remain in the eye of the storm, i.e. pandemic within a pandemic, underpinned by anti-black racism and the dominance of Whiteness . The subsequent neglect experienced by Black people aligns with their encounters which metaphorically situate them in stormy terrain. Therefore, purposeful recognition from institutions which take seriously how the inequitable structures of society infiltrate university cultures remains integral if Black people are to be given the space to locate the tools required to navigate the storm. --- Limitations As with all qualitative research, the findings in the current study are not generalisable; rather, the purpose of this study was to tell the stories of Black students and academic staff navigating the US and UK Academy during the COVID-19 pandemic. One of the benefits of CRT is that these voices, which are often marginalised from discourse or homogenised with other ethnic groups, are centred and brought to the fore. It is recognised that wider perspectives may have further informed this area of research, such as exploring the experiences of Black professional service staff as well as scholars. Conversely, narrower perspectives may also benefit further study, such as more detailed exploration of the experiences across the intersection, such as gender, class or disability. --- Implications and recommendations For Black students and academic staff, this study points to the complex and intersecting nature of two pandemics, COVID-19 and racism. While the present study initially set out to explore the COVID-19 pandemic, it was impossible to separate participants' experiences of racism. Together, they continue to reproduce greater inequalities for Black communities, reinforcing the need for each to be considered urgent public health crises. Specifically, both should be recognised as a type of collective, cultural trauma , as reflected in the current study and wider literature and particularly in respect to the murder of George Floyd . This recognition must take place at every level, with appropriate fiscal investment to ensure culturally responsive support is available. This might, for example, look like embedding in policy the recognition of racial trauma as a serious life event, as is the case at Goldsmiths, University of London. It could also look like reform within university mental and pastoral care provision, such as improving racial literacy among counsellors or engaging in more robust data monitoring practices. Perhaps also, universities and policymakers might consider abolition of precarious contracts , or local ringfencing of funding for equality and diversity resources. At the very least, senior leaders should be initiating courageous conversations , which start from the premise that racism is central and pandemic. Regardless of the approaches taken, there must be a clear shift in where the labour of this work sits, not just formally and in the more observable spaces, but in the quieter moments where everyday racism takes place. ---

The permanence of systemic racism in the UK and USA means that Black people are disadvantaged in myriad ways, including within the Academy. While the disproportionate impact of COVID-19, alongside the Black Lives Matter movement, has increased awareness of the challenges faced by Black communities, these issues remain, both in and beyond higher education. Furthermore, there is still a paucity of research individualising the experiences of Black people, who are often homogenised with other ethnic minority groups. This paper explores the impact of COVID-19 on UK and US Black students and academic staff, utilising a critical race theory (CRT) framework. Analysis revealed that Black students and staff experienced COVID-19 against the backdrop of racism as a "pandemic within a pandemic" (Laurencin and Walker, Cell Systems 11:9-10, 2020), including racial (re)traumatisation, loneliness and isolation. Other themes included precarious employment and exploitation. Recommendations are offered for penetrative interventions that can support Black students and staff in the wake of strained race relations neglecting their adverse experiences and a global pandemic.

I. INTRODUCTION Grassroots civic tech, or software for social change, is an emerging practice where people create and then use software to create positive change in their community, such as using citizen sensing to record evidence of air pollution to support environmental activism [1]. It is related to Free/Libre and Open Source Software movement and participants are diverse, calling themselves for example civic innovation, tech for good, civic crowdsourcing, free software, and community technology [2]. Like their names for the field, the participants' and groups' motivations and working practices are diverse. Civic tech itself is a wider field [3,4], with studies for example in: examining the role of information technologies in governance [5]; supporting participatory democracy practices of technology in society [6]; understanding the role of software in society [7]; developing agile methods and techniques to support software development in society [8,9] and, more recently architecting smart cities, open data with and for people [10,11]. However, civic tech software development processes from a systems perspective have been less often studied, especially when it comes to grounds-up development processes. Similarly, there have been studies on what drives individuals to participate and keep contributing in FLOSS [12,13] or tech for good [14], but less analysis on how volunteer motivations interact once the development process is ongoing. We propose that this is an open question that is worth addressing, since development processes in action [15,16] evolve based on team needs. A study on motivations of a project in action can reveal how different volunteer motivations can be negotiated towards a compromise. In this paper, we apply Engeström's expanded activity theory [17] as a theoretical lens to analyze motivations, how they relate to for example group goals or development tool supported processes, and what contradictions emerge when a civic tech group's development process is in progress. The research case is based on the activities of Code from Ireland civic tech group from 2018 to 2019, with the group structure described in more detail in [18]. During the case study period, Code for Ireland was engaged in two projects: Transparent Water for publishing data on water quality and Finding Vacant Homes for identifying residences that could alleviate the Irish housing crisis. More specifically, our research question is: What motivation, motive, or goal contradictions emerge in a grassroots civic tech project in action? In the following Section 2, we detail related research on grassroots civic tech and motivations. We continue then in Section 3, about case study and analysis. Findings are presented in Section 4 and conclusions in Section 5. [13]. They found that key motivational factors were social aspects and intrinsic motivations, such as fun, learning and altruism. Similarly, Barcomb et al. [12] found that social norms, satisfaction, and community commitment were supporting factors in episodic volunteers. Coelho et al. [19] who studied motivations of core developers mainly had similar results in friendly community and opportunity to engage in volunteer work. By contrast, core developers were also motivated by technical quality, such as improving projects they use, lack of project complexity, and lack of complex or buggy code. Huang et al. [14] studied specifically FLOSS for social good project participant motivations and found that they are more motivated by project impact, such as the societal issue being addressed, and who are the project owners. Community and social characteristics were also important. Pern and Kitchin [20] analyze frictions from a broader social and cultural systems perspective in civic tech development, using an earlier Irish case study as a data source. They found that complex urban issues can have only partial or imperfect solutions, that can in turn lead to or reveal further complications. The main frictions Pern and Kitchin discuss were stakeholder engagement, shifting goals, and varying commitment from participating developers. --- III. RESEARCH CONTEXT AND ANALYSIS The case we studied is focused on a grassroots group creating civic tech, Code for Ireland, which describes its mission as "developing innovative and sustainable solutions to real-world problems faced by communities across Ireland, by fostering collaboration with civic-minded individuals, businesses and public sector organizations." The research and data collection occurred during 2018 and 2019. The process is summarized in this Section and is reported in further detail in an earlier study on the group [18]. Case study data includes interviews, development activity logs from GitHub, and an ethnographic study with a researcher embedded with the group for six months in the first half of 2018. Of these, the main data source was semi-structured interviews with six key group contributors. The interviewees included the community leader, a project manager, an open data advocate, and several programmers. The interviews were recorded, transcribed, and analyzed with qualitative data analysis software. The interview guide is archived in the Zenodo repository 1 . Other collected data was used to contextualize the analysis findings. Other related research on the group has been performed by Perng and Kitchin [20], who discussed Code for Ireland from a civic and cultural perspective and defined Code for Ireland as a civic hacking group, where "civic hacking binds together elements of civic innovation and computer hacking, with citizens quickly and collaboratively developing technological solutions" [20, pp. 2]. --- A. Analysis method To accomplish our goals, we framed our research as an interpretive case study [21,22], using an iterative qualitative data analysis process aiming towards increasing abstraction [23,24]. Case study research was found to be the most suitable for the purposes of this research. It is a qualitative approach in which the investigator explores a bounded system over time, through detailed, in-depth data collection involving multiple sources of information, and reports a case description and case-based themes [21,22,25]. Initial codebook for motivations was based on the work of Barcomb et al. [12], Silva et al. [26], and Gerosa et al. [13]. Novel goals and motives were coded inductively when encountered. For rigour in our research process, we used a selection of techniques by Lincoln and Guba [27] for qualitative research. We used prolonged engagement to get rich data from the context, used referential adequacy in the transcription and coding process, and used peer debriefing to get additional neutral viewpoints into data analysis. --- B. Theoretical lens From an activity-theoretical perspective, human life is fundamentally rooted in participation in activites that are oriented towards objects [28]. These activities are driven by their own or collective purposes, mediated by social rules, shared with a community, and facilitated by tools [29]. For this paper, we use Engeström's cultural-historical activity theory [17], which distinguishes short-lived goal-directed actions and long-term activity systems [29]. This version is suitable for the study of work and technologies [30] and has been used in IS [31], software engineering [32], and HCI [33]. In Engeström's cultural-historical activity theory [17], activity systems are depicted as a set of mediation triangles, such as in Fig. 1. It displays the subject, object, and facilitating tools. The activity is moderated by rules, community, and the agreed division of labour between actors. Another key concept within activity systems are disturbances and contradictions. While rules, objects and motives give the system stability, they are often contradictory internally or between actors, and also tilt the system towards increasing instability. An activity system therefore constantly evolves towards a new balance based on internal contradictions in what Engeström has named "cycles of expansive learning." Contradictions can occur inside the key concepts or between them. Based on the distances, the contradictions are labeled as follows [17]: 1) primary , 2) secondary , 3) tertiary , or 4) quarternary . When applying Engeström's cultural-historical activity theory as a theoretical lens, we use the diagram representation to contextualize the motives and contradiction levels to analyze where the contradictions in motives occur. In the following sections where we present the analysis, we use terminology from the theory, which discusses actions, outcomes, and motives, instead of the adjacent word motivation. In this paper, we define motivation as a big picture "energization and direction of behavior" [34]. By contrast, motive is a specific reason for taking an action. --- IV. FINDINGS In this Section, we present our qualitative data analysis outcomes step by step as we progress towards increasing level of abstraction, following lightweight qualitative data analysis practises by Saldana [23] and Miles et al. [24]. In the first step of analysis, open coding, we analyze the interviews and note any motivations and motives. In the second round of analysis, we extract activity descriptions from interviews and online materials, and relate the motives to activities using our theoretical lens. In the final step of analysis, where we generate a situated explanation of the case, we present levels of contradictions inside and between the activities. --- A. Discovering Motives in Case In this Section, we present the findings of open coding analysis where we discovered individual motivations and motives, and then performed a sorting exercise to group the discovered motivations into larger themes. Four most common themes are listed in Table I, in a descending order of frequency. --- B. Contextualizing Motives in Activities In the second step of analysis, we performed another iteration of open coding to detect activity system concepts in two first --- Motivation Representative example To have an practical impact on society "Okay, we're working on software projects. It's very different, it's something good for the community, it's something to give back." -P5 Commitment to community "But, really, I'll do whatever I can to try and keep the Code For Ireland going on." -P3 Learning and developing skills "I do at some point maybe see myself managing some people so getting some project management experience, those help as well." -P5 Enjoyment and fun "I think the reason for that is that they like the freedom of these projects because if you have worked in a big company like SAP or Facebook, you're pretty much told what to do." -P1 stages of the software project: planning and design. The stages were originally discovered and defined in [18]. Then, we used activity theory mediation triangles to contextualize and visualize the constituent interactions in activity systems of 1) project planning and 2) software design. Finally, we analyzed how the stated big picture motivations translate to individual activity system motives. Planning and design activity systems were selected for presentation in this paper due to them featuring key contradictions. Implementation and evaluation stage analyses were omitted from this short paper due to size and scope limitations. 1) Planning Activity: In the first activity system, the key motive for project participants is to negotiate software development practises, which current urban issue should be addressed, and work division between current projects. Used tools involve communication tools such Slack and WhatsApp, project management tools such as Trello, and meeting memos. The key actors are the community leader and project managers. The optimal outcome would be a project plan that all participants agree on. The activity system is visualized in Fig. 2. 2) Design Activity: In the second activity system, the key motive for a development group participants is to decide in more detail what the final software artefact should be like, and how to divide the work inside the group. Used tools include wire-framing, design sessions, and comparison between visual designs. The outcome of the activity system is a design document. The activity system is visualized in Fig. 3. --- C. Contradictions of Motives in Activities In the final stage of analysis, contradictions inside activity systems, between system concepts, and between activity systems were analyzed. Inside each activity system concept, participant motivations were mainly in alignment. Whether participant was motivated by impact, commitment to friends, learning, or fun, they agreed with the overall motives and what the activity system should accomplish. What caused most contradictions inside activity systems was how these goals should be accomplished: in Project participant envisioned better processes, such as co-design or better support in design. However, the current system had no capability to implement them, leading to tensions. --- Quarternary --- Design activity ⇔ programming activity The activity systems were only loosely coupled, programming activity system sometimes disregarding the outputs and motives of design activity. II. Between activity systems, two main contradictions were loose coupling between software engineering stages . --- D. Discussing the Findings Some of the largest contradictions emerged in communication and task management, where a phone-based instant messenger was used for communication and task tracking. A second major issue occurred in design and choosing the technical specification: Design went through several rapid iterations, was loosely coupled, and several competing (instead In our case study, we found that the big picture motivations of participants were in alignment and what differs is how participants want to achieve those goals. This follows expectations set by Huang et al. [14], where they mention that tech for good group membership can be self-selecting for aligned goals. Our findings indicate that a richer view into participant motives appears when examining a process in action in detail, even though earlier survey-based research indicated that in tech for good projects, motives are initially in alignment. We found that contradictions rose up in the details of decisionmaking and in varying levels of commitment. These included how to divide tasks, communication tools and details, how to select implementation details, and how to select the technical approach. Majority of the participants were highly motivated, but the volunteer nature of participants and the lack of specialized facilitators meant that good compromises were not achieved on all viewpoints. --- V. CONCLUSIONS In this paper, we analyzed what motivation, motive, or goal contradictions emerge in a grassroots civic tech project in action. The main contradictions occurred inside activity systems on details of implementation or between system motives, instead of big picture motivations. Two most significant contradictions involved planning, and converging on design and technical approaches. Our recommendations for practice are what perhaps professionals in established volunteering and non-government organisations already know: Have skilled and committed facilitators that can keep volunteers pointed in the same direction and sustain participant engagement. Based on our case study, it appears that it is essential to have: efficient communicators, shared task management tools, and a compromise on technologies that most participants can commit to. Finding an acceptable lowest common technology denominator is important, since participant skill levels in civic tech can vary [35,36]. We extend the body of knowledge in civic tech software engineering through two main contributions. First, we demonstrate activity theory can be used as a theoretical lens to evaluate software engineering methods in action [15,16] as evolving, inaction activity systems. Second, we demonstrate that examining participant motives together as an activity system can shed light on contradictions that arise during civic tech or tech for good development. The trade-off between contextual richness of a single case study and generalizability of a multiple case study points towards a potential limitation in our work. However, seminal literature [37,38] acknowledge single case study as a method of inquiry, especially when undertaking novel, exploratory research to capture the complexity of the context in which the case is situated . Looking forward, we advocate for more comparisons between different civic tech software engineering activity systems for broader findings and more generalizability.

Grassroots civic tech, or software for social change, is an emerging practice where people create and then use software to create positive change in their community. In this interpretive case study, we apply Engeström's expanded activity theory as a theoretical lens to analyze motivations, how they relate to for example group goals or development tool supported processes, and what contradictions emerge. Participants agreed on big picture motivations, such as learning new skills or improving the community. The main contradictions occurred inside activity systems on details of implementation or between system motives, instead of big picture motivations. Two most significant contradictions involved planning, and converging on design and technical approaches. These findings demonstrate the value of examining civic tech development processes as evolving activity systems.

Introduction Throughout human history, pornography has continued to be an issue of concern. Initially, it drew attention for its depictions of nudity and sexual acts and, more recently, for the pervasive violence in its content and the consequential societal impact . In recent decades, the availability of pornography on the internet has given rise to a remarkable surge in its popularity, attracting an everwidening audience . Indeed, its presence has now become normalized for most of Western societyparticularly among men-leading to an escalation in both supply and demand for pornographic content . However, this consumption continues to be shrouded in social taboos , particularly in the context of interactions between young people and adults . As a result, fostering critical thinking about pornography consumption has proven to be challenging . Furthermore, there are even studies suggesting a waning influence of universities teaching on critical thinking and a diminishing level of critical discourse among teaching staff concerning pornography . A pivotal step in addressing this issue could involve destigmatizing the subject within the educational community, although this still seems to be a distant goal . In this context, the present study aimed to analyze how pornography is represented by young people who will become educators and teachers in the future. --- Pornography Consumption Among Young People Concerning the younger demographic, recent Spanish research suggests that the age at which users start consuming pornography is currently 8 years . Moreover, according to some reports , more than 60% of Spanish adolescents between 14 and 17 years old tend to regularly watch pornography on the internet . One of the repercussions of pornography consumption among adolescents and young adults is the phenomenon of perceptual distortion. This concept refers to the formation or deformation of attitudes since the cognitive and emotional development of young individuals may not yet be sufficiently mature to contextualize and interpret what they see . More specifically, the content of mainstream pornography is predominantly biased toward catering to male gratification, a trend characterized by the objectification and commodification of women to satisfy male desires . Furthermore, this portrayal fosters a sexist relational model, wherein women are frequently reduced to the status of sexual objects available for the gratification of men and in which instances of both physical and verbal assault against women are disturbingly prevalent . Consequently, some researchers suggest that adolescents who habitually consume pornography are more likely to internalize and endorse rape myths, leading to the adoption of negative sexual values or a propensity toward sexual aggression . Therefore, this consumption of mainstream pornography also has an impact on their perception of sexual relationships and may directly affect their sexual activity . In addition, the consumption of pornography is more stigmatized among women as they may fear potential judgement or disapproval of their reported consumption . --- Pornography as an Educational Tool While comprehensive affective-sexual education has been in place for more than 50 years in some European countries, such as Sweden , this is unfortunately not the case in Spain. In this country, adolescents often rely on pornography as their primary source of education, given the absence of regulated affective-sexual education that could enable them to develop a critical perspective . Additionally, educators working with children, adolescents, and young people have expressed a growing concern about the effects of pornography on their students . It is important to note that, historically, educators have been ill-equipped to address this issue, primarily because it is considered a taboo subject and difficult to approach . Hence, several academics have also emphasized the importance of training future teachers on the issue of pornography . However, as a starting point, it is important to determine how future educators understand and relate to pornography . In this context, the present research draws on social representations theory ) to investigate how pornography is understood and represented by future educators. According to this theory, social representations are systems of values, ideas, and practices with two functions. First, they establish an order that allows people to navigate and comprehend the world they inhabit, and second, they enable communication between members of a community . This theoretical framework allows us to explore how future educators understand pornography based on both their knowledge and their daily social discourse concerning this issue . We understand their representations of pornography as serving three fundamental functions: a knowledge function, enabling understanding and communication; an orientation function, guiding behavior and practices; and a legitimization function, rationalizing positions and behavior . Therefore, the present study focused on university students of education. This particular demographic was chosen because they are young individuals, who are a primary target of pornography consumption. Additionally, they hold the potential to become the future educators of the next generation and can thus play a pivotal role in driving changes in sex education. Thus, the general goal of the present study was to explore the representations of pornography among these future educators, that is, students of Infant Education, Primary Education, and Social Education. The specific objectives of the study were as follows: • To describe, using a free association exercise, how pornography is represented by trainee teachers. • To identify whether the types of pornography or reasons for its consumption are also incorporated into these representations. • To explore whether these representations also include reflections on the potential impact of pornography on their affective-sexual lives. --- Methodology --- Instrument and Procedure The data for this study were collected online from January to May 2021 in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University of the Basque Country . The questionnaires were administered into two parts. First, we collected socio-demographic data including age , gender , the subject they were studying , and whether they had received affective-sexual education . Second, a free association exercise was administered based on the grid elaboration method to analyze the participants' social representations of pornography . This method allows for gaining a deeper insight into the shared representations of the participants when compared with interviews or other types of questionnaires as it elicits not only a cognitive but also a behavioral and affective representation of the issue under analysis. This methodology has previously been used to explore the shared representations of young people on a range of controversial and even taboo subjects . Specifically, participants were asked to write down the first four ideas that came to their minds when they thought about the word "pornography." Then, they had to write each word or idea in a box and fill in the four empty boxes. Subsequently, participants were asked to complete their responses by clarifying as thoroughly as possible the meaning of each of their ideas to gather further information and explanations about the elicited items. Again, there was no word limit in this exercise. This allowed us to obtain a complete explanation about each word or idea, which formed the basis of the subsequent analysis. --- Data Analysis Method The Reinert method using the Iramuteq software for lexical analysis ) was employed to analyze the text corpus. This method has frequently been used in the study of social representations , confirming that the results obtained agree with those of other methods used in this field of research . Two types of analysis were carried out using the Iramuteq software. The first was based on the Reinert method, and the second employed lexical similarity analysis. The use of the Reinert method using the Iramuteq software for lexical analysis eliminates problems of reliability and validity in text analysis as it is an automatic instrument . First, the software creates a dictionary of "whole words" . Then, the initial text corpus is broken down into segments that have the approximate length of a sentence or two . The corpus is analyzed in terms of the presence of whole words in the segments. The segments and reduced forms are then used to create a contingency table, which shows the distribution of vocabulary per segment. From this contingency table, the program generates a squared distance matrix, implying that two segments are close if they share some of the words analyzed . Subsequently, the software, following the Reinert method, runs a descending hierarchical cluster analysis on this distance matrix, which yields classes of segments that best differentiate the vocabulary. In so doing, this software assists in the interpretation of texts. It extracts sets of words that are referred to as classes, which co-occur and are best distinguished from other classes. The software does not name these classes but instead assigns each of them with a number. In essence, the software identifies the words and text segments with the highest chi-square values, that is, those words and text segments that best identify each class or idea that the participants have repeatedly mentioned. In accord with previous research using the Reinert method , the raw data were entered into the Iramuteq software, and the most significant items of vocabulary in each class were selected based on three criteria: an expected value of the word greater than 3; evidence of an association based on the chi-square statistic, tested against the class ; df = 1) and; the word appearing mainly in that class with a frequency of 50% or more. The software also determined which text segments were associated with each class or group of words and classified them according to their chi-square value. In this study, the text segments with the most significant chisquares of each class were recorded. Once these "lexical universes" were identified, they were associated with "passive" variables . In the present case, the passive variables were gender and whether they had received affective-sexual education. Consequently, the analyst obtained a series of classes composed of typical words and typical text segments with the highest chi-square values . This provided the basis for "interpreting" the classes as lexical worlds. The Reinert method produces statistical, transparent, and reproducible data until the final point of interpretation, where the analyst then assigns a label. Finally, the researchers will give a title to the group of words and text segments grouped by the software . In this final phase of the study, we used a systematic process to create the labels or titles of each class. During this process, two of the researchers independently named each class based on the words and associated quotations, after which a third researcher created a final label that was approved by all three researchers. Second, the software also conducted a lexical similarity analysis. This type of analysis looks at the corpus as a whole, regardless of whether the responses were from one participant or another. It considers that the more times two elements are treated in the same way, the closer they will be in the representational structure to the corresponding object. To do this, the software identifies the cooccurrences between words according to their connections in the text, helping to identify the structure of the content of the textual corpus by generating a visualization in graphic form, which illustrates the content of the social representation of the object studied and its internal organization, its common elements, and specificities . --- Results First, to analyze the main discourses produced by the participants, the response corpus was entered into the Iramuteq software. This allowed us to clarify which terms were used to represent the concept of "pornography." The full corpus contained 22,595 words, of which 2385 were unique. --- Reinert Method Results The Reinert method's descending hierarchical analysis divided the corpus into 567 segments and six classes. The results of this analysis are observed in Fig. 1. Fig. 1 The hierarchical clustering dendrogram of the free association exercise, showing the most frequent words and those with the greatest association χ 2 , p < 0.001 extracted by the Reinert method The analysis identified the main ideas held by the participants regarding pornography, elicited through the free association procedure. Each issue or idea is represented by a set of typical words and text segments, referred to as a class. The results revealed six different classes. Four of these classes refer to the reasons why participants distance themselves from pornography or cause them to evaluate it with more negative connotations and were named as the following: "Violence and sexual exploitation," "Denigration of women," "Removed from real life," and "Pornography bodies." However, the other two classes are much more personal and explain young people's relationship with pornography. Among the classes that express the notion of being distanced from pornography, the first of these is "Violence and sexual exploitation," which emerged with a weight of 21.41% and was more frequently mentioned by women . Within this class, it can be observed how pornography has been defined as an industry that carries out sexual exploitation and is sexist, abusive, dominant, and mistreats its actresses. The characteristic text segments of this class should help to provide a context for these words. Thus, the following are the most significant quotations within this class: "Abuse, sexism, manipulation, in many of the videos women are forced to do certain things. Pornography promotes sexist violence; they want you to believe that this is reality when it is only fiction. People who watch pornography without any sex education tend to show the same behaviors." ; "I think that many of the people who work in pornography are being raped all the time. There will be many things they do out of obligation, but the current reality of pornography is one of massive violence towards women showing violent sexual scenes." ; "It makes me sick to see how a man treats a woman like this and she seems to like what he does. Women are sexually exploited, and the pornography industry is very sexist." . Second, and continuing with the reasons why participants reject pornography, the "Denigration of women" class emerged with a weight of 24.74%. This class was also more frequently mentioned by women . Within this class, pornography is characterized as a male-centered cinematic genre that denigrates and disrespects women, often displaying aggression and sexualization through the abuse of power. The following are some of the most significant quotations of this class: "Many of the freely recorded videos usually represent scenes in which a woman is raped or abused, and I am disgusted because I cannot understand how there can be sick minds that enjoy violence." ; "Porn revolves around men's pleasure and on many occasions, women are denigrated. Nothing you can see is real, there is a lot of violence, especially towards women, and there are very strong scenes." ; "On many occasions women are forced to maintain relations during prolonged periods of time, with different people and even with several people at the same time. Pornography is made for male consumption; the woman appears as a completely passive subject and many times is viewed as an object to be treated in a violent and abrupt manner. Besides, the vast majority of acts end when the man or the dominant person reaches climax regardless of whether the other person has gained pleasure." . The third class, named "Removed from real life" emerged with a weight of 10.36%. This class emphasizes the stark disparity between what is portrayed in pornographic movies and real-life experiences, noting the exaggerated and largely unrealistic nature of pornographic content. Additionally, some participants pointed out that attempting to replicate what is seen in these movies can be risky and potentially harmful in real life. The following are some of the most significant quotations of this class: "Porn is a lie because it's not real, it's all a role and it's exaggerated to be seen more. This disgusts me because it is unreal and creates false expectations." ; "It is unreal, young people make movies in their heads and then try to bring to real life those violent actions learned from pornography." ; "False, stereotyped, forced, unnatural. It is not a real thing that happens in people's lives the way they show it. It always follows the same pattern, it's all fake, it's not real sensations, it's sex that doesn't happen in people's lives and they sell it to us as if it were real. This affects the image of sex that people can create." . The final class to express a rejection of pornography was named "Pornography bodies" . This class argues that the bodies of pornographic actors and actresses are disproportionate and unrealistic and, in some cases, can cause individuals to develop certain complexes. The most significant text quotations of this class are the following: "Most women's bodies follow the same pattern, which is what attracts the male population. Only sexual attributes are highlighted, they are not looking for intelligent educated people or those capable of working at NASA, they are looking for sexualized bodies that sell." ; "Pornography presents totally perfect bodies. Perfect for the canons of beauty, which prevail in this society: depilated of certain sizes or shapes ." ; "The bodies have to be with exaggerated penises and breasts of exaggerated sizes and small vaginas. Therefore, many people end up creating complexes for not achieving what is shown in those videos or for not having a body like the one in those videos." . In contrast, the last two classes explain young people's relationship with pornography. The first of these classes was labeled as "Why do we consume pornography?" and emerged with a weight of 21.14%. This class was significantly associated with men and individuals who have received affective-sexual education . In this class, the participants indicate the two main reasons why they consume pornography. First, they state that they consume pornography to masturbate, and second, they use it to learn about sex, given the inadequate provision of affective-sexual education. Some of the most significant quotations of this class are the following: "When I watch pornography it gives me pleasure even though I know that it shows a wrong image of healthy sex. However, in pornography you can see many types of sex and that is why it is a tool that I use to masturbate." ; "Relaxing, exciting, quality pleasure, while watching videos I don't worry about other things, it makes me horny. Normally I relate it to masturbation." ; "I have always watched pornography to masturbate, and I combine porn with masturbation a lot. It excites me to watch pornography and it gives me pleasure to masturbate watching pornography although I am aware that porn is not real and that there is a lot of money behind this industry." ; "When teenagers start to get interested in sex and want to investigate, they consume pornography, either to satisfy their sexual desire or to learn. But they learn little or rather they learn badly. But that's normal, who are they going to talk to about explicit sex-with their parents? At school someone comes to explain everything but it's not a place where you can talk about it either" ; "Nowadays it is the only sex education we have. Because what is done in schools is a lecture that is useless. Porn is probably often the main source of learning about different subjects." . The final class was labeled "Influence of pornography on our sexual relations" with a weight of 13.32%. This class was significantly associated with individuals who have received affective-sexual education and indicates how the consumption of pornography shapes the sexual experiences of young people. The most significant text quotations of this class are the following: "After all, pornography as sex is pleasurable. You can discover new things in porn and then you can explore the new things you have discovered and put them into practice in your life." ; "I like what I see, and at the same time I dislike it, but then my relationships with my partner are different." ; "In the society in which we live, pornography is frowned upon, but then we consume OnlyFans and there are pictures and videos." ; "I look for other types of empowered women's porn and watching it makes me feel more comfortable." . --- Lexical Similarity Analysis Second, a lexical similarity analysis was conducted to generate an image to represent the co-occurrences between all the words in the corpus beyond their division into classes. The idea was to analyze how the words of the corpus were interconnected on a common plane. For this purpose, the lexical similarity analysis was carried out only with the words having a frequency greater than 20, the results of which are displayed in Fig. 2. The similarity analysis revealed that the corpus is structured around different nuclei. First, at the uppermost part of the figure, the word "sex" appears, associated with penetration and the role of pornography in masturbation, which differs from real life. Second, as we move to the center of the figure, pornography emerges as something that makes money through the Internet and is consumed globally in any location and at any time. Additionally, the OnlyFans phenomenon is also linked to this concept . Moving on to the nucleus composed of the word "woman," this is related to the violence and sexism suffered in pornographic videos. In addition, the female bodies are also questioned. In the lower part of the figure, we also see that pornography is related to sexual relations, sex education, and sexual exploitation. Finally, and linked to the nucleus of pornography, we can observe a complex emotional response with various emotions mentioned by young people such as pleasure, fear, disgust, or guilt. --- Discussion This study set out to examine the primary representational patterns elicited by pornography in young people who will be future educational professionals. The results of this research offer invaluable insights into identifying how pornography is represented by trainee teachers, whether the types of pornography or reasons for its consumption are also incorporated into these representations and to explore whether these representations also include reflections on the potential impact of pornography on their affective-sexual lives. One of the most significant findings of this study is the ambivalent relationship that young future educators have with pornography. On the one hand, they maintain a distant and critical perspective, associating pornography with violence, sexual exploitation, denigration of women, and disconnection from real life and real bodies. However, on the other hand, they utilize it for both masturbation and as a source of pleasure and sexual education . Therefore, this ambivalence in the social representation of pornography will be key since it will shape the understanding and communication of this subject, influence behaviors and practices, and impact its legitimization or delegitimization . To begin with, following the representational patterns extracted from the cluster analysis, it becomes apparent that the young participants in this study perceive pornography as conflicting with many of their core values. This distant, and at times negative, assessment is underpinned by several ideas rooted in two main arguments. First, both men and women express the belief that pornography deviates significantly from real-life sexual relations by presenting an exaggerated portrayal that includes unrealistic bodies. These bodies often feature disproportionate shapes and sizes of penises, vulvas, and breasts. The discourse surrounding the "authenticity" of pornography among viewers has been explored previously, concluding that viewers can, and should, make judgements about what is and is not real in pornography, and that this is an integral part of the experience of viewing pornography . Second, being unreal or fictional is not the only factor that leads young people to express a distant representation of pornography. A further important reason for this Fig. 2 Results of the lexical similarity analysis with words that emerged with a frequency higher than 20 distant-and sometimes negative-evaluation is that participants in this study, particularly women, contend that a significant quantity of pornographic content is characterized by violence and sexism . In support of this perspective, some participants highlight that, in addition to demeaning women, the pornography industry also sexually exploits women . To further corroborate this argument, some participants express the view that almost all pornographic content is centered around providing pleasure to men, often featuring scenarios where women are raped by one or more men who do not respect their right to say no . The notion that pornography influences many people to disregard the importance of respecting consent is well established in both academic research and public discourse following high-profile rape cases . Along these lines, some participants in this study, who are aspiring educators, express concerns about this potential risk. They advocate for the need to generate debate and implement concrete measures aimed at strengthening affective-sexual education for children and young people. The aim is to help them understand the importance of sexual consent and the notion that sex should be joyful, equal, and free . Indeed, alternative approaches within the industry, such as feminist pornography , seem to have limited reach among this young population. This might suggest that such content, characterized by a distinct representation of pleasure and the role of women, is predominantly consumed by an older audience with greater purchasing power and have not yet had an impact on the younger generation. Previous research has suggested that young people often simplify their consumption of mainstream pornography, characterizing it from a distanced and risky standpoint, in contrast to the more nuanced considerations of pleasure and arousal that adults are allowed to make . However, our findings indicate an association between such considerations and active pornography consumption among young individuals. Young people cite two main reasons for watching pornography: to masturbate and for educational purposes. Previous studies have identified pornography consumption as a source of information about sexual practices, bodies, sexual orientation, and desires, often catering to pre-existing or emerging fantasies and interests . Furthermore, along with the consumption of pornographic material, masturbation remains a significant sexual taboo . Several participants closely associate these two activities, with some even admitting that they cannot masturbate without watching pornography . It is important to note that while male participants talk more openly about this issue, we must recognize that both pornography consumption and masturbation are taboo subjects that are rarely discussed among women . Moreover, in line with previous research, it is also worth considering that boys tend to consume more pornography than girls . Many of the participants highlight the inadequate provision of affective-sexual education for young individuals, let alone a platform where sex can be talked about explicitly. They assert that sex is not discussed openly within either families or in the affective-sexual education that is offered in schools. Consequently, youngsters often turn to pornography, even though they are aware that it can perpetuate stereotypes, present unrealistic portrayals, and even promote violence . Thus, creating high-quality affective-sexual education represents a significant contemporary challenge . In Spain, the legal framework has been established to address this need, since the current Education Law has regulated the implementation of Affective-Sexual Education in compulsory education . In addition, our participants also explain how pornography can affect the sexual relationships or behaviors of young individuals. While some acknowledge its role as a learning resource , others express concerns that, due to pornography, many young people's sexual relationships prioritize male pleasure, with limited emphasis on female satisfaction . Some participants also state that while traditional pornography may be frowned upon, platforms such as OnlyFans are highly successful . Therefore, it becomes apparent that the impact of these new social networks on the emergence of fresh approaches to consuming or even creating pornography among young people warrants further investigation. Moreover, one of the most striking findings to emerge from our research is that those participants who have received affective-sexual education tend to speak more openly about their reasons for consuming pornography. This implies that although the notion of pornography being "bad" is prevalent among young people, they consume it nonetheless . Thus, our results suggest that affective-sexual education could encourage individuals to talk more openly and even reflect on why they consume pornography . Additionally, it is important to address empathy within the context of affective-sexual education, since previous studies have indicated that low levels of cognitive and affective empathy are associated with extensive pornography consumption and a greater prevalence of associated risk behaviors. For this reason, it is important to focus on empathic disconnection processes . In essence, the fact that pornography remains a taboo and stigmatized subject, it does not deter young people from its consumption, let alone promote thoughtful consumption . The similarity analysis also highlights the concerning widespread availability of pornography and at increasingly younger ages, underscoring the pressing need to urgently address this issue . Moreover, on an emotional level, our analysis reveals that the emotions linked to pornography consumption are, on the one hand, pleasure, but on the other hand, disgust, fear, and guilt, an inconsistency that has also been observed in other studies . Therefore, it would be interesting to analyze to what extent pornography is considered a "guilty pleasure," that is, a way of furtively satisfying oneself through taboo acts . This view of pornography, and potentially sex itself, perpetuates the legitimization of taboo and discourages open communication, an issue that needs to be addressed within the context of affective-sexual education. Finally, this research has several limitations. First, this work employed a cross-sectional design with a non-probabilistic sample recruited from a specific region-the north of Spain. Therefore, our conclusions cannot be generalized to any other society or context. Furthermore, the degree subjects being studied by our participants have a clear gender imbalance, and it would therefore be interesting to analyze students of subjects with a more balanced gender ratio to determine how they are influenced by their context. In addition, in future studies it would be advisable to examine the impact of the variables of gender identification and type of pornography consumed. Nonetheless, the strength of the present study is that it lays bare the voices of young people and reveals novel findings that have not been reported in previous studies. In conclusion, this research suggests that pornography serves as both a source of pleasure and a trigger of emotions such as disgust, fear, and guilt among young people. Indeed, the symbolic perceptions of pornography held by future educators are determined by their gender and the affectivesexual education they have received. Moreover, it is essential to further explore whether critical appraisals of pornography are rooted in social desirability or reflect a genuine capacity for critical analysis of this subject. Consequently, we stress the importance of implementing high-quality affective-sexual education rooted in emotional awareness and a gender perspective, not only for children but also for current and future educators. Furthermore, it is imperative that this affective-sexual education, aimed at young adults , begins by recognizing these ambivalent emotions and promoting the values of respect and self-empowerment, thereby dismantling societal taboos. In other words, we need to equip educators with the tools to address the issue of pornography with both adolescents and younger children. However, as a first step, it is necessary to encourage these future educators to reflect on the possible impact of pornography on young people in general, as well as on their own perceptions, while taking into account the gender perspective. --- Availability of Data and Material The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Code Availability Software application or custom code: Iramuteq. --- --- Author Contribution All authors contributed equally. Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This research was funded by KideOn Research Group of the Basque Government, ref.: IT1342-19 and UPV/EHU . --- Conflicts of Interest The authors declare no competing interests. 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://creativecommons.org/licenses/by/4.0/. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Introduction Although pornography has existed throughout human history, contemporary society finds itself immersed in a digital era that significantly facilitates its accessibility. This heightened ease of access particularly affects the perceptions and practices of sexuality among youngsters. In this context, the present study aimed to explore the representations of pornography among education students. Methods A total of 276 university students participated by completing a free association exercise based on the grid elaboration method (GEM). Results Future educators distance themselves from pornography, perceiving it as a medium that projects an unrealistic portrayal of reality. Moreover, participants also linked pornography to situations of violence, exploitation, and denigration of women. However, students also show a direct relationship with pornography, using it to both masturbate and learn about sex, particularly in the case of men. Educational students also recognize the influence of pornography on their own sexual life and knowledge. Finally, they report that pornography produces mixed emotions such as pleasure, disgust, fear, and guilt.This study highlights the urgent need to equip future education professionals with the necessary tools for reflective engagement with pornography and its consumption, paving the way for a more thoughtful and informed approach to educating on this subject.

Introduction Childhood obesity has increased dramatically in developing and developed countries. 1 Unhealthy diet and physical inactivity are two major risk factors for childhood obesity. 1 Caregivers may influence children's diet and physical activity making child care, defined as parental, informal or formal , a potentially important, modifiable factor in childhood obesity. In developed Western countries, one cross-sectional study 2 and three prospective cohort studies [3][4][5] have found informal rather than parental child care during infancy or early childhood to be associated with obesity, whereas associations were less clear for formal child care. 2,[5][6][7] It has also been suggested that the association of informal child care with obesity may be limited to more advantaged socio-economic groups. 5 In Western populations, both child care and childhood obesity are socially patterned, such that low socio-economic position is associated with both informal child care [7][8][9][10] and childhood obesity. 11,12 However, in a non-Western population , non-maternal care was not associated with obesity in young children, 13,14 but informal care was not distinguished from formal care. Examination of this question in another developed setting with a different social patterning of informal child care and childhood obesity can clarify whether the observed associations are socially confounded or are an attribute of informal child care where intervention could be justified. Moreover, it is not known whether it is the current exposure to informal child care that matters, the cumulative exposure or whether there are any critical periods when informal care has a long-lasting effect. Hong Kong has a recently developed, post-industrial service-based economy with a gross domestic product per head and social infrastructure comparable with Western Europe or North America. In Hong Kong, the labour participation rate for women aged 25-34 years increased from 47.8% to 78.7% over the past three decades. 15,16 Parents generally rely on informal child care provided by extended family or domestic helpers. 17 Childhood obesity is less clearly socially patterned than in Western settings. 18 Here, we took advantage of a large, contemporary, populationrepresentative Hong Kong Chinese birth cohort 'Children of 1997' to examine the association of child care at different ages with body mass index and the presence of overweight at $11 years. We also examined whether associations varied with sex or family SEP. --- Methods --- Source of data The Hong Kong 'Children of 1997' birth cohort is a population-representative Chinese birth cohort covering 88.0% of all births from 1 April 1997 to 31 May 1997. It has been described in detail elsewhere. 19 The study was initially established to investigate the effect of second-hand smoke exposure on infant health. 20 Families were recruited at the first post-natal visit to any of the 49 Maternal and Child Health Centres in Hong Kong, 20 which parents of all newborns are encouraged to attend for free postnatal care, developmental checks and vaccinations until the child reaches 5 years. Baseline characteristics obtained at recruitment included SEP, birth characteristics, infant feeding and second-hand smoke exposure, using a selfadministered questionnaire in Chinese. Follow-up at regular 'well-baby' visits to the MCHCs was continued until 18 months post-natally. In 2005, the birth cohort was resurrected and information enabling record linkage to routinely collected data was obtained. Lifetime routine height and weight measurements taken by the Department of Health at 'well-baby' or 'well-child' visits, public hospital discharges from the Hospital Authority and deaths from the Death Registry have since been retrieved. In early 2007, a programme was instituted to re-establish direct contact with the birth cohort families. It included repeated mailings to their last known addresses and schools, telephone calls and general publicity about the cohort including a press conference and a television programme. A postal survey was sent in July 2008, then re-sent a second and third time as necessary to non-respondents over the following 9 months. The survey, in Chinese , included questions on food frequency, activities, developmental progress and family and medical history. With each wave of data collection, any missing baseline data were updated and any discrepancies between waves reconciled. As shown in Figure 1, of the 8327 Chinese children in the birth cohort, by June 2010, 12 were known to have died, 22 had permanently withdrawn, 79 were known to have migrated and 281 were untraced, probably migrated, whereas 7933 remained living in Hong Kong. Of these 7933, 3682 returned Survey I. The 3682 respondents and 4251 non-respondents to Survey I were similar in sex, birth weight, birth order, mother's birthplace, weight status at $11 years of age and measures of baseline SEP as shown in Table 1. The Cohen effect sizes were small . 21 The study obtained ethical approval from the University of Hong Kong-Hospital Authority Hong Kong West Cluster, Joint Institutional Review Board and the Ethics Committee of the Department of Health, Government of the Hong Kong Special Administrative Region. Exposure Our primary exposures were child care at 6 months and at 3, 5 and 11 years, obtained from this question in Survey I: 'Who was your child's main caregiver at the following different ages: 6 months, 3 years, 5 years and now?' Our participants were $11 years old when Survey I was conducted. Responses were categorized as parental care or informal care, with the latter including care by grandparents, other family members or domestic helpers. The small number reporting formal care in a nursery or cre `che was excluded. Those reported to have more than one main caregiver were also categorized as 'informal care' making parental and informal care mutually exclusive in this study. --- Outcome The primary outcome was adiposity at $11 years proxied by age and sex-specific BMI z-score relative to the 2007 World Health Organization growth standard. 22 To facilitate comparison with other studies, we also defined childhood overweight as a BMI for age and sex corresponding to an adult BMI525 kg/m 2 using the International Obesity Task Force cut-offs. 23 Due to the relatively small number of obese children , overweight here included those who were obese. Not all height and weight measurements were at exactly 11 years of age, so we used the closest measurement between 9 and 12 years. We interpolated the WHO references and the IOTF cut-offs on a daily scale using the 'akima' package 24,25 in R . 26 Among the 7933 cohort members alive, living in Hong Kong and not withdrawn, 6782 children had a BMI measurement at $11 years of age, excluding children with formal child care. Of these, 52.5% had missing data for child care and 24.4% had missing data for mother's birthplace, birth weight, highest parental education, highest parental occupation or household income. Multiple imputation is the gold standard in this situation, because the assumptions required for validity of a complete case analysis are a superset of those required for multiple imputation. 27 We predicted missing values of these exposures and confounders based on a flexible additive regression model with predictive mean matching 28 incorporating data on the primary outcome , 29 exposures and other factors potentially associated with child care . We imputed missing values 10 times using the 'Hmisc' package in R 31 and analysed the 10 complete data sets separately. We summarized the results into single estimated b-coefficients with confidence intervals adjusted for missing data uncertainty. 27 We also performed a complete case analysis for comparison. --- Statistical analysis We used multivariable linear and logistic regression models to estimate the association of child care with BMI z-score and the presence of overweight at $11 years of age, adjusted for sex, mother's birthplace , birth weight z-score, highest parental education, highest parental occupation, monthly household income per head and the interaction of mother's birthplace and highest parental education. 18,32,33 Following Mishra et al., 34 we compared models representing child care in terms of a critical period or an accumulation of exposures . Preliminary analysis suggested that child care at older ages had most influence. We also assessed whether there was any additional contribution of child care at earlier ages than that of current child care, by adding exposures at previous ages sequentially. We used model fit Akaike Information Criterion to compare these different representations of child care. 28 A lower AIC indicates a better fitting model. We assessed whether the association of child care with the outcomes varied with sex or markers of SEP, from the heterogeneity across strata and the significance of interaction terms. We compared the statistical significance of the likelihood ratio test of the difference between models with and without interaction terms between child care and sex, highest parental education, occupation or income. Interaction terms with other relevant confounders were also included to ensure that any observed difference by sex or SEP was not due to confounding by other interactions with child care. 35 We used residual plots and Hosmer-Lemeshow-Cressie tests 36 to check the goodness-of-fit of linear and logistic regression models, respectively. Cohen effect sizes have three levels: 0.1 for small, 0.3 for medium and 0.5 for large. 21 Cohen effect sizes are calculated as ˇ 2 /p 0i ) where the sum is over the categories and p 0i is the proportion in the ith category in the respondents and p 1i is the proportion in the ith category in the non-respondents. --- Results Informal child care was quite common at 6 months , 3 years and 5 years but less so at 11 years . Table 2 shows the association of child care with socio-demographic characteristics before imputation using an available case analysis, i.e. pairwise deletion of observations with missing data. At younger ages, informal care was slightly more likely for boys than girls. At all ages, informal care was more likely for children with mothers born in Hong Kong. Highest parental education, occupation and monthly household income were all positively associated with informal care. Residual plots for linear regression models showed no violation of normality . For the logistic regression models, P-values from Hosmer-Lemeshow-Cressie tests were all 40.8, suggesting no evidence of significant lack of fit. There was no indication that considered separately, the association of child care at each age with BMI or overweight varied with sex, highest parental education, occupation or household income . Table 3 shows that, considered separately, informal care at 3, 5 or 11 years was associated with higher BMI z-score and with the presence of overweight at $11 years. The associations were similar when informal care was separated into grandparents, other family members or domestic helpers . Table 4 shows that considering child care at 5 and 11 years only produced the best fitting model for both BMI z-score and the presence of overweight . This was better than additionally including child care at younger ages, or than considering child care as an accumulation of exposures to informal care. The associations for informal childcare as an accumulation of exposures are shown in Supplementary Table S2 available at IJE online. Table 5 shows how child care jointly at 5 and 11 years contributed to BMI z-score and overweight to 0.59) and the presence of overweight or obesity . The complete case analysis produced similar results . --- Discussion In this large, population-representative birth cohort of Chinese children from a developed non-Western setting, informal child care at each of 3, 5 or 11 years considered separately was associated with higher BMI z-score and the presence of overweight at $11 years. Current informal care had the strongest association; however, informal child care at 5 years also made a contribution. Considering only child care at 5 and 11 years produced the most parsimonious model and was preferable to considering the accumulation of exposures to informal child care at all four ages. As such, our findings are largely consistent with those from Western populations, 2-5 although we did not find different associations by family SEP, which was only found in one previous study. However, that study did not explicitly test for differences by SEP, 5 therefore those findings may simply be the chance variation which would be expected on stratification. Our study extends previous studies by showing that the association between child care and childhood BMI occurs at older ages. Our study also adds by replicating these findings in a setting where informal care is common and SEP has different associations with child care and childhood obesity from those in Western countries. 18 The consistency of results in Western and non-Western contexts suggests that these findings may be due to attributes of informal child care rather than residual confounding. There are several limitations in our study. First, BMI at 9-12 years of age might have been assessed before Survey I, which was first sent in July 2008, when our cohort members were aged $11 years. However, informal child care at earlier ages was also prospectively associated with higher BMI z-score and the presence of overweight . Second, information on child care was based on parental recall and could be subject to recall error. However, there is no reason to think that parents' answers depended on their child's BMI. Third, we do not have information on child care duration during the day, so those who reported more than one main caregiver were classified as informal care. However, there were relatively few in this category and excluding these children did not change the overall pattern of results . Misclassification of child care may bias any associations towards the null and make our results conservative. Fourth, we do not have complete information on child care. However, we used multiple imputation which uses all available data, preserves uncertainty for missing data, 37 minimizes inclusion bias and increases statistical power. 27 Moreover, a complete case analysis produced similar results. Fifth, Western studies showed inconsistent results for formal child care. 2,[5][6][7] However, in our setting, formal care is not as common as in Western countries. Formal child care services provided by nurseries take up only a small market share. Moreover, most families are concerned that there may not be adequate attention given to their children in a group care setting. 17 In our cohort, the low use of formal care makes it impossible to evaluate its influence on obesity. Finally, we do not have detailed information on the children's diet and physical activity at each age. Caregivers may influence children's dietary style and physical activity. 38,39 In a Japanese cohort, children with maternal care were less likely to snack and more likely to be physically active. 14 We were unable to compare these characteristic between parental care and informal care, so we were unable to determine which factors may underlie the observed associations. There are several possible explanations for our findings. First, informal child care, particularly at 6 months of age may represent a different infant feeding pattern; for example, greater use of formula feeding, a shorter duration of breastfeeding or an earlier introduction of solid foods. However, in this birth cohort, there was no association between breastfeeding and childhood BMI, 40 nor in this study was there any association between child care at 6 months and adiposity at 11 years of age. Second, it is possible that parents may be more aware of health recommendations than grandparents or domestic helpers. Neither grandparents nor domestic helpers in Hong Kong are usually trained in child care. Grandparents and domestic helpers may also be less exposed to health advice if parents take the child for routine 'well-baby' and 'well-child' visits when health advice may be given. Moreover, domestic helpers in Hong Kong are usually from South East Asia and may not be able to read information leaflets in Chinese. Third, parents may also be more likely to adhere to health recommendations on child rearing. An older generation may have experienced poverty and so may consider fatness a symbol of wealth 41 and health 42 possibly resulting in overfeeding of their grandchildren. Similarly domestic helpers come to Hong Kong from other parts of South East Asia, to earn remittance money, and may also have grown up in relative poverty. If overfeeding was an issue we might have expected informal child care to be more strongly associated with adiposity in boys given the traditional Asian cultural preference for boys. However, we did not find any evidence of differences by sex. Fourth, parents may put more effort into controlling the child's lifestyle, for example, by supervising the child's eating or encouraging physical activity. In contrast, grandparents may put more emphasis on playing with the child. 17 Grandparents may also be more likely to spoil the child by giving energy-dense favourite foods as rewards or by allowing large amounts of video games and television viewing. 4 In Hong Kong, domestic helpers are usually hired for housework as well as child care, so they may not be able to spend much time and effort on the child nor be able to play a parental role in child care according to working mothers' views from semi-structured interviews. 17 Fifth, children may mimic the caregiver's behaviour. 38 Grandparents and domestic helpers may have different patterns of physical activity from parents 4 and may discourage the child's physical activity. 14 From a public health perspective, this study suggests that grandparents or others providing child care may be a potential target for education and intervention to counter the emerging epidemic of childhood obesity. Further research delineating the knowledge of child care of informal caregivers, their attitudes concerning child health, their child care practices and their own behaviour may help in designing effective interventions. --- Conclusion In this population-representative Chinese birth cohort, informal child care, rather than parental care, at both 5 and 11 years was associated with higher BMI z-score and the presence of overweight at $11 years. Our study contributes evidence of a prospective influence of informal child care on childhood obesity in a different socio-historical context in Asia. Future research is needed to understand better the characteristics of different types of child care, so that effective, targeted interventions can be developed to reduce the increasing prevalence of childhood obesity. --- Supplementary Data Supplementary Data are available at IJE online. Conflict of interest: None declared. --- KEY MESSAGES Western studies suggest that informal child care is associated with childhood obesity. However, in Western populations, informal child care and childhood obesity are socially patterned, making these observations vulnerable to residual confounding. In a developed, non-Western setting, informal child care was also associated with childhood obesity. Consistency in different contexts suggests that the association may be due to attributes of informal child care. Interventions might be warranted with informal caregivers for childhood obesity prevention.

Background In Western populations, informal child care is associated with childhood obesity. However, informal child care and obesity share social patterning making evidence from other settings valuable.We used multivariable linear and logistic regression models to estimate the association of child care at 6 months and at 3, 5 and 11 years with body mass index (BMI) z-score and overweight (including obesity) at 11 years in a Hong Kong Chinese birth cohort. We assessed whether associations varied with sex or socio-economic position (SEP). We used multiple imputation for missing exposures and confounders.Of the original 8327 cohort members, 7933 are alive, participating and living in Hong Kong. At $11 years, 6796 had their BMI clinically assessed. Higher SEP was associated with informal care. After imputation, informal care at each of 3, 5 or 11 years was separately associated with higher BMI z-score [3 years 0.10, 95% confidence interval (CI) 0.03-0.18, 5 years 0.12, 95% CI 0.04-0.21, 11 years 0.17, 95% CI 0.04-0.31] and with the presence of overweight [odds ratio (OR) 3 years 1.19, 95% CI 1.03-1.37, 5 years OR 1.20, 95% CI 1.03-1.40, 11 years OR 1.21, 95% CI 1.02-1.45], adjusted for sex, SEP and birth weight z-score. Current informal care had the strongest association. However, informal child care at 5 years also contributed. There was no evidence of differences by sex or SEP. Conclusions In a developed, non-Western setting, informal child care was associated with childhood obesity. Modifiable attributes of informal child care warrant investigation for obesity prevention.

leading to emotionally-rich communication between involved peers. Consequently, instant communication should require specific tools and models for analysis, that are capable of covering these predominant features. Nowadays, IRC channels are still one of the most used platforms for collective real-time online communication and are used for various purposes, e.g. organization of open-source project development, Internet activism, dating, etc. Our dataset , consists of 20 IRC channels covering topics as diverse as music, sports, casuals chats, business, politics, or computer related issues -which is important to ensure that there is no topical bias involved in our analysis. For each channel, we have consecutive daily recordings of the open discussion over a period of 42 days, which amounts to more than 2.5 million posts in total generated by more than 20.000 different users. We process our analysis as follows: first, we look into the communication patterns of instant online discussions, to find out about the average response time of users and its possible dependence on the topics discussed. This shall allow us to identify differences between instantaneous chatting communities and other forms of slower, persistent communication. In a second step, we look more closely into the content of the discussions and how they depend on the emotions expressed by users. Remarkably, we find that most users are very persistent in expressing their positive or negative emotions -which is not expected given the variety of topics and the user anonymity. This leads us to the question in what respect online chats are different from offline discussions which are mostly guided by social norms. We argue that even in instantaneous, anonymous online chats users behave very much like ''normal'' people. Our quantitative insights into user's activity patters and their emotional expressions are eventually combined to model interacting emotional agents. We demonstrate that the stylised facts of the emotional persistence can be reproduced by our model by only calibrating a small set of agent features. This success indicates that our modeling framework can be used to test further hypothesis about emotional interaction in online communities. --- Results User activity patterns. An IRC channel is always active, and enables the real time exchange of posts among users about a specific topic. User interaction is instantaneous, the post written by user u 1 is immediately visible to all other users logged into this channel, and user u 2 may reply right away. Fig. 1 illustrates the dynamics in such a channel. As time evolves new users may enter, others may leave or stay quiet until they write follow-up posts at a later time. To characterize these activity patterns, we analyzed the waitingtime, or inter-activity time distribution P, where t refers to the time interval between two consecutive posts of the same user in the same channel and ask about the average response time. We find that t is power-law distributed P , t 2a with some cut-off , with an exponent a 5 1.53 6 0.02. The fit is based on the maximum likelihood approach proposed by Clauset et al. 9 and the power-law nature of the distribution could not be rejected . This finding is inline the power-law distribution already found for diverse human activities [1][2][3][5][6][7] and classifies the communication process as belonging to the regime where posts arrive faster than they can be processed. We note that for a , 2, no average response time is defined . Further, we observe in the plot of Fig. 1B a slight deviation from the power-law at a time interval of about one day, which shows that some users have an additional regularity in their behavior with respect to the time of the day they enter the online discussion. Such deviations were usually treated as power-laws with an exponential cut-off, and can even be explained based on simple entropic arguments 10,11 . However, because of the ''bump'' around the one day time interval, our distribution also seems to provide further evidence to the bi-modality proposed by Wu et al. 12 . We should note, however, that the tail is better fitted by a log-normal distribution rather than an exponential or a Weibull one as shown in Fig. 1B. Here, KS stands for the Kolmogorov-Smirnov statistical test; the smaller this number, the better the fit. We now focus on an important difference between online chats and previously studied forms of communication, such as mail or email exchange, which mostly involve two participants. Due to the collective nature of chats, a chatroom automatically aggregates the posts of a much larger amount of users, which allows us to study their collective temporal behavior. If v denotes the time interval between two consecutive posts in the same channel independent of any user , we find that the distribution P is is still fat-tailed, but does not follow a powerlaw. Interestingly, the time interval between posts significantly depends on the topic discussed in the channel . Some ''hot'' topics receive posts at a shorter rate than others, which can be traced back to the different number of users involved into these discussions. Specifically, we find that the average inter-event time AEvae ch depends on the amount of users in the conversation and becomes smaller for more popular channels, as one would expect. If we rescale the channel dependent inter-event distribution P ch using the average inter-event time AEvae ch per channel and plot AEv ch ae P ch versus v ch /AEv ch ae, we find that all the curves collapse into one master curve . The general scaling form that we used is P 5 F, where F is independent of the average activity level of the component, and represents a universal characteristic of the particular system. Such scaling behavior was reported previously in the literature describing universal patterns in human activity 13 . We fit this master curve by a stretched exponential [14][15][16] P v ð Þ~a c v h i e {b c v v h i À Á cð1Þ where the stretched exponent c is the only fit parameter, while the other two factors a c and b c are dependent on c 14 . A histogram of the c values across the 20 channels is shown in Supplementary Figure S2. Using only the regression results with p , 0.001 we find that the mean value of the stretched exponents is AEcae 5 0.21 6 0.05. We note that stretched exponentials have been reported to describe the inter-event time distribution in systems as diverse as earthquakes 15 and stock markets 16 . These systems commonly exhibit long range correlations which seem to be the origin of the stretched exponential inter-event time distributions 14 . Long range correlations have also been reported in human interaction activity 5,17 , and we tested their presence in the temporal activity over IRC communication. As shown in the Supplementary Figure S3, we verified the existence of long range correlations in the conversation activity. We found that the decay of the autocorrelation function of the inter-event time interval between consecutive posts within a channel is described by a power-law C Dt ð Þ* Dt ð Þ {nvð2Þ with exponent n v ^0:82. In addition, we applied the Detrended Fluctuation Analysis technique 18 , described in detail in the Methods section, and we found a Hurst exponent value, H v ^0:6, which is well in agreement with the scaling relation n v 5 2 2 2H v . For a more detailed discussion about scaling relations, and memory in time series please refer to 19 . In conclusion, our analysis of user activities have revealed a universal dynamics in online chatting communities which is moreover similar to other human activities. This regards the temporal activity of individual users and the inter-event dynamics across different channels, if rescaled by the average inter-event time . We will use these findings as a point of departure for a more in-depth analysis -because obviously the essence of online communication in chatrooms, as compared to other human activities, is not really covered. From the perspective of activity patters, there is not so much new here, which leads us to ask for other dimensions of human communication that could reveal a difference. Emotional expression patterns. Human communication, in addition to the mere transmission of information, also serves purposes such as the reinforcement of social bonds. This could be one of the reasons why human languages are found to be biased towards using words with positive emotional charge 20 . Humans, from the early stages of our lives, develop an affective communication system that enables us to express and regulate emotions 21 . But emotions are also the mediators of our consumer responses to advertising 22 , and many scientists acknowledge their importance in motivating our cognition and action 23 . However, despite the increasing time we spend online, the way we express our emotions in online communities and its impact on possibly large amounts of people is still to be explored. Consequently, we are interested in the role of expressed emotions in online chatting communities. Users, by posting text in chatrooms, also reveal their emotions, which in return can influence the emotional response of other users, as illustrated in Fig. 1A. To understand this emotional interaction, we carry out a sentiment analysis of each post which is described in detail in the Methods section. This automatic classification returns the valence v for each post, i.e. a discrete value {21, 0, 11} that characterizes the emotional charge as either negative, neutral, or positive. Instead of using the real time stamp of each post as in the analysis of the user activity, we now use an artificial time scale in which at each time step one post enters the discussion, so the number of time steps equals the total number of posts. We then monitor how the total emotion expressed in a given channel evolves over time. We use a moving average approach that calculates the mean emotional polarity over different time windows. In Fig. 2A we plot the fraction of neutral, negative and positive posts as a function of time, for different sizes of the time window. While it is obvious that the emotional content largely fluctuates when using a very small time window, we find that for decreasing time resolution the fractions of emotional posts settle down to an almost constant value around which they fluctuate. From this, we can make two interesting observations: the emotional content in the online chats does not really change in the long run , i.e. we observe fluctuations that depend on the time resolution, but no ''evolution'' towards more positive or negative sentiments. For the low resolution, the fraction of neutral posts dominates the positive and negative posts at all times. In fact there is a clear ranking where the fraction of negative posts is always the smallest. Both observations become even more pronounced when averaging over the 20 IRC channels, as Fig. 2B shows. Our findings differ from previous observations of emotional communication in blog posts and forum comments which identified a clear tendency toward negative contributions over time, in particular for periods of intensive user activity 24,25 . Such findings suggest that an increased number of negative emotional posts could boost the activity, and extend the lifetime of a forum discussion. However, blog communication in general evolves slower than e.g. online chats. Hence, we need to better understand the role of emotions in real time Internet communication, which obviously differs from the persistent and delayed interaction in blogs and fora. To further approach this goal, we analyse to what extend the rather constant fraction of emotional posts in IRC channels is due to a persistence in the emotional expressions of users. For this, we apply the DFA technique 18 , to the time series of positive, negative and neutral posts. Since our focus is now on the user, we reconstruct for every user a time series that consists of all posts communicated in any channel, where the time stamp is given by the consecutive number at which the post enters the user's record. In order to have reliable statistics, for the further analysis only those users with more than 100 posts are considered . As the examples in the Supplementary Figure S4 show, some users are very persistent in their emotional expressions , whereas others are really antipersistent in the sense that their expressed emotionality rapidly changes through all three states. The persistence of these users can be characterized by a scalar value, the Hurst exponent H, which is 0.5 if users switch randomly between the emotional states, larger than 0.5. if users are rather persistent in their emotional expressions, or smaller than 0.5 if users have strong tendency to switch between opposite states, as the antipersistent time series of Fig. S4 shows. If we analyse the distribution of the Hurst exponents of all users, shown in the histogram of Fig. 3A, we find that the emotional expression of users is far from being random, and that it is clearly skewed towards H . 0.5, which means that the majority of users is quite persistent regarding their positive, negative or neutral emotions. This persistence can be also seen as a kind of memory in changing the emotional expression, i.e. the following post from the same user is more likely to have the same emotional value. The question whether persistent users express more positive or negative emotions is answered in Fig. 3B, where we show a scatter plot of H versus the mean value of the emotions expressed by each user. Again, we verify that the majority of users has H . 0.5, but we also see that the mean value of emotions expressed by the persistent users is largely positive. This corresponds to the general bias towards positive emotional expression detected in written expression 20 . The lower left quadrant of the scatter plot is almost empty, which means that users expressing on average negative emotions tend to be persistent as well. A possible interpretation for this could be the relation between negative personal experiences and rumination as discussed in psychology 26 . Antipersistent users, on the other hand, mostly switch between positive and neutral emotions. Are the more active users also the emotionally persistent ones? In Supplementary Figure S6 we show a scatter plot of the Hurst exponent dependent on the total activity of each user. Even though the mean value of H does not show any such dependence, we observe large heterogeneity on the values of H for users with low activity. Furthermore, in Supplementary Figure S7 we show that the Hurst exponent of a very active user varies only slightly if we divide his time series into various segments and apply the DFA method to these segments. Thus we can conclude that active users tend to be emotionally persistent and, as most persistent users express positive emotions, they tend to provide some kind of positive bias to the IRC, whereas users occasionally entering the chat may just try to get rid of some negative emotions. This leads us to the question how persistent the emotional bias of a whole discussion is. While Fig. 3A has shown the persistence with respect to the different users, Fig. 3C plots the persistence for the different channels, which each feature a very different topic. This persistence holds even even if we analyse only certain segments of the channel, as it is shown in Supplementary Figure S8. So, we conclude that the persistence of the discussion per se reflects a certain narrative memory. Precisely, for each chat, we observe the emergence of a certain ''tone'' in the narration which can be positive, negative or neutral, dependent the emotional expressions of the persistent users. If we reshuffle these time series such that the same total number of positive, negative, and neutral posts is kept, but temporal correlations are destroyed, then the persistence is lost as well as Fig. 3C shows. We note that we could not find evidence of correlations using the autocorrelation function of the emotion time series, while the observed persistence in the fluctuations of user emotional expression, as captured by the Hurst exponent is very robust. This indicates that the chat community assumes an emotional memory locally encoded in the current messages , while the size of the conversation is too large to detect it through averaging techniques. An agent-based model for chatroom users. After identifying both the activity patterns, and the emotional expression patterns of users in online chats, we setup an agent-based model that is able to reproduce these stylized facts. We start from a general framework 27 , designed to model and explain the emergence of collective emotions in online communities through the evolution of psychological variables that can be measured in experimental setups and psychological studies 28,29 . This framework provides a unified approach to create models that capture collective properties of different online communities, and allows to compare the different emotional microdynamics present in various types of communication. The case of IRC channel communication is of particular interest because of its fast and ephemeral nature. Thus, we have designed a model for IRC chatrooms, as shown in Fig. 4A. The agents in our model are characterized by two variables, their emotionality, or valence, v which is either positive or negative and their activity, or arousal, which is represented by the time interval t between two posts s in the chatroom. The valence of an agent i, represented by the internal variable v i , changes in time due to a superposition of stochastic and deterministic influences 27,30 : _ v i ~{c v v i zb à h z {h { ð ÞÃv i zA v j ið3Þ The stochastic influences are modeled as a random factor A v j i normally distributed with zero mean and amplitude A v , and represent all changes of the individual emotional state apart from chat communication. The deterministic influences are composed of an internal decay of parameter c v , and an external influence of the conversation. The change in the valence caused by the emotionality of the field is measured in valence change per time unit through the parameter b. Previous models under the same framework 27,31 had an additional saturation term in the equation of the valence dynamics. This way the positive feedback between v and h was limited when the field was very large. But, as we show in Fig. 2, chatrooms do not show the extreme cases of emotional polarization observed in other communities. Thus, we simplify the dynamics of the valence without using any saturation terms, since a large imbalance between h 1 and h 2 is unrealistic given our analysis of real IRC data. In general, the level of activity associated with the emotion, known as arousal, can be explicitly modeled by stochastic dynamics as well 31 . Here, the activity of an agent is estimated by the time-delay distribution that triggers the expression of the agent, i.e. by the power-law distribution P , t 21.53 shown in Fig. 1B. Assuming that an agent becomes active and expresses its emotion at time t, it will become active again after a period t. The agent then writes a post in the online chat the emotional content of which is determined by its valence . This information is stored in an external field common for all agents, which is composed of two components, h 2 and h 1 , for negative and positive information, and their difference measures the emotional charge of the communication activity. Since we are interested in emotional communication, we assume that all neutral posts entered, or already present, in a chatroom do not influence the emotions of the agents participating to the conversation. Thus, the dynamics of the field is influenced only by the amount of agents expressing a particular emotion at a given time: N 1 5 S i ) and N 2 5 S i ), where H is the Heaviside step function. Therefore, the time dynamics of the fields can be described as: _ h + ~{c h h + zc à N + t ð Þð4Þ These two field components, h 1 and h 2 , decay exponentially with a constant factor c h , i.e. their importance decays very fast as they move further down the screen . Each field increases by a fixed amount c from every post stored in it. The values of the valence of the agents are changed by the field components, as described by Eq. 3. In contrast with traditional means of communication, online social media can aggregate much larger volumes of user-generated information. This is why h is defined without explicit bounds. Chatrooms pose a special case to this kind of communication, as they can contain large amount of posts but limited amount of users. Most IRC channels have technical limitations for the amount of users that can be connected at once, which in turn is reflected in the total amount of posts present in the general discussion. In our model, h might take any value, but the empirical activity pattern combined with the fixed size of the community dynamically constraints it to limited values. Whenever an agent creates a new post in an ongoing conversation, the variable, s i , obtain its value in the following way: s i {1 if v i vV { z1 if v i wV z 0 otherwise: 8 > < > :ð5Þ The thresholds V 2 and V 1 represent a limit value of the valence that determines the emotional content of each post, and in general can be asymmetric, as humans tend to have different thresholds for the triggering of positive and negative emotional expression. Each action contributes to the amount of information stored in the information field of the conversation, increasing h 2 if s 5 21 or h 1 if s 5 11. We emphasize that the way we model the agent behavior is very much in line with psychological research, where emotional states are represented by valence and arousal, following the dimensional representation of core affect 32 . The valence, v, represents the level of pleasure experienced by the emotional state, while the arousal represents the degree of activity induced by the emotional state, and determines the moment when posts are created. Continuously the agent's valence relaxes to a neutral state and is subject to stochastic influences, as show empirically in 33 . The effect of chatroom communication on an agent's emotionality is modeled as an empathydriven process 34 that influences the valence. In the valence dynamics we propose in Eq. 3, agents perceive a positive influence when their emotional state matches the one of the community, and a negative one in the opposite case. When a post is created, its emotional polarity is determined by the valence, as it was suggested by experimental studies on social sharing of emotions 26,35 . All the assumptions of our model are supported by psychological theories. Parameter values and dynamical equations can be tested against experiments in psychology, providing empirical validation for the emotional microdynamics 28,29 . Furthermore, our model provides a consistent view of the emotional behavior in chatrooms leading to testable hypotheses that can drive future psychology research. We performed extensive computer simulations using different parameter sets . By exploring the parameter space, we identified which parameter sets lead to similar conversation patterns as observed in the real data. We used such set to simulate chats in 10 channels, and we analysed the agent's activity and their emotional persistence. The results are shown in Fig. 4B,C. Specifically, we find that the distribution of Hurst exponents for individual agents is shifted towards positive values similar to the one observed in real data, this way reproducing the emotional persistence of the conversation without assuming any time dependence between user expressions. Further, we reproduce the empirically observed stretched exponential distribution for the rescaled time delays v9 between consecutive posts, without any further assumptions. We do note, however, that the stretched exponent, c 5 0.59 , of the simulated distribution is different from real IRC channels where it was c 5 0.21, i.e. there is a faster decay in the simulations. This could be explained by the fact that in the real chat users usually write after they have read the previous post, i.e. there are additional correlations in the times users enter a chat. These, however, are not considered in the simulations, because agents post in the chat at random after a given time interval t, i.e. there is no additional coupling in posting times. Following the same approach as we did for the real data, we calculated the Hurst exponent of the inter simulated event time-series of the discussions. We found that H v9 5 0.75, however, we did not observe a power-law decay of the autocorrelation function . This suggests that the observed correlations are due to the power-law distributed interevent times used as input to our model, and it is inline with the above discussion about the absence of coupling that also explains the difference in the stretched exponents. Eventually, we observe the emotional persistence in the simulated conversations. The mean Hurst exponent for the 10 simulated channels is H s 5 0.567 6 0.007, whereas for the real IRC channels H r 5 0.572 6 0.021 was found. These results suggests that our agentbased model reproduces qualitatively the emergence of emotional persistence in the IRC conversation and thus, based on all findings, is able to capture the essence of emotional influence between users in chatrooms. --- Discussion We started with the question to what extent human communication patterns change on the Internet. To answer this, we used a unique dataset of online chatting communities with about 2.5 million posts on 20 different topics. Our analysis considered two different dimensions of the communication process: activity, expressed by the time intervals t at which users contribute to the communication, and v at which consecutive posts appear in a chat, and the emotional expressions of users. With respect to activity patterns we did not find considerable differences between online chatrooms and other previously studied forms on online and offline communication. Specifically, both the inter-activity distribution of users and the inter-event distribution of posts followed the known distributions. Thus, we may conclude that humans do not really change their activity patterns when they go online. Instead, these patterns seem to be quite robust across online and offline communication. The picture differs, however, when looking at the emotional expressions of users. While we cannot directly compare our findings on emotional persistence to results about offline communication, we find differences between online chatrooms and other forms online communication, such as blogs, fora. While the latter could be heated up by negative emotional patterns, we observe that online chats, which are instantaneous in time, very much follow a balanced emotional pattern across all topics , but also with respect to individual users, which are in their majority quite persistent in their emotional expressions . This observation is indeed surprising as online chats are mostly anonymous, i.e. users do not reveal their personal identity. However, they still seem to behave according to certain social norms, i.e. there is a clear tendency to express an opinion in a neutral to positive emotional way, avoiding direct confrontations or emotional debates. One of the reasons for such behavior comes from the ''repeated interaction'' underlying online chats. As the daily ''bump'' the activity patterns also suggest, most users return to the online chats regularly, to meet other users they may already know. This puts a kind of social pressure on their behavior to behave similar to offline conversations. In conclusion, we find that the online communication patters do not differ much from common offline behavior if a repeated interaction could be assumed. Eventually, we argue that the emotional persistence found is indeed related to the nature of human conversations. After all, the correlations shown in the emotional expressions of different users indicate that there is some form of emotional sharing between participants. This suggests the presence of social bonds among users in the chatroom 26 and confirms similarities between online and offline communication. The fact that we could reveal patterns of emotional persistence both in users and in topics discussed, does not mean that we also understand their origin. One important step towards this ''microscopic'' understanding is provided by our agent-based model of emotional interactions in chatrooms. By using assumptions about the agent's behavior which are rooted in research in psychology, we are able to reproduce the stylized facts of the chatroom conversation, both for the activity in channels and for the emotional persistence. Specifically, our model allows us to test hypotheses about the emotional interaction of agents against their outcome on the systemic level, i.e. for the chatroom simulation. This helps to reveal what kind of rules are underlying the online behavior of users which are hard to access otherwise. --- Methods Data collection and classification. The data used in this article is based on a large set of public channels from EFNET Internet Relay Chats , to which any user can connect and participate in the conversation. Based on the assessment of the initially downloaded set of recordings, 20 IRC channels were selected aiming to provide a large number of consecutive daily logs with transcripts of vivid discussions between the channel participants, measured in number of posts. The finally used data set contained consecutive recordings for 42 days spanning the period from 04-04-2006 to 15-05-2006. The general topics of discussions from the selected channels include: music, sports, casuals chats, business, politics and topics related to computers, operating systems or specific computer programs. The IRC data set contains 2,688,760 posts. The total number of participants to all this channels is 25,166. However, because some people participate to more than one channel, the total number of unique participants is 20,441. On average, the data set provides 3055 posts per day. In the recorded period 15 users created more than 10000 posts. The distribution of the user participation i.e. the number of posts entered by every user, is shown in Supplementary Figure S1. The mean of the distribution is 97 posts per user, and as we can see from Fig. S1, it is skewed with most of the users contributing only a small number of posts. The acquired data was anonymized by substituting real user ids to random number references. The text of each post was cleaned by spam detection and substitution of URL links to avoid them from influencing the emotion classification. The emotional content was extracted by using the SentiStrength classifier 36 , which provides two scores for positive and negative content. Each score ranges from 1 to 5, and changes with the appearance of emotion bearing terms from a lexicon of affective word usage, specifically designed for this purpose. Each word of the lexicon has a value on the scale of 25 to 5 which determines the strength of the emotion attached to it. The classifier takes into account syntactic rules like negation, amplification and reduction, and detects repetition of letters and exclamation signs as amplifiers. When one of this patterns is detected, SentiStrength applies transformation rules to the contribution of the involved terms to the sentence scores. It has been designed to analyze online data, and considers Internet language by detecting emoticons and correcting spelling mistakes. The perception of emotional expression varies largely across humans, and traditional accuracy metrics are not useful when there is lack of an objective space. Human ratings of emotional texts have certain degree of disagreement that needs to be considered by sentiment analysis in order to have a valid quantification of emotions. SentiStrength scores are consistent with the level of disagreement between humans about how they perceive written emotional expressions 37 . This classifier combines an emotion quantization of proved validity with a high accuracy, and is considered the state of the art in sentiment detection 38 . Due to the short length of the posts in chatrooms, we calculate a polarity measure by comparing the two different scores of SentiStrength. The sign of the difference of the positive and negative scores provides an approximation to detect positive, negative and neutral posts. The accuracy of this polarity metric was tested against texts tagged by humans and messages including emoticons from MySpace 39 and Twitter 40 , which are of a similar length to the ones in our chatroom data. The data are freely available for research purposes, and are provided as Supplementary Material. Detailed information about their structure is provided in the ''Data section'' of the Supplementary Information text. Detrended Fluctuation Analysis. The method of Detrended Fluctuation Analysis 18 is a useful tool in revealing long-term memory and correlations in time series 5,15,16 . The method maps the system into a one-dimensional random walk, and enable us to compare the properties of the real time series with the time series produced by the random case. The DFA analysis of a time series x with length T, which can be divided into N segments is performed as follows: First we integrate the time series, by calculating the profile Y t ð Þ~P t t' x t' ð Þ{vx t ð Þw ½ . Next, we divide the integrated time series into N boxes of equal length Dt. Each box has a local trend, which in a first level approximation, can be fitted by a linear function using least squares. We denote with y Dt the y coordinate of the straight line segments that represent the local trend in each box, and we subtract this local trend from the integrated time series Y. Next we use the function F Dt ð Þ~ffi ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 N X N k~1 Y k ð Þ{y Dt k ð Þ ½ 2 v u u tð6Þ to calculate the root-mean-square fluctuation of the integrated and detrended time series, and we characterize the relationship between the average fluctuation F, and the box size Dt. Typically, F will increase with box size as F , H , which indicates the presence of power-law scaling. Therefore, the fluctuations can be characterized only by the scaling exponent H that is analogous to the Hurst exponent 41 , and it is calculated from the slope of the line relating logF to logDt. If only short-range correlations exist in the time series, then it has the statistical properties of a random walk. Therefore F , 1/2 . However, in the presence of long-range power-law correlations H ? 1/2. A value H , 1/2 signals the presence of long range anti-correlations, while a value H . 1/2 signals the presence of long range correlations . --- --- Additional information Supplementary information accompanies this paper at http://www.nature.com/ scientificreports Competing financial interests: The authors declare no competing financial interests. License: This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/

How do users behave in online chatrooms, where they instantaneously read and write posts? We analyzed about 2.5 million posts covering various topics in Internet relay channels, and found that user activity patterns follow known power-law and stretched exponential distributions, indicating that online chat activity is not different from other forms of communication. Analysing the emotional expressions (positive, negative, neutral) of users, we revealed a remarkable persistence both for individual users and channels. I.e. despite their anonymity, users tend to follow social norms in repeated interactions in online chats, which results in a specific emotional ''tone'' of the channels. We provide an agent-based model of emotional interaction, which recovers qualitatively both the activity patterns in chatrooms and the emotional persistence of users and channels. While our assumptions about agent's emotional expressions are rooted in psychology, the model allows to test different hypothesis regarding their emotional impact in online communication.ow do human communication patterns change on the Internet? Round the clock activities of Internet users put us into the comfortable situation of having massive data from various sources available at a fine time resolution. But what to look at? Which aggregated measures are most appropriate to capture how new technologies affect our communicative behavior? And then, are we able to match these findings with a dynamic model that is able to generate insights into their origin? In this paper, we provide both: a new way of analysing data from online chats, and a model of interacting agents to reproduce the stylized facts of our analysis. In addition to the activity patterns of users, we also analyse and model their emotional expressions that trigger the interactions of users in online chats. Validating our agent-based model against empirical findings allows us to draw conclusions about the role of emotions in this form of communication. Online communication can be seen as a large-scale social experiment that constantly provides us with data about user activities and interactions. Consequently, time series analyses have already revealed remarkable temporal activity patterns, e.g. in email communication. Such patterns allow conclusions how humans organize their time and give different priorities to their communication tasks [1][2][3][5][6][7] . One particular quantity to describe these patterns is the distribution P(t) of the waiting time t that elapses before a particular user answers e.g. an email. Different studies have confirmed the power-law nature of this distribution, P(t) , t 2a . Its origin was attributed either to the burstiness of events 2 or to circadian activity patterns 3 , while a recent work shows that a combination of both effects is also a plausible scenario 4 . However, the value of the exponent a is still debated. A stochastic priority queue model 6 allows to derive a by comparing two different rates, the average rate l of messages arriving and the average rate m of processing messages. If m # l, i.e. if messages arrive faster than they can be processed, a 5 3/2 was found, which is compatible with most empirical findings and simulation models [1][2][3]8 . However, in the opposite case, m $ l, i.e. if messages can be processed upon arrival, a 5 5/2 was found together with an exponential correction term. The latter regime, also denoted as the ''highly attentive regime'', could be verified empirically so far only by using data about donations 7 . So, it is an interesting question to analyze other forms of online communication to see whether there is evidence for the second regime. In this paper, we analyze data about instant online communication in different chatting communities, specifically Internet Relay Chat (IRC) channels, where each channel covers a particular topic. Prior to the very common social networking sites of today, IRC channels provided a safe and independent way for users to share and discuss information outside traditional media. Different from other types of online communication, such as blogs or fora where entries are posted at a given time (decided by the writer), IRC chats are instantaneous in real time, i.e. users read while the post is written and can react immediately. This type of interaction requires much higher user activity in comparison to persistent communication e.g. in fora. Further, it is more spontaneous, often

Introduction People are exposed to vast amounts of information through different social platforms. Individuals can collect and share information through these platforms at unprecedented speeds, influencing both their lives 1 and decision-making, 2 as such platforms can disseminate information faster than conventional websites. 3 Individuals generally use social media to stay in touch with friends and family, while some post, search, and share information through it. Females generally use social media for information sharing in different situations compared to males. 4 A problem with social media is the spreading of false or fabricated information that can harm others, cause hysteria, or make people disregard accurate information presented on various social media platforms. 5 Inaccurate or misinformation can evoke unwanted emotions, such as confusion and anxiety, which can subsequently motivate individuals to share incorrect information with others on social media. 6 --- Social Media and Pandemics Comments on social media reflect people's thoughts and interests. 7 They are usually affected by news and content posted on social media. Particularly, comments offer some insight into individuals' perspectives, as witnessed during the 2009 H1N1 pandemic, when comments had a significant role in influencing people's decision-making, opinions, and perspectives. 7 Diseases that are highly discussed in the media are perceived as more serious than ones with a low frequency of discussion. 8 Furthermore, negative tweets are associated with higher stress, and high levels of fear are associated with a greater frequency of negative tweets. Positive emotions, however, are associated with tweets reflecting happiness. Thus, data and posts on social media can influence individuals' well-being, and social media platforms can help predict stress levels and emotional states. 9 Notably, individuals' attitudes toward social media are generally positive, even in the case of medical issues. 10 Coronavirus disease 2019 is the latest pandemic that has disrupted the world. On December 31, 2019, this novel disease was first identified in Wuhan, Hubei Province, China. On January 10, 2020, the World Health Organization issued technical guidelines, advising countries on how to detect the disease and manage potential cases. On March 11, COVID-19 was characterized as a pandemic. 11 After COVID-19 started spreading around the globe, many people used social media to share information. Some materials and posts were inaccurate. However, most information was reliable and practical, especially when shared on popular social media. 12 On the one hand, social media had a vital role in managing communication and operations during the pandemic among university employees in the Philippines, 13 and helped to disseminate medical information much faster than what was possible 30 years ago. 14 On the other hand, academicians found that social media was employed as a platform for spreading conspiracy theories during the Arab Spring and the initial release of WikiLeaks in 2010, as well as in 2020 when conspiracy theorists on Twitter linked the COVID-19 pandemic to the 5G cellular network. 15 It was even suggested that the virus accidentally escaped or was even deliberately released from a Chinese biological weapons laboratory. This rumor was closely related to the 5G conspiracy theory. According to another conspiracy theory, the COVID-19 pandemic is a hoax and poses no significant threat other than economic damage caused by stay-at-home restrictions. Bill Gates and his foundation were also placed center stage in conspiracy theories involving global vaccination and surveillance. 16 Many social media conspiracies are discussed on different platforms and messaging applications. 15 The emotions expressed in social media differ according to the platform's purpose. Negative emotions appear more frequently due to the need for socialization and information sharing about events and updates. Conversely, positive emotions are identified with entertainment and information related to products, services, ideas, and inspirations. 17 The expression of negative emotions has been rated appropriate for WhatsApp, followed by Facebook, Twitter, and Instagram. Moreover, the expression of positive emotions is more appropriate for WhatsApp, followed by Instagram, Facebook, and Twitter. 18 --- Social Media and Information Dissemination As the use of social media platforms rose, people obtained information from more than one source. However, the sources are not always professional but do have a strong presence online . Accordingly, such media can influence people by communicating messages. 19 If used properly, health communication can help protect public health through proper planning and education. Such forms of communication can maximize awareness and peoples' adaptability to changing circumstances, particularly when health plans are made from the perspective of a targeted group and are relevant to their needs. 20 Many people employ multiple messaging applications to communicate with family members and friends. Several applications permit and support cross-application conversations, enabling users to send and receive messages. Each messaging application fulfills a specific communication role; for instance, people choose a certain application according to their experience with it. 21 --- WhatsApp as a Social Media and Messaging Application WhatsApp is a popular messaging platform for sharing information and keeping in touch with individuals and groups. 22 It is an instant messaging application available on smartphones and employs the Internet for communication, allowing users to send images, videos, and audio media in addition to their location, 23 while maintaining privacy. 24 WhatsApp accounts for a large proportion of typical daily smartphone use. 25 Many people believe it is an appropriate platform to express feelings and emotions. 18 Hence, WhatsApp helps shape decision-making, improves communication, and allows data to be sent to others, 24 especially essential information that can help resolve daily life issues. 26 Females use WhatsApp more than males, and younger people tend to use it more than older ones. WhatsApp is apparently employed more than other messaging applications and social media, such as Facebook. 25 WhatsApp helps deepen friendships; it allows multiple participants to be added to a group, thus supporting a collective experience. 27 According to several journalists, WhatsApp is a fast way of circulating information. Ordinary citizens report incidents and news to the media before any reporter. A partnership has developed between ordinary people and media channels. For instance, citizens may record a video of an event on the spot and send it to different media channels using WhatsApp. Therefore, it is a way of exchanging information and news immediately among groups. 28 --- Pandemics and COVID-19 Throughout history, infectious diseases have resulted in the demise of countless humans. People's perception of and response to such risks and threats affect their decision-making during epidemics and pandemics. 29 During the SARS epidemic , people in infected areas mostly avoided public transport, which was perceived hazardous. Large gatherings of people at restaurants, entertainment facilities, hospitals, and stores were also regarded as risky, though regional variations were evident. Asian respondents reported that they were less likely to avoid restaurants and theaters and more likely to avoid doctor visits. European respondents indicated that stores and hospitals were riskier places and should be avoided. 30 Cognitive variables during the swine flu pandemic generally showed low social sensitivity. However, a few key threat representations helped predict behavioral change, such as perceived control, understanding, and fear of disease. Thus, public information about infectious diseases is necessary to tackle inaccurate beliefs and help prevent diseases from spreading. 31 During the H1N1 pandemic, the content on Twitter observably changed over time. The public's attitude and behavior altered during the pandemic, and public concern and precautionary behavior increased when the threat of the outbreak grew and decreased when the perceived risk declined. Personal accounts of H1N1 increased with time, and the number of humorous comments fell. Besides, a large amount of misinformation was not even detected. 32 Some measures have been adopted by various countries and organizations during epidemics and pandemics to curb the spread of diseases. Examples include school closures during the 1918, 1957, and 1968 pandemics in the United States , France, Hong Kong, and some Australian cities, which had some effect on the outbreak rate among children. 33 Implementing social distancing in Mexico during the H1N1 pandemic caused the outbreak to pause, but it did not stop the disease completely between waves. 34 Social distancing is the practice of increasing the space between people by about two meters or more to decrease the chance of a disease spreading. In addition to working remotely and avoiding public transport, social distancing is one of the main measures recommended to decrease the spread of COVID-19. 35 Governments in most countries closed schools and educational institutions temporarily as a precautionary measure to prevent COVID-19 from spreading. 36 School closure likely reduces influenza transmission during outbreaks. 37 United Nations International Children's Emergency Fund urges schools to provide children with vital information on how to protect themselves and their families against COVID-19. The International Federation of Red Cross and Red Crescent Societies, UNICEF, and the World Health Organization issue guidance on protecting children and supporting safe school operations. 38 During the COVID-19 pandemic, many people relied on social media and social messaging to acquire information and keep abreast with changes. Twitter posts allowed users to read about inspiring stories, global efforts on fighting COVID-19, and ideas and humorous ways of approaching medical quarantine. 5 Except for Twitter, people's activities on social media during emergencies and life-changing situations have generally been under-researched. Furthermore, Twitter is more accessible for research purposes than other social media and messaging platforms. 4 Whether accurate and reliable or incorrect and undependable, news and information spreading on social media and messaging platforms is dependent on the group dynamics of those engaged with the topic. 13 The Present Study This study evaluates the messages of a WhatsApp group of friends in Saudi Arabia concerning the COVID-19 pandemic to capture their attitudes, emotions, and perception with no reservations. This aspect is not possible on other social media because most other social media applications are open to the general public; therefore, individuals' interactions might not be as transparent as on WhatsApp. Moreover, we did not find a study investigating the emotions and attitudes of individuals on a private social media application such as WhatsApp. This research answers the following question: do WhatsApp messages and posts influence users' perceptions, emotions, and attitudes? This study also examined the use of WhatsApp during the pandemic. WhatsApp is the main social application used for communication among friends and family members in Saudi Arabia and is more popular than Twitter, Facebook, Instagram, and other social applications. 39 --- Methods This research combines a quantitative exploratory study and qualitative content analysis of a sample of a WhatsApp group's interaction messaging during the COVID-19 pandemic in Saudi Arabia. --- Quantitative Study This quantitative exploratory study intends to determine the utilization of WhatsApp as a social communication and messaging application during the COVID-19 pandemic in Saudi Arabia. --- --- Tools A questionnaire was developed to obtain relevant information to better understand WhatsApp utilization. It constituted 24 questions, among which four questions obtained demographic data, 13 questions had a four-point Likert-scale answer , five questions had a three-point Likert-scale answer , and two questions permitted the participants to choose more than one answer. The survey was available as a Google Drive link on WhatsApp, Instagram, Twitter, SMS, and email. The questionnaire was sent to eight university academics with the face validity form so that its face validity could be tested. Overall, six members of our expert panel agreed that the questionnaire is valid at 90% according to its wording, clarity, and relation to the study. However, two of our experts suggested some changes to the wording of some questions, which was taken into consideration while finalizing the questionnaire. --- Procedures The questionnaire was available as a Google Drive link on WhatsApp, Instagram, Twitter, and SMS, and was also sent as an email. Individuals that were interested to participate answered the questionnaire and were also asked to forward it to family and friends if interested. --- Results With regards to the use of different social apps, where participants were allowed to select more than one app, the proportions of users were as follows: WhatsApp, 93% of respondents; Snapchat, 71%; Twitter, 58%; Instagram, 53%; Facebook, 15%; and TikTok, 9%. Thus, WhatsApp was used more than any other application. The survey findings are summarized in Tables 2 and3. Participants were asked if they had downloaded the Saudi Ministry of Health application on their mobiles. Notably, 47.2% downloaded the appointment application, 40.6% downloaded the health application, 4% downloaded the health for doctor's application, and 39.5% did not download or follow any Ministry of Health applications during the pandemic. Some respondents, however, indicated that they downloaded or followed other health applications, such as Asefni , Sehaty , Zeyara , Nabd , a WHO app, and the Ministry of Health's Twitter account and SMS. We asked participants about moments when they began feeling anxious or worried about COVID-19. The highest proportion was 20%, witnessed after the announcement regarding the entry of infected people into Saudi Arabia . Subsequently, this news led to the closure of schools and mosques, marking the beginning of the lockdown. The lowest percentages were seen when COVID-19 was first announced in January and when it spread to other countries, such as Italy and Spain . --- Qualitative Study This qualitative study aimed to determine how a WhatsApp group of friends interacted after the COVID-19 outbreak in terms of their attitudes, emotions, and perceptions. --- Data Collection We assessed posts and messages of a WhatsApp group. The group consisted of 12 Saudi women aged 49-52 years, all university graduates. In particular, 10 participants had children studying in schools or universities in Saudi Arabia or --- 2373 other countries. They had known one another since school. This group was chosen because the members interacted with each other on WhatsApp daily and felt comfortable discussing everything among themselves, including COVID-19 and everything related to it, with no reservations. All 12 of the participants gave their consent to participate and for their WhatsApp communications to be analyzed for this study. --- Data Analyses and Results A qualitative approach was followed to categorize the data obtained from our participants, whereby all messages were classified according to the type of message , sentiments and prayers, and prevention measures for COVID-19 during the study time frame . We examined and categorized all messages and posts in the WhatsApp group about COVID-19 since the first message appeared in their messaging history . After analyzing the group's messages and posts, a pathway of how the communications affected emotions, attitudes, and perceptions related to COVID-19 was identified . Initially, all the posts shared from January 28 contained general information concerning the new disease and situation in Wuhan. The comments were mostly about doctors reporting the most affected groups, and the group was generally not happy with the situation. A month later, on February 24, one post implied that the coronavirus was like a new plague. However, the number of infected people in other countries was still small. The Umrah to Makkah was halted by Saudi authorities to protect against the virus. However, most group comments denied any immediate danger. The first rumor appeared in the group around the end of February. It was a post on social media about the commencement of school's final exams being held earlier than usual because of the virus. The following day, the rumor was discredited, and all group members agreed to rely on information from original sites, not posts from social media. On March 2, the announcement of the first case of infection in Saudi Arabia, together with the closing of the central main part of the Great Mosque in Makkah, was discussed among the group members. Overall, three emotions were noticeable. Hope was evident through comments and posts about the virus and infected people, with positive thoughts and prayers being relayed for the --- 2374 infected. Sadness was also evident in the remarks and discussions because of the closure of the Great Mosque in Makkah. Lastly, pride was observable in relation to the government's intervention to protect the people. On March 6, the closure of schools and universities generated anger. Infected people had entered the country without informing immigration officers about countries they had visited before entering Saudi Arabia. The change in emotion was due to the sadness caused by the virus, which was clear in the group's comments about schools' closure. This move was considered a major change, indicating the danger the virus represented. With the initiation of online learning, most of the mothers in the group were not happy with the new circumstances because they had to spend the whole day teaching their children. Around mid-March, anxiety became prominent in the comments as the number of infected people rose. The members started discussing the importance of being cautious about everything related to the virus and how to avoid it for the first time. Anger became a common theme in their remarks as some people were not following the guidelines issued by the Ministry of Health. New posts appeared about making sanitizers and producing health guidelines in different languages. Prayers started to increase in messages and notes, especially after the Friday prayers were halted in mosques. Posts about lockdown around the country started to trend, and panic was widespread. Many members were purchasing food and stocking their kitchens. The government announcements, however, reassured the people that food markets would not close, promising some relief. Weeks after the school closure, children and their parents began feeling bored. Hence, posts about group games, exercises, and activities at home appeared in large numbers, and comments about daily goings-on were discussed. By the second week of April, a lockdown was implemented in Riyadh. Sadness and anxiety became very clear in the comments through the usage of tear and broken heart emojis. By mid-May, some posts asked for the lockdown to be eased. The comments reflected the fear generated by the virus. The remarks also suggested that some people were not adhering to precautionary measures. Nevertheless, some group members expressed hope that life would return to normal. --- Type of Messages We identified 2220 messages and posts about COVID-19 from January 28 to June 27, 2020, in the WhatsApp group. Specifically, 51.6% were messages and discussions among group members and 27.5% were posts from other social media platforms and unknown sites . Humor accounted for 5% and rumors accounted for 1% of the total number of messages. The social media posts shared within the group were mostly from Twitter , followed by Instagram . Notably, Facebook posts only accounted for 2.5%. --- Sentiments and Prayers Among messages sent by the WhatsApp group members reflecting sentiments and prayers, prayers had the highest proportion . Fear was noticeable in 18.46% of posts, followed by sadness , anger , caution , hope , happiness, and stress . The lowest sentiment was pride , as seen in Figure 3. These findings reveal that the proportions of fear and prayers were close, possibly indicating that fear led to prayers in the group. Most of the messages indicating fear occurred when the number of infected people and deaths increased. That emotion was followed by sadness because of the suffering the disease had caused. There was anger because some people were not following instructions on how to protect against the virus and respect quarantine rules and the lockdown. --- Prevention Measures Messages and posts regarding prevention measures against COVID-19 started in the group on February 26. The first post was concerned with how pilgrims from infected countries would be denied entry into Saudi Arabia for Umrah. On February 27, the first post appeared about protection against the virus. We found that 80 posts were about prevention measures applied by the government and roles and regulations on social distancing, hygiene, hand sanitizer utilization, and wearing masks. In 35 messages, the comments presented views on the prevention measures implemented by the government and how they would affect the members' lives . Comments and posts sent in the WhatsApp group during the 38 days, from February 26 to April 4, were regarding preventive measures and were taken from official websites and different social media channels. All the posts and comments stopped around the beginning of April. That could be because interest in protecting against a disease usually starts when major related events occur close to home. In the group's case, it was around the time that Umrah in Makkah was not allowed, which underlined the danger and importance of such measures. --- Discussion We studied the use of WhatsApp during the COVID-19 pandemic in Saudi Arabia and its effect on the attitudes and emotions of a WhatsApp group consisting of childhood friends. The extent to which WhatsApp was used was also evaluated. We found 40 while ensuring high quality and improving communications capability. This platform helps its users to share their thoughts and feelings with individuals and within different groups. 41 Contrarily, other social media are usually used for information collection and self-expression. 40 This study found that most participants employed WhatsApp as the main platform of communication with family and friends. They shared information about important topics, in tandem with images and posts from different social and Web platforms. This outcome is consistent with those of Alsanie 39 and Azeema and Nazuk, 28 who state that WhatsApp is not used only for chatting purposes but also for professional practices. We found that a high proportion of participants resent information about COVID-19 to their WhatsApp groups. In particular, 45% of the participants' messages were posts taken from different social media platforms. This result is in line with Kant, 23 who observed that WhatsApp users sent images and posts from various social media platforms. Emotions and attitudes evident in the WhatsApp group chats signified a dismissal of the disease after the candidates first received information about it. This ignorance resulted in rumors regarding the effect the disease might have on the members' lives, especially regarding schools and examinations. However, other rumors and conspiracy theories about COVID-19 and the reality behind the pandemic, such as the relationship between the disease and the 5G cellular network or the possibility of it being a biological weapon, did not appear in the shared posts or in the discussions. This result is at odds with Ahmed et al, 15 who found that conspiracies were discussed on different social platforms, including WhatsApp. The discrepancy between the results could be due to differences between the participants' characteristics and interests. We found that the attitude towards the disease at the beginning of the pandemic was not fear or panic; instead, the prevalent attitude was denial because the event was presumably considered distant. This behavior is usually exhibited during pandemics and is known as optimism bias. The problem here is that it can lead people to underestimate a situation and ignore public health warnings. 29 When COVID-19 reached Saudi Arabia, emotions were conveyed through comments and posts shared by the group members. Changes began occurring around the country with the closing of schools, workplaces, and mosques. Accordingly, sadness, hope, pride, and anger became evident in the comments. The closer the disease, the more prominent the emotions, and all posts and discussions were primarily around the effect of the pandemic. This result concurs with Jonung and Roeger, 42 who state that psychological effects begin to show when key changes occur and precautionary measures are adopted in a community. Posts and remarks concerning disease prevention began appearing along with the need to contain and ideally prevent the disease. 31 Overall, the emotions and changing attitudes were clear with every post sent by the group members. Thus, posts and discussions among people affect their perceptions, attitudes, and emotions. When the posts addressed government prevention measures for COVID-19 and people started following those guidelines, pride became evident. Conversely, when posts revealed that groups of people were disregarding social distancing and precautionary measures, anger became clear in the comments. Figure 5 indicates our interpretation of the pathway of communications effect on emotions and attitudes, where posts and events are presented in triangles, and emotions and attitudes are depicted through oval shapes. This finding is in agreement with Lin et al, 17 who state that social media users are impacted by what they read on social media. It also concurs with Raude and Setbon, 31 who found that the public's attitude and behavior changed over time with social media posts during the H1N1 pandemic. This finding could be because the information can be shared faster on social media than via any other conventional mechanism. 3 Information about COVID-19 appeared widely on social media. Many people received such information through WhatsApp, as shown in our results. Over 50% of our participants indicated that they obtained the latest updates regarding COVID-19 from WhatsApp. Social media information can be false or fabricated. 5 Most of our participants stated that they confirmed the validity of the information received before resending it. We observed that the same behavior was detected in our WhatsApp group. In their group statements, members would add that they ensured the accuracy of the information they were sharing. If they failed to confirm a point, they would be clear that they had just resent the information without validating it. --- Conclusions Overall, our results show that as a social communication and chatting platform, WhatsApp affects its users' attitudes, perceptions, and behavior just as much as any other social media, or perhaps even more. Posts and comments shared on WhatsApp reflected different emotions, ranging from fear, anger, and sadness to hope, happiness, and pride. They also included prayers and cautious behaviors were advocated. Notably, the main emotion observed was fear. Our finding agrees with those of Bavel et al, 29 who also determined that the central emotion during the pandemic was fear, which generated a certain defensive response. However, in our study, fear led to prayers, which reportedly increased during COVID-19 in many countries and faiths. 43 Emotions, perceptions, attitudes, and behaviors indicated by our participants were influenced by what they had been reading and hearing on WhatsApp and perhaps on other social media platforms. --- Practical and Theoretical Implications Our findings could make important practical contributions to parents, teachers, social media managers, and any individual using WhatsApp because the results showed that interacting on WhatsApp during the pandemic impacted emotions and attitudes. Therefore, WhatsApp and other social media could be utilized to help people during crises and hard times in the future. Each social media platform has a different use and a different way of interacting with its users, where the users might not know each other in person, such as on Twitter and YouTube, nor would they be able to amass hundreds of followers. On WhatsApp, the interaction between group members could be more personal; hence information dissemination and discussion could be different from other social media platforms, and more studies should look at WhatsApp as a tool affecting behavior in different situations. This research had several limitations. Most of our survey respondents were living in cities. Therefore, the results reflect behaviors predominant in only cities, and the patterns could differ between villages, towns, and cities. Future research should address how chatting on social media platforms affects users' emotions, lives, and well-being. Such platforms are personal, and people there are open and their communications are more transparent than on other platforms. WhatsApp as social media platform appears to affect emotions and attitudes during the first wave of COVID-19. It could also affect people's behavior under different circumstances; therefore, social media is an important tool to study and explore. --- Ethical Approval The Institutional Review Board of King Saud University approved this study . Participants gave their consent, and the principles specified in the Declaration of Helsinki were followed. --- Disclosure The author reports no conflicts of interest in this work.

Background: Social media is an integral part of daily life as people are exposed to considerable amounts of information via social media. In particular, WhatsApp is a messaging application used for sharing information and keeping in touch with individuals and groups. It is a platform considered appropriate for expressing emotions. Purpose: This study analyzed a WhatsApp group of friends in Saudi Arabia and their messages regarding the coronavirus disease 2019 (COVID-19) pandemic during the first wave of infections to capture their attitudes, emotions, and perceptions related to the pandemic. We also investigated WhatsApp usage during COVID-19. Methods and Participants: A quantitative exploratory study and qualitative content analysis of a sample of WhatsApp group interactions during the pandemic in Saudi Arabia were combined. Results: We found that WhatsApp was used more frequently than other social media platforms for communication with family and friends. Moreover, denial was the first behavior evident at the beginning of the pandemic, followed by hope, sadness, and anxiety, and then, panic, fear, and happiness. Conclusion: WhatsApp affects its users' attitudes, perceptions, and behavior just as much as any other social media. Posts and comments shared on WhatsApp reflected different emotions, ranging from fear, anger, and sadness to hope, happiness, and pride.

When an academic culture based on self-enhancement values undermines female students' sense of belonging, self-efficacy, and academic choices The higher education system in the western countries is becoming increasingly selective , creating a negative social interdependence dynamic and thus, an academic culture emphasizing self-enhancement values . We argue that depicting academic success as being linked to these values may have a deleterious effect on women's academic experience. We lay the ground for this proposal by looking first at more nuanced indicators on gender inequalities in the western higher education system, and at how past research may be limited in fully addressing these dynamics. --- Gender in Higher Education: From general tendencies to more nuanced indicators Women today represent the majority of students in higher education in the western countries . However, despite their higher enrolment percentages, which may paint a positive preliminary picture regarding women's experience in the academic system, particular indicators come into play to nuance this view. One of the most documented indicators regarding the gender issue in higher education is the variable of academic choices, generally in reference to the gender dynamics in STEM fields. In France as well as the majority of European Union countries , women are still underrepresented in the fields of sciences, technology, engineering and mathematics. However, even when women enroll and persist in these fields, other features of the higher education environment may negatively interfere with their academic path. For instance, a study conducted in the United Kingdom on a population of chemistry doctoral students revealed that women's interest in pursuing an academic career in the field decreased significantly during their doctoral studies, in part because of the competitive culture within the research teams. While in the first year, 72% of women and 61% of men expressed their interest in pursuing an academic career, by the third year, the percentage decreased to 37% among women compared to 59% among men. Moreover, there is often a smaller percentage of women than men among doctoral graduates even in domains in which they are in the majority at the undergraduate level. For example, according to a national French report , 62% of the bachelor's degree students enrolled in economics and administration were women. However, they represented only 42% of the students enrolled in the PhD programs in the same field. Research has also revealed that women experience higher belonging uncertainty , including in fields where they are not targeted by a negative stereotype. These findings, which are illustrative rather than exhaustive, bring into question the idea that because they represent the majority of students, women would have an utterly positive psychological experience in higher education. Previous research addressing women's poorer academic experience in higher education has focused on two main explanations: social identity threat and the goals congruity hypothesis . The former occurs when women fear that they may be devalued in an academic setting because of their gender . The latter proposes that gendered endorsement of social goals leads to differentiated career paths, with women less likely to study in the most lucrative and prestigious STEM fields. It is worth noting the more recent approach of Leslie and colleagues that generalizes the deleterious effect of gender stereotypes beyond the STEM versus non-STEM framework. Their findings show that the percentage of women in a field of study may be predicted by the shared beliefs that success in the field requires brilliance, a feature stereotypically associated more with men . In the current research, we aim to go beyond these approaches, arguing that regardless of whether an academic setting is more or less congruent with women's goals or more or less targeted by negative stereotypes, the cultural marks associated with academic success in the increasingly selective realm of higher education may also negatively affect women's academic experience. --- A culture of selection in higher education Many Western European countries with minimal university entry requirements have large first year enrollment and limit masters and doctoral students to those deemed competent or having merit . This selection process creates a "struggle" to be perceived as competent, increasing students' willingness to demonstrate their own competences relative to others' . In the same vein, Stephens and colleagues illustrated that western higher educational institutions focus primarily on norms of independence . The cultural frame created by these selection dynamics is theorized in Schwartz's model of values as the self-enhancement values . Self-enhancement corresponds to emphasizing competence in terms of prevailing social standards and the attainment of a dominant position within a social system. On the opposite side of the cultural spectrum, emphasizing the promotion of supportive social relations are the self-transcendence values . This psychological opposition is well illustrated in the context of higher education. A low level of concern for others, often going as far as trying to impede others' progress, could be reasonably perceived as increasing the chances of success for oneself. For example, Toma and Butera showed that competition led group members to withhold unshared information in a decision-making task. To sum up, we argue that an academic culture that emphasizes self-enhancement values and deemphasizes self-transcendence values has become a pervasive feature of the current western academic system, regardless of whether a particular field is a STEM, non-STEM or brilliance-focused field. Because women endorse self-enhancement values less and self-transcendence values more than men do , they may have a more challenging academic experience, notably in terms of belonging uncertainty, decreased academic selfefficacy, with downstream consequences on women's academic choices. --- Sense of belonging, academic self-efficacy and academic choices A sense of belonging in a given academic setting captures the extent to which students feel accepted, included, and supported by others . Students who experience a sense of belonging in educational environments are more engaged in classroom activities, more motivated, and report increased academic self-efficacy . Academic self-efficacy captures students' belief that they are capable enough to master academic tasks and obtain successful results . Thus, in the academic contexts in which women are less certain that they belong, they may consequently feel less academic self-efficacy. Furthermore, research has shown that academic self-efficacy in an educational field predicts students' course choices . However, so far, no research has tested such causal paths between these three psychological mechanisms within the same study. Thus, the second goal of the present research is to examine within the same study how an academic culture promoting self-enhancement values may affect women's sense of belonging, self-efficacy, and course choices in both STEM and non-STEM academic settings. --- Overview of the research Past literature suggested that the culture often promoted in the STEM as well as brilliance-focused fields emphasizes self-enhancement values . However, it remains unclear whether an individualistic and competitive culture would still have a negative effect outside a STEM or "brilliance-focused" type of setting. Indeed, in these settings self-enhancement values are embedded in an already threatening environment, with women being negatively stereotyped . Therefore, the goal of the present research is to isolate the effect of values alone on women's academic experience: first, within a STEM field but in a course not negatively stereotyped for women and second, within a non-STEM field in a course framed as highly relevant and useful . We argue that depicting self-enhancement values as strongly linked to success in a given academic setting might undermine female students' expected sense of belonging, which would affect their anticipated self-efficacy, and ultimately would decrease the likelihood of integrating into the given academic setting, even if it is presented as highly relevant for their career. In contrast, due to the intrinsic nature of self-transcendence values, which foster a feeling of acceptance and integration for all students, both female and male students should report relatively high levels of sense of belonging, academic self-efficacy and interest in pursuing the presented academic opportunity in an academic culture depicting success in terms of self-transcendence values. Such a prediction mirrors the documented findings of men's positive experience in feminine occupations . Thus, we hypothesized that assigning high success to selfenhancement values would impair female students', but not male students', sense of belonging, academic self-efficacy and academic choices. Studies 1 and 2 tested this hypothesis on a population of engineering students. We considered that our hypothesis would receive stronger support if validated among women who had chosen a STEM field of study considered as especially prestigious and demanding in the French higher education system, hence embracing a career perceived as more favorable to self-enhancement than to self-transcendence values. Indeed, this will demonstrate that the negative effect of self-enhancement values remains strong even among women who chose such an orientation. We employed an experimental paradigm designed to minimize for women the potential social identity threat associated with the setting, by offering engineering students the opportunity to attend a psychology course. Moreover, in addition to lacking threatening STEM-related content, psychology is not perceived as a field in which brilliance is the primary requirement for success . In order to make the decision to participate to the course more consequential, and Procedure and materials. At the end of a regular class, the participants were told that their engineering school collaborated with the psychology department and that we were interested in their opinions on the potential implementation of a new psychology course the following academic year. After the course-description, participants either read that the students who succeed brilliantly in this course like being the ones making decisions, leading others, showing their competence and outperforming others or that the students who succeed brilliantly in this course think it is important that all persons be treated equally, and like helping other students and listening to other people's different opinions even if they do not agree . To ensure that the participants read and understood the text, they had to write down four adjectives that, in their opinion, characterized students who succeeded in this course. Next, they completed the 7-item sense of belonging scale , and the 7-item academic self-efficacy scale . Finally, the participants reported the likelihood of enrolling in this class the following academic year . --- Results and Discussion --- Manipulation check. Three blind judges estimated the extent to which each series of adjectives provided by the participants referred to self-enhancement values and self-transcendence values . The Cronbach's alpha coefficient showed good interrater reliability for both self-transcendence and self-enhancement ratings; therefore, the three --- Overview of the analyses. We conducted regression analyses to test our hypotheses. The regression model contained three predictors: condition , gender , and condition x gender. --- Sense of belonging. Regressing the sense of belonging scores on the predictors revealed a significant effect of condition, B = -0.21, 95% CI [-0.37, -0.05], F = 6.60, p = .012, ɳp² = .06. When led to believe that success in the course was achieved through selfenhancement values , the participants expected a lower sense of belonging than when led to believe that succeeding involved the endorsement of selftranscendence values . The main effect of gender was not significant, B = 0.10, 95% CI [-0.06, 0.26], F = 1.40, p = .239, ɳp² = .01. More importantly, the effect of condition was qualified by a significant condition x gender interaction effect, B = 0.19, 95% CI [0.03, 0.35], F = 5.37, p = .022, ɳp² = .05 . Analyses of simple effects indicated that the women anticipated a significantly greater sense of belonging in the self-transcendence condition than in the self-enhancement condition, B = -0.40, 95% CI [-0.66, -0.14], F = 9.54, p = .003, ɳp² = .08, whereas the men anticipated approximately the same sense of belonging across the two conditions, F < 1, p = . 838, ns. In the selfenhancement values condition, the women expected a significantly lesser sense of belonging than male students, B = 0.29, 95% CI [0.05, 0.52], F = 5.80, p = .018, ɳp² = .05, whereas in the self-transcendence values condition, female and male students did not significantly differ regarding their anticipated sense of belonging, F < 1, p = .410, ns. Academic self-efficacy. Regressing the academic self-efficacy scores on the predictors revealed that neither the main effect of condition nor the main effect of gender was significant = 1.59, p = .209, for gender). Although showing the expected means-pattern , the condition x gender interaction effect did not reach significance, B = 0.20, 95% CI [-0.04, 0.45], F = 2.77, p = .099, ɳp² = .03. --- Likelihood of enrolling in the course. Regressing the likelihood of enrolling scores on the predictors did not reveal any significant main or interaction effect . We found that in an experimentally created academic context, linking success to the endorsement of self-enhancement values undermined female students', but not male students', sense of belonging, creating a significant gender disparity. The effects expected on self-efficacy and likelihood of enrolling were not significant. One plausible explanation might be that the students' relatively busy timetable in engineering school lowered the general probability of choosing to attend a course that was neither essential nor mandatory for their training program. Furthermore, as we only had access to the students who attended the class that day, our sample sizes were quite small. To offset these limitations, we replicated Study 1 with the following methodological improvements. First, we used a-priori power analyses to determine the number of participants we needed to make sure we ran a sufficiently powered study. Second, we included in the description of the training more elements likely to appeal to engineering students . Since this modification in the content may cause the course to be seen as a "male" domain that may be threatening for women, we decided to assess students' perception of the course in terms of gender relevance. Third, because the one-item scale that we used to estimate the likelihood of enrolling in the course was quite limited in capturing participants' overall interest in the course, we improved the scale by including three more items. --- Study 2 --- Method Participants. We conducted first a power analysis using G*Power 3 to determine how many participants we needed to achieve 80% power to detect a small to medium effect . This analysis suggested that we needed 2 The results-pattern does not change when controlling for the study setting. likely, 10 = very likely; α = .93)3 . Finally, the participants rated on a 7-point scale whether the proposed training was intended more for women , for men or both . --- Results and Discussion --- Manipulation check. Two blind judges evaluated the adjective series. Three participants were excluded because their adjectives were unrelated to the actual description of the training. The interrater agreement was satisfactory for both self-transcendence and self-enhancement ratings . Supporting the efficacy of our manipulation, the adjectives provided in the self-enhancement values condition were considered to refer to self-enhancement values to a greater extent, F = 76.15, p < .001, and to self-transcendence values to a lesser extent, F = 134.52, p < .001, than those provided in the self-transcendence values condition . --- Perception of the training. A frequency analysis of this variable revealed that the vast majority of participants, namely, 89.6% of the total sample, considered that the training was intended for both men and women , while 7.8% conveyed a rather gender neutral evaluation by choosing the responses 3 and 5 on the scale4 . Overview of the analyses. As in Study 1, we contrast-coded our independent variables . Because students' age correlated with condition, standardized-age and its interactions with the independent variables were included in the final model 5 . The regression model contained seven predictors: gender, condition, condition x gender, age, and the three interactions with age. --- Sense of belonging. Regressing the sense of belonging scores on the predictors revealed a significant effect of condition, B = -0.24, 95% CI [-0.34, -0.14], F = 23.71, p < .001, ɳp² = .10. When led to believe that success in the course was achieved through selfenhancement values , the participants expected a lower sense of belonging than when led to believe that succeeding involved the endorsement of selftranscendence values . The main effect of gender was not significant, B = 0.05, 95% CI [-0.05, 0.15], F = 1.11, p = .292, ɳp² = .01. As expected, we obtained a significant condition x gender interaction effect, B = 0.12, 95% CI [0.02, 0.22], F = 5.71, p = .018, ɳp² = .03 . Analyses of simple effects indicated that the female students anticipated a significantly greater sense of belonging in the self-transcendence condition than in the self-enhancement condition, B = -0.36, 95% CI [-0.50, -0.22], F = 24.79, p < .001, ɳp² = .10. The difference was not significant for the male students, B = -0.12, 95% CI [-0.27, 0.01], F = 3.28, p = .071, ɳp² = .01. Furthermore, in the selfenhancement values condition, women expected a significantly lower sense of belonging than men did, B = 0.17, 95% CI [0.03, 0.31], F = 5.63, p = .019, ɳp² = .03, whereas in the self-transcendence values condition, female and male students did not significantly differ regarding their anticipated sense of belonging, F < 1, p = .333, ns. We also obtained a significant interaction between age and condition, B = -0.15, 95% CI [-0.27, -0.04], F = 6.54, p = .011, ɳp² = .03. In the self-enhancement condition, the older the participants, the lower their expected sense of belonging, B = -0.22, 95% CI [-0.41, -0.04], F = 5.61, p = .019, ɳp² = .03. In the self-transcendence condition, age was unrelated to sense of belonging . Academic self-efficacy. Regressing the academic self-efficacy scores on the predictors revealed main effects of condition and gender. When led to believe that success in the course was achieved through self-enhancement values , the participants anticipated lower academic self-efficacy than when led to believe that succeeding involved the endorsement of self-transcendence values , B = -0.18, 95% CI [-0.32, -0.04], F = 6.79, p = .010, ɳp² = .03. Overall, the male students anticipated higher academic self-efficacy than the female students , B = 0.14, 95% CI [0.004, 0.28], F = 4.09, p = .044, ɳp² = .02. The condition x gender interaction effect was significant, B = 0.21, 95% CI [0.07, 0.34], F = 8.99, p = .003, ɳp² = .04 . Analyses of the simple effects indicated that the female students anticipated higher academic self-efficacy in the self-transcendence condition than in the selfenhancement condition, B = -0.39, 95% CI [-0.58, -0.19], F = 14.79, p < .001. ɳp² = .06. The male students anticipated comparably the same academic self-efficacy across the two conditions . Moreover, in the self-enhancement values condition, the women expected lower academic self-efficacy than men did, B = 0.35, 95% CI [0.15, 0.54], F = 11.95, p = .001, ɳp² = .05, whereas in the self-transcendence values condition, female and male students did not significantly differ regarding their expected academic selfefficacy, F < 1, p = .479, ns. Finally, although the interaction between age and condition was significant, B = -0.19, 95% CI [-0.35, -0.02], F = 5.14, p = .024, ɳp² = .02, the simple effects of age did not reach significance . --- Likelihood of enrolling in the course. Regressing the likelihood of enrolling scores on the predictors revealed a significant effect of condition, B = -0.54, 95% CI [-0.83, -0.25], F = 13.28, p < .001, ɳp² = .06. When led to believe that success in the course was achieved through self-enhancement values , the participants reported a lower interest in the course than when led to believe that succeeding involved the endorsement of self-transcendence values . The main effect of gender was not significant, F < 1, p = .531, ns. We obtained a significant condition x gender interaction effect, B = 0.48, 95% CI [0.19, 0.77], F = 10.78, p = .001, ɳp² = .05 . Analyses of simple effects indicated that the female students reported a greater interest in the course in the self-transcendence condition than in the self-enhancement condition, B = -1.02, 95% CI [-1.44, -0. Mediation analyses. To test the hypothesized mediated moderation model , we used the PROCESS macro for SPSS , in which we entered the interaction term between condition x gender as the X variable, and the main effects of gender and condition in the covariates window. Age and its interactions with the independent variables were also included as covariates. The interaction effects of condition x gender on likelihood of enrolling and sense of belonging have already been established as significant. The effect of sense of belonging on academic selfefficacy, when controlling the condition x gender interaction effect , was also significant, B = 0.61, 95% CI [0.45, 0.78], t = 7.29, p < .001. Finally, the effect of academic self-efficacy on likelihood of enrolling, when controlling the effects of sense of belonging and condition x gender interaction , was significant, B = 0.51, 95% CI [0.25, 0.77], t = 3.91, p < .001, whereas the interaction effect between condition x gender on likelihood of enrolling was not, B = 0.23, 95% CI [-0.02, 0.47], t = 1.83, p = .068. Moreover, the indirect effect of condition x gender on likelihood of enrolling through the sense of belonging and academic self-efficacy in the sequence was significant, B = 0.04, 95% CI [0.01, 0.09]. Taken together, Studies 1 and 2 consistently demonstrated that depicting selfenhancement values as strongly linked to success undermined female students', but not male students', sense of belonging. Study 2 went further in showing that such an association also undermined female students' academic self-efficacy, and interest in the course. Importantly, the fact that the course proposed to the participants was a psychology course, a field that is not negatively stereotyped for women and which was not perceived as a male domain, allowed for an independent test of the effect of values. Finally, we obtained an interesting result involving students' age. In an academic culture depicting self-enhancement values as highly linked to success, the older the students the lower their anticipated sense of belonging and their interest in the course. This may be explained by the fact that the older the students, the more they have been exposed to such a culture, generating more avoidance tendencies. Although this effect applies to both male and female students, it does so in an additive manner for women only, because, independent of age, they are affected more by the self-enhancement culture compared to men. We argued that similar selection processes should apply to all fields of study. Thus, even a "social" field of study could develop a culture of achievement based on self-enhancement values, and thus create a challenging academic context for female students. Therefore, the goal of Study 3 was to test the generalizability of the findings among psychology students and to highlight further the academic relevance of the findings. Moreover, the results of Studies 1 and 2 illustrated that, as expected, male students were unaffected by whether the achievement culture in the course emphasized self-transcendence or self-enhancement values. Thus, we conducted Study 3 on female psychology students and framed the psychology course as highly important for their career. --- Study 3 --- Method Participants. Based on a medium effect size , as suggested by the effect sizes obtained in Study 2 for women, a G*Power 3 analysis indicated that we needed at least 93 participants to achieve 80% power. Ninety-seven first and second year female psychology students completed the study . Procedure and materials. We used the same experimental paradigm and framed the psychology course as highly relevant for today's job market. Next, students completed the sense of belonging scale , academic self-efficacy scale , and estimated their likelihood of enrolling in this class the following academic year . In order to reinforce the perception of an actual academic choice they were about to make, the students were asked to report their name and academic identification number. --- Results and Discussion Manipulation check. Three blind judges evaluated the four adjective series with course was achieved through self-enhancement values , the participants were less likely to enroll in the course than when led to believe that succeeding implied endorsing self-transcendence values , B = -1.02, 95% CI [-1.50, -0.56 --- ], F = 18.78, p < .001, ɳp² = .17. --- Mediation analyses: What model best captures the path from condition to likelihood of enrolling? We hypothesized that that a culture of achievement based on self-enhancement values ultimately affects female students' academic choices through its effect on students' anticipated sense of belonging and academic self-efficacy. However, given our cross sectional data, we could not exclude the alternative, that the path from condition to likelihood of enrolling for the course flows through anticipated academic self-efficacy and expected sense of belonging . Thus, in order to assess which of the two models better fits our data, we conducted path analyses using a maximum likelihood estimator in R. To test our hypothesis , we ran a model with estimated paths from condition to sense of belonging, from sense of belonging to academic self-efficacy and from academic self-efficacy to likelihood of enrolling in the course. Moreover, to show full mediation, we constrained the direct path from condition to likelihood of enrolling in the course to zero. Additionally, to depict full meditation of the effect of condition on academic self-efficacy through sense of belonging, we also constrained the direct path from condition to academic self-efficacy to zero. Based on previous studies documenting a direct effect of sense of belonging on academic choices , we allowed for the direct path from sense of belonging to likelihood of enrolling to be estimated. Supporting our hypothesis, this model showed good fit to the data, χ 2 = 4.09, p = .130, CFI = .99, RMSEA = .10, 90% CI [.00, .25] 6 , and SRMR = .05 . To test the conceptual alternative model , we shifted the place of sense of belonging with academic self-efficacy. As in Model 1, to reflect full mediation, the direct path from condition to likelihood of enrolling was set to zero. Moreover, to mirror Model 1, we set the direct path from condition to sense of belonging to zero, while allowing for the direct path from academic self-efficacy to likelihood of enrolling to be estimated 7 . Model 2 showed poor fit to the data, with a statistically significant χ 2 test, χ 2 = 7.66, p = .022. These complementary sets of analyses support the directionality of our hypothesized model. Depicting self-enhancement values as strongly linked to success undermined female students' sense of belonging, academic selfefficacy, and corresponding choice regarding enrolling in the course. As the course was framed as a clearly useful and relevant opportunity for their future career, this result effectively illustrates the dynamic through which, even in a non-STEM field of study, women opt out of academic possibilities, with a potentially decisive impact on their professional career, when they perceive that they have to endorse self-enhancement values to succeed. 6 It is important to note that the RMSEA tends to be substantially elevated for correctly specified models with small degree of freedom . In fact, the authors recommend completely avoiding computing the RMSEA when model dfs are small, and particularly when the χ 2 test is not statistically significant, since, in such a case, we know that the model relatively closely reproduces the data. 7 Given that in the alternative model we did not have a specific prediction about the degree of mediation for the effect of condition on sense of belonging through academic selfefficacy, we also ran Model 2b, in which we included a direct path from condition to sense of belonging. Model 2b also showed relatively poor fit to the data, χ 2 = 2.98, p = .084. --- General Discussion Women's academic attainment in higher education is qualified by a certain paradox. Despite the fact that a higher proportion of women than men attend university, women's psychological experience in various academic contexts is qualified by decreased sense of belonging, and academic self-efficacy, even in non-STEM settings, contradicting interpretations in terms of social identify threat only . We proposed that a possible explanation may be located at a cultural level, namely in the emergence, within an increasingly selective higher education system, of an achievement culture based on self-enhancement values. Because of gender differences in the endorsement of self-enhancement values , we expected that an academic setting depicting success in terms of self-enhancement values would diminish female students' sense of belonging, academic selfefficacy and interest in pursuing the given academic course. Indeed, we showed that independent of fields, STEM or non-STEM , an academic culture depicting self-enhancement values as highly linked to success had a negative impact on female students' psychological academic experience through sense of belonging , self-efficacy and academic choices . This is important because even a "social" field of study, in which the specific goals afforded by the future career , may be close to self-transcendence values, could still develop a culture of achievement based on self-enhancement values. This may be the case because similar selection processes apply to all fields of study, STEM or social sciences . Moreover, by presenting female engineering students with a course that does not pose a particular threat to their gender identity, we assured that the effect is due to the value-mismatch. Thus, our findings nicely disentangled the effect of values from the effect of social identity threat, which are relatively intertwined in the literature. Furthermore, we extended the goals congruency hypothesis by highlighting a complementary academic dynamic in which the concept of values is essential, but which is orthogonal to the students' primary choices for specific fields of study. In other words, both students who chose either a STEM-related or a female-dominated field of study may be confronted during their studies with a culture of success emphasizing self-enhancement values. Understanding how students navigate this climate is of particular importance. Thus, an additional contribution of the present results is to depict the psychological mechanism underlying students' academic choices. Based on the literature , we construed a mediational process engaging as key variables sense of belonging and academic self-efficacy. Although the cross-sectional nature of our data demands caution, the various statistical techniques employed support the idea of sense of belonging affecting academic self-efficacy, which ultimately shapes students' interest in different academic courses. The effect on self-efficacy, which was modeled as fully mediated by sense of belonging, is particularly interesting since it reveals that women's belief in their capacity to master academic tasks and obtain successful results is affected by cultural features of the setting unrelated to the actual content of the course. --- Limitations and perspectives In the current research, we argue that self-enhancement values are associated with a high level of academic success, notably because of the selection function inherent in the functioning of the higher education system. However, we have not fully addressed the direct causal relationship between the selection practices of the university and the perception of the academic culture as being based on self-enhancement values. For instance, showing that the link between self-enhancement values and academic success is stronger when the type of evaluation in the course is based on relative performance as opposed to personal improvement or in fields practicing numerus clausus would nicely test this relationship. Despite these limitations, our findings have significant practical implications. In particular, given the selection processes that take place in most European academic systems all throughout the undergraduate level, an academic culture in which self-enhancement values are emphasized is rather common, even in a non-STEM field of study. For example, the fact that a program is "highly competitive" is often presented as an argument to support its quality. Such messages imply a display of selfenhancement values and, as suggested by our findings, could prevent female students from taking full advantage of their academic opportunities, even in a field stereotypically in accordance with their predominant values and goals. This is particularly important because it suggests that all things being equal, women may have a more challenging academic experience, even if they perform as well as their male counterparts ). Women's generally good performance in college may minimize and even conceal their more subjective experience in terms of sense of belonging or self-efficacy, with significant downstream consequences for their academic choices during their academic years. Finally, in the spirit of self-transcendence values, higher educational institutions could certainly gain from implementing collaborative projects framed with a prosocial purpose. For example, the opportunity of working together toward common ends was found to increase motivation for challenging tasks, as well as greater interest in and enjoyment of the task . Our results suggest that such collaborative projects might also provide all students with a more positive academic experience. --- satisfying inter-rater agreement for both self-transcendence and self-enhancement ratings . Confirming that our manipulation worked, the adjectives provided in the self-enhancement values condition , the participants anticipated a lower level of academic self-efficacy than when led to believe that succeeding implied endorsing selftranscendence values , B = -0.39, 95% CI [-0.60, -0.17], F = 12.60, p = .001, ɳp² = .12.. --- Likelihood of enrolling in the course. Regressing the scores of likelihood of enrolling in the course on condition revealed that when led to believe that success in the

Although overall women are better represented in higher education than men, women's psychological experience in various academic contexts is qualified by a decreased sense of belonging and academic self-efficacy, including in fields where they are not targeted by a negative stereotype. To clarify this phenomenon, we develop the hypothesis of a mismatch between female students' values and the values associated with success in the increasingly selective realm of higher education. We argue that, whatever the fields of study, these values are self-enhancement values (competitiveness, self-affirmation, dominance). Three studies showed that when success was depicted in terms of self-enhancement values, women-but not men-expressed a lower sense of belonging, had lower self-efficacy and were less likely to pursue a given academic opportunity both in STEM and non-STEM fields of study. These effects did not appear in an academic context depicting success as being rooted in selftranscendence values (helpfulness, cooperation, benevolence).

Background There is no shortage of media coverage following disasters, and no shortage of criticisms of the media coverage following disasters. It is a familiar pattern: the first media wave reporting on the event; the second wave of clarifications and professional opinions; then the third, a stream of splashy, sensationalist pieces that play towards a hungry audience's desire for more information. It all becomes so big and blurry. This is my memory of the media coverage of the 2011 earthquake, tsunami, and nuclear plant explosion in Fukushima, Japan, and my only concept of Fukushima prior to visiting for research. News stories of contaminated fish, mutated plants, terrified and irradiated evacuees. There was no sense of personal narrative -people lost control of their own stories. The same is true in the academic literature; 8 years later, ethnographic studies of the Great East Japan Earthquake are extremely few. During my time at Fukushima Medical University , my hope was to hear the experiences of students who were there when it happened. As half of my stay in Fukushima overlapped with the university's exam period, I imagined that students would be reluctant to speak with me for fear of compromising their studying, and that the interviews I did manage to schedule would be very brief. However, the students were more than eager to share both their time and their stories. While I initially came to FMU to pursue questions of ethics in post-disaster research, I quickly found that the participants were more passionate about other topics related to the Great East Japan Earthquake, and I was excited for the discussions to shift. Therefore, a new research focus emerged and I simply began asking participants, "What was it like?" This inductive and discovery-based approach allowed me to reframe and refine the objectives for this study. While there continues to be extensive news coverage and quantitative studies on Fuksuhima Prefecture's recovery, there remains a paucity of scientific literature on individual narratives during and following the GEJE. A holistic understanding of such an impactful event requires investigations from both qualitative and quantitative perspectives. Through open-ended interviews, this study aims to illuminate individual stories of medical students at FMU who lived in the Prefecture at the time. Here, we utilize a qualitative approach based in ethnography to delve into their complex lived experiences, explore their responsive belief systems and coping strategies as a result of the Earthquake, and lastly provide a space in which survivors can control the narrative in academic research. In the afternoon of March 11 th , 2011, a magnitude 9.0 earthquake off the coast of Japan's Tōhoku region induced a major tsunami. This earthquake -the Great East Japan Earthquake -was the most powerful ever recorded in Japan, and the fourth most powerful recorded worldwide [1]. The resulting 6-8 m tsunami hit areas of Iwate Prefecture, Miyagi Prefecture, and Fukushima Prefecture , causing enormous loss of life, as well as tremendous destruction of buildings, roads, and railways. As of December 10, 2019, the National Police Agency of Japan has declared 6,157 injured, 2,529 missing, and 15,899 dead [2]. As it reached the Fukushima Daiichi Nuclear Power Plant in Ōkuma, Fukushima Prefecture, the tsunami swept over the plant's seawall, flooding the lower grounds and knocking out the emergency generators, which were required to cool the reactors. In the absence of cooling, 3 of the plant's 4 units went into nuclear meltdown resulting in three hydrogen explosions which released radioactive material [3]. In the days following the accident, radiation released into the atmosphere lead the government to declare an evacuation zone around the plant, which grew to reach a 20-km radius [4]. An estimated 154,000 residents evacuated from the towns surrounding the plant to avoid contamination and relocated throughout Japan, principally in the central and western areas of Fukushima Prefecture [1]. There were significant challenges in evacuating patients from hospitals and nursing homes: the deaths of 51 "seriously ill patients" -e.g. terminally ill, suffering from dementia, requiring dialysis, and/or immobile -were attributed to insufficient transportation, inadequate medical facilities, lack of accompaniment by medical staff, improper heating, inappropriate seating for patients with limited mobility, and extremely prolonged travel due to the tsunami's destruction of roads and bridges [5]. As of 27 February 2017, the Fukushima Prefecture government identified 2,129 "disaster-related deaths [6]" -this figure does not distinguish between the deaths of people displaced by the nuclear disaster compared to those displaced by the earthquake or tsunami, but media analysis suggests that 1,368 of these deaths are attributable to the nuclear evacuation [7]. While today, the evacuation zone is gradually shrinking and some residents are returning to their homes for the first time in 8 years, many still remain in temporary housing, or else have decided not to go back. Despite ongoing fear regarding the potential health effects of the disaster, studies have shown no increase in miscarriages, stillbirths, developmental disorders in babies born after the accident, nor have they found a link between the nuclear disaster and incidence of thyroid cancer within Fukushima Prefecture [8]. However, the results of a large-scale mental health and lifestyle survey conducted throughout the prefecture by Fukushima Medical University indicated post-traumatic psychological distress among residents of Fukushima Prefecture, both in evacuees and non-evacuees, as well as increased rates of obesity, binge drinking and alcohol use disorder, depression, and suicide. The rate of psychological distress was noted to be particularly high in children and adolescents compared to the background population [9]. These effects are more prevalent among evacuees, who face "radiation stigma" -stigma due to the fear that these individuals may contaminate others, or that they might bear children with severe birth defects due to potential exposure [10]. --- Methods --- Context In the academic literature written in English, the events of March 2011 are referred to collectively as "the Tōhoku disaster, " "the Fukushima disaster, " or "the triple disaster, " or else as variations on "3/11 [11]. " However, in order to be consistent, accurate, and respectful, I will to refer to the earthquake-tsunami, the subsequent incident at the power plant, and the repercussions of all three components of the event in the same way that the research participants referred to them: the Great East Japan Earthquake , or simply, the Earthquake. The Icahn School of Medicine at Mount Sinai and Fukushima Medical University have enjoyed a partnership since 2012 -each spring, two third-year students from Fukushima Medical University come to the Icahn School of Medicine to study endocrinology, and each summer, two students from the Icahn School of Medicine conduct research at Fukushima Medical University. A major goal of this partnership is cross-cultural exchange: students from each school have the opportunity to learn about the healthcare system, current policy issues, and the lifestyle of medical students and doctors in their host country. As I was a student at the Icahn School of Medicine at the time of the study, this existing relationship paved the way for my participant recruitment. --- Study population and recruitment Interview participants were limited to students enrolled at Fukushima Medical University who lived in Fukushima Prefecture at the time of the Great East Japan Earthquake. Medical school provides students with a medical degree and general medicine education with which students will go on to specialize in residency. Medical education in Japan lasts 6 years and begins after high school. As a result, most students are generally somewhere between 18 to 24 years of age. In a student body of 792 students as of May 1 st , 2019, only 32% came from Fukushima Prefecture [12]. Exponential non-discriminative snowball sampling, where each participant recruited could provide one or more referrals, was the main method of participant recruitment. Initially, participants were invited to participate via announcements made at the beginning of classes for year 1 to 6 students as well as through class group chats in the LINE mobile messenger app. Two reminders, spaced two weeks apart, were sent in the class group chats. The first wave of recruitment garnered 5 participants. These participants then provided recommendations for other students to attempt to recruit, garnering an additional 3 participants in the second wave of recruitment. Concurrently throughout both waves of recruitment, members of a disaster medicine interest group called "Fukushima WILL" reached out to their club members. This garnered 2 participants. Of the initial group of 253 students, 10 students agreed to engage in interviews. There were no limits on the number of participants for the study. The natural endpoint for participation coincided with the conclusion of the exchange program . Participants were in years 1-5; no 6 th year students from Fukushima Prefecture were available to interview due to demanding curricular requirements during the summer months. --- Interviews Interviews were conducted from June to August 2019. Interviews were open-ended and an interview guide was used . The interview guide was developed by the authors, including two physicians , two medical students , two Japanese medical students at FMU, and a Japanese faculty member at FMU . The interview guide was updated throughout the study as needed. The goal of the open format was to encourage the participants to reveal and discuss what they wanted to reveal and discuss, rather than what the researchers wanted. Interviews took place in study rooms and classrooms on FMU's campus. Informed consent to participate was obtained both in English and Japanese at the start of the interview. Written consent for recording and publication was obtained as well. All but one of these interviews was private -unfortunately, a private room could not be reserved for Participant 2, and there were other students in the study room at the start of the interview. Additionally, Participants 8 and 9 requested they be interviewed together, and brought three friends with them who wanted to observe. Though the researcher verified many times prior to the start of the interview that this was alright with both participants, they insisted that they would rather have their friends stay. Ultimately, the researcher favored participants' comfort over a "sterile" interview environment. Following consent, participants were provided with the choice of conducting the interview in English or Japanese. The choice of language depended on both the participant's comfort with speaking English and their desire to practice. Interpretation from Japanese to English and vice versa was provided telephonically by a language access company, Pacific Interpreters. Interpreters hired by the company have passed both language proficiency and interpreter readiness tests in English . Throughout the interviews, the interpreter would convert from Japanese to English or vice versa in real time. Interpreters would sometimes ask clarifying questions to either the English or Japanese speaker to more accurately translate the meaning of the speaker's words. The researcher did not perceive a difficulty in understanding or being understood by any of the participants. During one interview involving Participants 8 and 9, an telephone interpreter through Pacific Interpreters was unavailable. Instead, a friend observing who was fluent in English served as an interpreter for Participant 8. Of the 10 participants, 5 conducted the interview in Japanese via an interpreter, and 5 conducted the interview in English. Interview times varied from approximately 30 min to over 3 h, but the average session length was around 1.5 h. All interviews were audio recorded and were later transcribed by a researcher . Only the English language portions including words spoken by the interviewer, interpreter, or participant were transcribed . A second researcher listened to all the audio recordings along with the transcribed document to ensure no errors in transcription were made. Throughout this study, interview participants are referred to by the order in which they interviewed -for example, the third participant to be interviewed is called "Participant 3" -in an attempt to anonymize participants. The researchers recognize the relative futility of this effort; as anthropologist Rebecca Nelson points out in her wonderfully titled essay "How Can We Hide Participants' Identities When They're on Pinterest?, " the ubiquity of social media and the necessarily specific identity markers of these interview participants allow curious readers to "play detective, " making it difficult to respect the privacy and ethical commitments made to participants [13]. Therefore, personal and identifying information is withheld from the text whenever possible. All participants, regardless of the language spoken, are quoted in first-person -translators from Pacific Interpreters relayed the words of Japanese-speaking participants directly -with the exception of Participant 8, who was interpreted in third person rather than first. --- Methodological approach, data analysis, and rigor The topics discussed in this text are the result of an inductive thematic analysis of the interview transcripts under the lens of ethnographic anthropology, so chosen for the discipline's goal of exploring cultural experiences from the point of view of a study's subject rather than from the perspective of the researcher. The researchers began the analysis while collecting the data. The inductive thematic analysis process comprised the following steps: familiarizing oneself with the data, generating initial codes, defining and naming interpretative codes for the entire data set into themes, identifying patterns across all data to derive themes for the data set, and lastly defining and naming themes. AS, ML, CK, and RY were involved in all steps of the thematic analysis and served as reviewers. This measure minimized the risk of a researcher's individual bias, personal preconceptions, or positionality overly influencing the data analysis. Team meetings were organized frequently to share the research path and reframe objectives as needed. Credibility was obtained through prolonged engagement of all the researchers with the setting . Credibility was also assured by gathering rich data from the interviews, transcribing the interviews verbatim, and analyzing the data line by line. --- Findings Three themes naturally emerged in the inductive analysis: first, that the events following the Earthquake influenced not only these students' decisions to become doctors, but also the ways in which they hope to practice medicine in the future . Second, that these students were motivated to share their experiences by a want to change Fukushima's public image . And lastly, that the students viewed the opportunity to discuss their experiences through these interviews as healing, both for themselves and for the future . --- Effect on future medical practice All of the interview participants attributed some part of their budding medical careers -be it the medical specialties they're interested in or their wish to study or practice in Fukushima -to their experiences following the Earthquake. For some of the students with whom I spoke, the Great Earthquake acted as a catalyst of sorts in their decisions to pursue a career in medicine. "I think, " Participant 1 observed, "what you did before you were medical student maybe makes you what you are today. " Indeed, of the 10 students interviewed, 6 cited their experiences following the GEJE as a motivating factor in applying to medical school. Qualitative evidence supports this trend. A 2015 survey of FMU medical students who had engaged in recovery volunteer work showed that these student volunteers displayed a statistically significant increase in desire to become a physician compared to non-volunteers. Authors Anderson et al. acknowledge that in the years that elapsed between the Earthquake and the survey, volunteers and non-volunteers alike certainly had other experiences that influenced their professional goals [14]. However, some of the students interviewed for this study explicitly connected the Earthquake to their chosen career paths. Some, such as Participant 1, expressed an interest in medicine prior to the Earthquake, but noted that this interest was heightened by the disaster. Participant 1 explained, "…After the earthquake, partly because my father's job is doctor, I thought that medical is essential for daily life, I thought about it stronger than before the earthquake. That's why I thought that I want to be a doctor. " Participant 5 similarly reported a strengthening of their pre-existing interest in medicine by the Earthquake. They explained, "I learned through this experience [of the Earthquake] that one of the most important things is helping people. That's the biggest thing I learned through the disaster. I think it's connected to what I want to do in the future. " Other student interview participants viewed their experiences following the Earthquake as essential in their choice of career -Participant 8 is one such example. Though Participant 8 had worked in healthcare prior to the Earthquake as a nursing home aide, they decided to apply to medical school after coming into contact with evacuees from the exclusion zone, including the displaced residents of another nursing home in Namie, a town directly downwind of the Fukushima Daiichi Nuclear Power Plant. Participant 8 explains,, "As a fellow citizen of Fukushima Prefecture, [Participant 8] was always thinking about the…evacuation zone. The people in the evacuation zone, he want[ed] to do something for them, but he [had] a daily life, and work…so he [couldn't] do enough for them. " Participant 2 was the most emphatic in describing the Earthquake's role in her choice of career. She recounted, "Before the Earthquake happened, I didn't know what I wanted to do in the future, I had no goal or dream. But then I'd seen so many friends die, and I went through that experience, and that made me want to save lives. " She went as far as to say, "I think that if I didn't experience the accident, there's no way I would have gone into medicine. " The aftermath of the Great Earthquake affected not only her decision to pursue medicine, but also her priorities as a future healthcare practitioner. After the Earthquake, she and her friends experienced several years of flashbacks and panic episodes triggered by images and videos of the disaster. "But what saved us, " she said, "was school counselors, doctors, and mental health professionals. So I [want] to be a doctor who can help, not just medically, but mental health as well. Not just physical health, but mental health as well. That's really important. " This emphasis on an integrated approach to medicine, particularly for children and adolescents, was shared by several of the participating students. Participant 1, who is not at present planning on going into psychiatry, expressed the importance of incorporating psychiatry into his future medical practice: --- It's important thing to have psychiatric mind, psychiatric comprehension skills. How to feel sympathy, how to open [patients'] minds, how to icebreak. I think, especially for child, like adolescent age…I think it's kind of like 100% skill of a doctor. It's essential to me. Psychiatry is important to every medical field…I strongly think so because I can emphasize on people's depression and people's negative minds. Especially in people who struggle in harder situations, like attacked by not only nature disaster but also violence. Maybe I can feel sympathy with them. I want to be a different doctor. Participant 1 describes most doctors as being "100% medical mind, " his shorthand for a scientific approach to the exclusion of the personal, but he describes himself as being "less medical than [his] classmates" at only 90% "medical mind, " a trait he attributes to living in Fukushima after the Earthquake. It is this trait that he hopes to use to become "a different doctor:" Majority of my classmates, I think…are kind of like more concentrated on their medical things, more than me. They tend to avoid to talk with ordinary people [who aren't in medicine]…Students totally forget about the normal mind because of the exposed by the medical things. But to have sympathy with "normal" people, maybe we have to still have a kind of normal mind about, like, everything. So I don't think 100% medical mind is a good thing. Participant 9 describes a similar mindset that they attribute to their experiences as a post-disaster recovery volunteer in Iwaki, the site of the Earthquake's epicenter. Of those displaced by the Earthquake, Participant 9 states, "It's very important to do healthcare for them. It's also very important to chat with them, to talk with evacuees, about a lot of things -for example, their families, their hobbies, and so on -while doing treatment. " While he states the Tōhoku Earthquake did not influence his decision to pursue medicine, he says it did change his field of interest from surgery to family medicine [15]. He explained, --- When I first wanted to be a doctor, I thought "It's cool to become a surgeon, very cool. " But after the big earthquake, my feeling changed because a lot of people were affected by a lot of hardship. If I become a surgeon, I will help one person per day. But if I become a home doctor, I will treat many people. All of the students with whom I spoke described a relationship between their medical field of interest and their experiences living in Fukushima following the Earthquake. While none of these students will have to decide until after they graduate from medical school, their current thoughts on specialties reveal the lasting impact of the Earthquake. Participant 7 is currently doing research in lifestyle medicine and preventative care, which she says "also includes what to do after a nuclear incident, so I guess I'm ending up studying life after the nuclear accident, because I think it's important to know how to prevent risks of being sick after that. " As a specialty, she's interested in "psychiatry, because you have to know how the disaster affects the mind. " Participant 9 is interested in family medicine for the same reason as his friend Participant 8. Participant 4 is considering pediatric psychiatry, inspired by loved ones who struggled with their mental health following the Earthquake -"I want them to know that they don't need to be ashamed, " she declared. Participant 3 described a similarly personal motivation for his field of interest. Due to space constraints, his was the only interview conducted in public, and, though nervous about his level of proficiency, he requested we switch our conversation to English for this portion of our discussion. His friends, none of whom spoke English, were sitting at another table on the other side of the room, and he glanced at them before sharing, After the disaster, I saw sick people who couldn't be treated properly, so I thought, you know, I want to provide proper care for sick people… My mother… had cancer. And the time is just when the earthquake happened. So when the earthquake happened, lots of doctors went away to other prefectures, so Fukushima Prefecture had little doctors, so...there were no people, she couldn't get treatment… And later -she was gone. She died. This difficult experience was the impetus for Participant 3's interest in a career in regenerative medicine. He explained, "Because of this disaster, we had less doctors, so patients couldn't be treated properly. Like my mother. So the development of new technology would help patients treat themselves, by themselves, for themselves. " Several participants also expressed a desire to stay in Fukushima to practice medicine once they graduate, out of a want to help heal their prefecture. While this is not unique to these participants -56% of Fukushima Medical University's last 5 graduating classes have stayed in the prefecture after graduation for their medical internships [12] -the students interviewed specifically cited a wish to help their home communities recover. Participant 2 says she wants to pursue internal medicine "since I want to go back and help the locals. " Participants 8 and 9 are both hoping to become home-visit family doctors because "Fukushima needs home doctors. " Fifth-year student Participant 6 felt that the Earthquake so transformed his plans in practicing medicine that he confided, "I think that 3/11 is good experience for my life. " He went on, [The Earthquake] was the trigger of my studying radiation and [now] I want to know more about Fukushima. If there were no disaster, I would want to go to some big city, but there was a disaster, so now I want to help Fukushima. It was a great experience for me. But someone who had big damage by the disaster would think that it's an imprudent point of view. --- Righting Fukushima's public image This desire to help Fukushima heal extends beyond the literal. Many individuals I spoke with in Fukushima, both in the context of interview and out, lamented the fear that many outside of the prefecture feel towards Fukushima. As Participant 2 put it, "…People, especially overseas, around the world, they think that Fukushima is a very dangerous place to live, still, and that everyone who lives in Fukushima is sick. But that's not true. Ideas like that, they should be fixed. " I have encountered this line of thinking myself, both in the United States and in Japan. In planning my visit to Fukushima, friends -including my classmates in medical school -asked me if it was safe to be there. A bartender in Osaka confided that she would never visit Fukushima, and that her friends from Korea have even asked her if visiting Tokyo is safe, given its "proximity" to Fukushima. A few weeks into my stay, in looking for the address of a shop in downtown Fukushima called Daisy Bell, I made the mistake of running a Google search for "Fukushima daisy. " Instead of the address I had hoped for, my search returned articles with titles such as "These 'Mutant Daisies' Near Japan's Nuclear Disaster Site Are Freaking Everyone Out [16], " flanked by images of a group of daisies that appear elongated, like several flowers merged together . They came to the internet via Twitter user @San_kaido, who uploaded them in May 2015 with the caption "The atmospheric dose is 0.5 μSv/h at 1 m above the ground [17], " which lead many viewers to assume that the mutations displayed in the photograph were caused by nuclear radiation from the Fukushima Daiichi Nuclear Power Plant. In reality, such mutations occur naturally worldwide, without the influence of radiation. [18]. These images, accompanied by the claim that the flowers shown had "nuclear birth defects, " were used in a 2016 study by the Stanford History Education Group to investigate the ability of Usonian high school and college students to evaluate the credibility of information that exists on the internet [19]. According to study co-director Sarah McGrew, "Most students -the vast majoritynoticed none of those problems" [18]. Eighty percent of Fig. 2 Viral photo of fascination mutation in daisies https:// www. twitt er. com/ san_ kaido/ status/ 60351 33719 34130 176. Used with permission from Twitter user @san_kaido high school students accepted the picture as proof of the continued effects of radiation in Fukushima [19]. The public's fixation on Fukushima as "the ultimate meme of nuclear power [20]" is still going strong, fed by an eager media. In April 2018, British tabloid newspaper The Sun ran a story with the outrageously dramatic headline "Japan 'covering up' Fukushima nuclear dangerzone radiation levels and blackmailing evacuees to return to radiated areas swarming with radioactive pigs and monkeys [21]. " In July of that same year, Netflix aired its original docuseries Dark Tourist, which follows journalist David Farrier as he visits, in the words of the show's Netflix Official Site, "unusual -and often macabre -tourism spots around the world [22]. " Episode 2, "Japan, " begins by following Farrier on an organized bus tour in an unspecified location near the Fukushima Daiichi Nuclear Plant as Farrier wonders in the narration, "Is it even safe to be here?" He and his fellow "nuclear tourists" are then shown in Tomioka, anxiously watching the numbers on their Geiger counters rise, and a woman comments, "Oh my god, the level of radiation, that's higher than around Chernobyl where no one's allowed to go. " The group later has lunch in Namie, where Ferrier says of the food, "It may be radioactive, but it's delicious [23]. " While it is not within the scope of this research to disprove the false rumors on the safety or lack thereof of Fukushima, it's important to point out that the high radiation levels observed by Farrier and his fellow tourists reflect the exposure they would have received in an hour, and that the threshold for sickness against which they compared their readings is the amount of continuous exposure needed per hour to cause damage over the course of a year. In fact, Farrier likely received more radiation flying from Japan to the United States to film his next episode than he did on his tour in Fukushima [24]. The Fukushima Prefectural government and the Reconstruction Agency are upset by more than Farrier's cavalier attitude towards the Tōhoku Earthquake ; they reported considering legal action due to concerns that "the video could fuel unreasonable fears related to the March 2011 disaster [25]. " The episode has certainly had an impact on the nuclear tourism community: despite the government's efforts, an estimated 100,000 foreign visited Fukushima in the year following the show's release, many attracted by the opportunity to see what Farrier called "a nuclear wasteland [23,26]. " In reality, it's not, and the Fukushima Prefectural government has been working hard to disabuse this notion. However, the wide-reaching impacts of the sensationalism and misinformation following the Earthquake have been so severe that some have dubbed the so-called "harmful rumors" the "fourth disaster" of the Triple Disaster [27]. The widespread view of Fukushima as dangerous, its food as radioactive, and its people as sick or even mutated is familiar to the interview participants. "Rumor itself, " Participant 8 said, "is a very evil thing, especially when there are a lot of vulnerable people in Fukushima. " Tabloid and gossip media, he argued, were equally "evil. " Participant 7 was in agreement. She observed, --- A lot of people outside Fukushima have fears about radiation, and I think those fears exaggerated and became rumors, all over the place. There was so much information after the nuclear accident. The media picked up so many stories. But I don't think many people visit the government site [to verify information], but I don't know. I'm sure some people went to the government site to get information, but there was just so much information on the media, and this is where many people went for their infor-mation…People inside of Fukushima, they tried to find out what really happened, but people outside of Fukushima -once you go outside of Fukushima, they're just not interested in seriously learning the truth, so they still go by biases about Fukushima. Once you go outside of Fukushima, a lot of people have those biased ideas, but I can't blame them because they don't get much information. As an evacuee from Minamisōma, Participant 7 is passionate on the topic of misinformation on Fukushima. Since the Earthquake, she has heard these claims from peers and adults alike. She shared: After the incident, people were so stupid sometimes, saying things like "People from Fukushima, they have runny noses all the time, " or "They don't live that long, " or "You shouldn't get married to women from Fukushima. " I heard all kinds of terrible, ignorant comments. As a teenager, I was so shocked to hear that, so that's why I'm interested in medicine now, [because] I was so shocked to hear those rumors from people. Before I went to college, I went to a prep school outside of Fukushima, and students and teachers at the prep school, those people would tell me their biased ideas, like that we're all sick or that we all have cancer, that we all have tumors. I was so shocked to hear this. Even the teachers said that to me…You know, when I heard that, I defended Fukushima, you know, saying "It's not like that. " I kept telling them that, but after a while, I just started ignoring those people and their words. No matter what I was going to say, they would never change their minds. Those biased ideas were already in there already, so I just stopped correcting people like that. However, many of the student participants felt that by sharing their experiences through these interviews with "foreign researchers, " they had the chance, as Participant 4 put it, "to tell everyone the right situation of Fukushima. " Participant 3 declared: Prejudice about Fukushima, I want to clear it…Any wrong information, such as the idea that Fukushima people cannot live because of the radiation -that's not true, that's not correct. Also, the idea that Fukushima produce is not edible, that's not true…When the incident occurred, there was a lot of news going around which discriminated against Fukushima, and I don't want that…There's still a lot of prejudice about Fukushima going around, so by speaking out, I hope I can clear some of it up. This is important to me…That's the reality, and I want people to know about it. Participant 6 also implored me to tell people in my "home country" about the reality of Fukushima's situation. He sees the Earthquake as an essential time in his life and in the lives of other individuals from Fukushima, and even called the incident "our disaster" several times throughout the interview. Despite his wish for Fukushima to move forward, he feels this cannot be accomplished until the misconceptions have been cleared. He explained, --- I don't want to keep in the past, in bad things. I want to go ahead, to forget about that disaster. But we have some impacts still from this disaster, so it is part of our, part of my life. It is part of myself, it is part of my identity. So to educate people is a good thing…I think that the information about this disaster, the information in foreign countries, it's not exact as in Japan. So some foreign students come to Fukushima and study about it and go back to their home country and spread the exact information, it's very important thing. He experienced this importance firsthand in 2018 as an exchange student in Belarus. Like its partnership with the Icahn School of Medicine at Mount Sinai, Fukushima Medical University has an "academic cooperation" with Belarusian State Medical University and Gomel Medical University, facilitated by Nagasaki University's Atomic Bomb Disease Institute [28,29]. Part of Belarus is still included in the Chernobyl Exclusion Zone following the Chernobyl disaster of 1986, and as Nagasaki University's international collaborative research coordinator wrote, "There has been active attitude in Japan toward learning lessons of Chernobyl accident aftermath and countermeasures since the Fukushima Nuclear Power Plant accident [29]. " Though Participant 6's curriculum focused on studying military medicine, shadowing thyroid surgeries, and practicing effective patient communication, what left the most lasting impression on him was how his concept of Chernobyl changed [30]. In our interview, he described, Before we visited there, I thought Chernobyl is very terrible, destroyed, but it is not exact information. --- Chernobyl is not terrible condition. And my understanding changed after I visit there, so I have good image about Belarus and about Chernobyl. [For foreign students] to come to Fukushima and go back to their home country helps them, they can have exact information. They can have a similar experience as I had in Belarus. He, like other participants, believes that international collaboration allows for increased opportunity to set the record straight when it comes to Fukushima. Participant 4 had a similar experience as Participant 6, though as a host rather than as a guest. In high school, she took part in a workshop in which Fukushima students partnered with French students as they toured through the prefecture in an effort to show what Participant 4 calls "the right thing" to the visiting students. She elaborated, --- People should just know the right thing. I know the right thing because I live in Fukushima. I want to tell the world the right thing because I know the right thing…I [thought] I could tell them [the French students] the right things of Fukushima. The right things instead of harmful rumors about farmers. And I could show them the nice sightseeing places of Fukushima. And the delicious food of Fukushima. Participant 4 explained that she felt she had, and still has, a responsibility as a Fukushima resident to represent her prefecture and to ensure that others have an accurate understanding of the place that she calls home. She expressed a wish to replace the damaging, inaccurate ideas I may have received with positive ones, and called this wish her "duty. " She said: That earthquake was big turning point of my life. And my high school teacher said that in the future, we are going to be asked about 3/11 and the accident at Daiichi, so we need to know the right situation of Fukushima, and we have duty, we need to tell people around Fukushima, around Japan. We have a duty. So I realized that I have to know the situation of Fukushima, the right situation of Fukushima. Do you know the bad rumors? The bad rumors about Fukushima? ... We, the people who live in Fukushima, have to tell [others] the right things about Fukushima, I think. However, not all those from Fukushima Prefecture are as clear as Participant 4 as to "the right things about Fukushima. " Participant 8 explained, "[We don't] know which was rumor and which is truth. Even the people, the local people, [we don't] know that. " He noted that the media's platform has been enhanced by a continued lack of trust in the government. Of his fellow Fukushima residents, he said, "They can't trust the government's state-ment…They couldn't think that they should believe all the information from the government. " When I asked Participants 8 and 9 where they believe the mistrust of the government comes from, they discussed in Japanese for a moment and laughed before offering, "The government said that the nuclear plant was safe. But it wasn't. " Distrust of the government in Fukushima following the Earthquake has been well documented. The Washington Post describes "a culture of coverups and denials that contributed to the nuclear accident and continues to dog Japan's efforts to restart its nuclear industry, " referring to the Tokyo Electric Power Company 's attempt to deny that the nuclear disaster had occurred for months after the accident [27]. Subsequently, TEPCO's president Naomi Hirose apologized for the lack of disclosure, admitting, "I would say it was a cover-up [31]. " In January 2019, a member of TEPCO's external advisory committee said at a news conference, "If TEPCO does not improve their communication, it will be very difficult for them to regain the public trust [27]. " Due to this local wariness of government information, Participant 8 expressed that unlike many of the other participants, his top priority was not to educate "not the whole world, or all Japanese people's minds, but the people around him. " He was adamant on this point, insisting several times that he wants to enact change "in a small way" rather than on a large scale. His friend translated, He has a strong feeling about caring for the people in his community. He has feelings for his community, but his community is not so big. He doesn't want to change public opinion or something. He just wants to change his people. Everyone has to care for their own. For some students, the desire for individuals outside of the prefecture to know "the right situation of Fukushima" stemmed from, as Participant 3 phrased it, a want "to be treated like everybody, just like anyone. Not anything special, we're just like anyone. I want to be treated the same. Equal. " Participant 7 remembers expressing this wish as well and described being treated differently than her classmates. As a high school student, she noticed that her teacher was projecting his idea of an evacuee on her rather than recognizing her true feelings. She recounts: My teacher treated me like an evacuee. He felt so sorry for the kids like me in his class. As a kid, moving to my grandparent's house, moving to the city, I was excited, but the teacher didn't understand how I was feeling. I think the teacher pitied me. He didn't treat me like a regular student, and when I was 14, I just wanted to be like everybody else. Teenage life by itself is confusing, but with the accident affecting that time of my life, it was all much more complicated. However, the motivating factor expressed most often by participants was their love of Fukushima. Participant 5's to-the-point explanation summarizes the thoughts of many of the students. She shared, In general, if you see the news, you know that people are kind of horrified about the lasting effects of the nuclear accident. But, you know, I have to tell them, those effects are not as bad as they think…A lot of people [are] living here, in Fukushima, and I don't want to present that image of Fukushima, because I love it here. Participant 4 was of the same mind; she believes that many residents of Fukushima such as herself want to rectify the negative image of their prefecture "because they love Fukushima, " and repeated, "I want everyone to know the situation of Fukushima. " This was the sixth time in our conversation that she'd said this -she then laughed and said of her goal, "Always the same. " --- Research as healing For the participating students at Fukushima Medical University, these interviews appeared to serve multiple purposes. The students were clearly motivated to speak with me by a passion to change the public opinion of their prefecture, and some hoped to practice their English, or else to learn more about North American medical education. But most students also said that they wished for their interviews to be healing, both to themselves and to potential future survivors of similar disasters. Participant 7 said that because of the Earthquake, "everyone knows Fukushima, " and that the prefecture is "famous now. " "I know it's because of the nuclear accident, " she said, "but I think we can somehow use it for good, " for both her fellow Fukushima residents, and for the future. As Participant 1 said, the interview helped to "evaporate some negative feelings, " and he declared, "I think it was a good thing for me to be interviewed. " Participant 1 felt that this was an ideal time to discuss his experiences. He said that while there was plenty of opportunity and encouragement to talk about his thoughts and feelings in the months following the Earthquake, that has since died down, and at present, he feels such activities to be inadmissible. Of the interview, he said: It's good for me to share the emotions. And, um, actually -actually three years, four years after the earthquake, the kind of atmosphere changed. Soon after the Earthquake, everybody thought, "Okay, we have to share the information and share the emotions because of the struggle with the earthquake. " But…years after the earthquake, majority people thought that we mustn't share the emotions, we mustn't speak about the earthquake. The main atmosphere shows that we're already finished with the earthquake. So I think it's kind of taboo [to] reopen old wounds. And for me, it's only my family to share my emotions about the earthquake, about evacuation…And I think that it was a precious experience and good experience for me, to share emotions, not only with my family and relatives but also with other people. Participant 3 also felt that he no longer had occasions to discuss his feelings. His most memorable experience of the day of the Earthquake was leaving school in a bus and looking out the window to see "snow at the same time as thunder. " Of that sight, he said, "That was something I will never forget. " He went on, "I was with my classmates when this happened, so once and awhile, we [used to] talk about it. The five years after the incident, we'd meet up and talk about it. But not anymore. " Participant 7 agreed that she is now in a position to benefit from an interview for different reasons than Participants 1 and 3. She explained that she decided to volunteer for an interview because "I think it's a good time for me to look back at my past, at what happened during 3/11. " Though in our conversation she recounted difficult memories, her tone was determined, even hopeful. She explained: I was so afraid of doing an interview or a survey before, when I was young. I was afraid to remember my experiences, the bad things that happened. Like Participants 1 and 3, she found that talking about her experiences was a crucial part of her recovery. Participant 8 also found discussion helpful, calling the interview "a rare situation" in that it offered him a chance to reflect while moving forward. He and Participant 9 discussed in Japanese for a moment before Participant 9 relayed, "These interviews made me rethink about the disaster. Reflect. Think back. Remember my situation and know what other people have to say, and combine these things to give me new ideas and so on. " Participant 8 agreed, and his friend translated the following: --- [Participant 8] thinks that he feels like he was healed by talking about his experience…He feels very ashamed that he almost forgot about his experience with the disaster, and this is a very good opportunity to remember about the experience, what it was like… These [interviews] give us the chance to think about [the Earthquake]. Not forgetting is very important, I think. It's not a good experience, but something good can come because of it. This "something" that he and his fellow interviewees hope will come of their experiences following the Earthquake is a healing in two directions, internally and externally. The internal healing appears to come from the interview itself, while the external healing will come from what they hope will be done with their words. As Participants 8 and 9's friend translated, "They don't want the memories of the tragedy to decay. They want people to remember the memories of tragedy…And they want to continue the memory, for the future. " Participants 8, 9, and 1 are all members of the University's student group Fukushima WILL, a club, Participant 1 defined via message, devoted to studying disasters and related issues as well as "radiation/radiation education in Fukushima. " Though capitalized, the name is not an acronym; Participant 1 says that the emphatic "WILL" refers to the "strong mind" of the organization, their determination to keep fresh "the situation [of the Earthquake], people's mind, [the] difficulties. " Their hope is that by helping to spread the stories of what the residents of Fukushima went through, survivors of future disasters won't feel as lost as they did. This goal is not unique to members of Fukushima WILL; many participants described a hope that their words would improve conditions for survivors of future disasters. As Participant 7 said, "I hope this kind of disaster never happens again, but in case it does, I want to let everyone know, as a good example, what happened here. " Participant 8 expressed a nearly identical sentiment, saying "There are a lot of nuclear power plants around the world, so accidents will happen again in other countries, so people will need to have the right things to treat those accidents. People from Fukushima had a very interesting [unique] experience, so these things should be taken to other people. " Participant 5 also highlighted the uniqueness of the experiences of those after the Earthquake. In particular, Participant 2 hopes that by sharing his experiences, he can offer more insight into the psychological effects of such a disaster. In addition to his want to "help Fukushima get better, help people get better, " he stressed the need for increased attention to mental health in the aftermath of similar events as a reason for his participation in the interview. He explained, I wanted people who never had an experience like this to know about the mental health issues that me and my friends had due to the earthquake, the tsunami, and the power plant disaster. It shouldn't be forgotten, and I don't want people to forget…I want them to know, other people in the world, I want them to know what we experienced…I want people to understand, when there's a nuclear accident, an atomic power accident, with a natural disaster like an earthquake, that the damage that causes to people's mental health, that lasts many, many years. It needs to be dealt with. The sharing of these post-disaster stories does more than provide an opportunity for healing, or memorialize history for the future, or "lead to improved medicine, " as Participant 9 put it. The experiences of these participants are deeply personal, a part of these students' lives. Therefore, to tell their stories is to create their legacies. In a quietly determined voice, Participant 8 -once a teacher, now one of the oldest students in his first-year class -told me and his friends why he decided to talk about what he went through: I may not do big things, achieve big things, in my life. But if I can tell about [my experiences] to other people, maybe they [those people] will help [other] people with hardships. These people who have had hardships, they might themselves have a high achievement, and that will give us the legacy of their achievement. So, knowing is very important, and conveying my feelings and my knowledge is very important…And people who are from other countries and other prefectures, if I give them some of the truth, maybe they will do things to help the refugees. They can do good things with that truth. --- Discussion and conclusions At present, Fukushima is a prefecture of dualities -at once, it is both injured and recovering; mourning and moving forward; unified with the government, and distrustful of it. The GEJE has a similarly complex significance for this study's interview participants. For most, if not all of these students, the event was career-defining -it opened doors not only to areas of study within the field of medicine, but for some, to medicine as the field of pursuit. At the same time, the lasting, burdening effects of the Earthquake endure: trauma, loss of life and destruction of property, stigma. And yet, even these consequences act as galvanizing agents. These students stand eager, motivated, and poised to make change. In this current study, through interviews and inductive thematic analysis, we explored the individual stories of medical students at FMU who lived in Fukushima during the time of the Great East Japan Earthquake. To our knowledge, this is one of the first studies to not only use a qualitative approach to detail common themes in the personal narratives of evacuees of the Earthquake, but also to provide a space wherein evacuees can control the narrative of the GEJE in academic literature. This study further colors our understanding of the disaster's impact on the behavioral, societal, cultural, and political aspects of the lives of residents of Fukushima. Our study has several strengths. The qualitative and ethnographic framework adopted for this study allows for a complex and nuanced investigation of the impact of a large-scale disaster. This approach lent itself well to broadening comparative, contextual, and cross-cultural perspectives on the GEJE. Furthermore, the focus on personal narratives and the common threads throughout these narratives adds a humanistic layer to current academic research on the topic. By placing these narratives at the forefront, participants were empowered to contribute to local and worldwide perceptions of Fukushima following the Earthquake. There are a few limitations to our study. Firstly, our study population exclusively included medical students, a group not necessarily representative of all those affected by the Earthquake. The themes discovered in this study may not be generalizable to the entirety of the evacuee population. Secondly, as with most ethnographic studies, the richness of the data depends on both the rapport of the researcher with the subjects and the openness and honesty of the subjects. Though the researcher felt that good rapport was established through close contact with participants and their friends prior to, during, and following the interview, this is difficult to fully ascertain. Interviews were conducted in private and anonymity was stressed at the beginning of the encounters to increase the likelihood of participants being open and honest. More chances must be given to residents of Fukushima Prefecture, both current and displaced, to provide insight into their experiences following the Great East Japan Earthquake and the associated tsunami and nuclear incident. Further ethnographic research should be conducted to expand upon the themes approached in this research, and to explore the wealth of other stories not touched here. Additionally, longitudinal studies of medical students from Fukushima Prefecture would permit more rigorous data on the impact of the Earthquake on career path. Finally, cross-cultural studies between survivors of nuclear accidents would allow for not only the healing described by the participants, but for the chance to pass on a unique expertise for potential future survivors. As the interview participants noted themselves, these stories are their legacy. --- --- Authors' contributions AS, ML, CK, and RY devised and designed the project. AS performed ethnographic interviews with help from ML. CK, SW and RY performed technical coordination. AS analyzed the data and wrote the manuscript with input from all authors. AS and ML revised the manuscript according to reviewer comments. The author read and approved the final manuscript. --- --- --- • At BMC, research is always in progress. --- Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: --- Competing interests On behalf of all authors, the corresponding author states that there are no competing interests. ---

Background The Great East Japan Earthquake and the resulting tsunami and nuclear disaster on March 11, 2011 have had a profound and lasting effect on residents of Japan's Fukushima Prefecture, particularly among evacuees. While there continues to be extensive news coverage and academic study of Fukushima Prefecture's recovery, there has been little exploration of individual narratives. This study aims to illuminate some individual stories of medical students at Fukushima Medical University (FMU) who lived in the Prefecture at the time of the Earthquake. Methods A qualitative approach was taken in order to investigate individuals' experiences with the goal of adding a personal dimension to quantitative studies on the subject. 10 open-ended ethnographic interviews were conducted with medical students at FMU in years 1-5 who lived in Fukushima Prefecture at the time of the Great East Japan Earthquake. All interviews were audio recorded and transcribed. Transcriptions were reviewed using inductive thematic analysis under the lens of ethnographic anthropology.Three major themes emerged from these interviews: first, that the events following the Earthquake influenced not only these students' decisions to pursue careers in medicine, but the ways in which they hope to practice medicine in the future. Second, that these students were motivated to share their experiences by a want to change Fukushima Prefecture's public image. And lastly, that the students viewed the opportunity to discuss their experiences through these interviews as healing, both for themselves and for the future. Conclusions While multiple factors undoubtably contributed these students' medical education, they cite the Earthquake as essential to their approach to their medical careers. Additionally, opportunities for the participants to discuss their experiences following the Earthquake appear to be rare but valued, as the students view their stories as their "legacies. " The enduring, burdening effects of the Earthquake appear to have galvanized the participating students to act on behalf of their communities and their Prefecture. Further qualitative studies in more generalizable populations are needed to improve and deepen our understanding of the societal, cultural, and personal impacts of the Great East Japan Earthquake.

INTRODUCTION This research deals with the range of vulnerabilities induced by climate change in the coastal areas of Bangladesh. Du e to the unique geophysical structure, the coastal areas of Bangladesh are extremely vulnerable due to climate change. The severe impacts of climate change are evident on the life and livelihoods of all walks of people specially in the coastal areas of Bangladesh . Poor and marginalized people are the worst victims in this case to be specific. However, due to this climate vulnerability, the vulnerable communities of the coastal areas have developed their own adaptation practices based on their indigenous coping mechanisms . To study the socio-economic vulnerability of climate vulnerable coastal area is the key motivation as well as the objective of this study. The main aim of this study is to carry out a vulnerability assessment of the selected coastal at-risk communities. The lack of knowledge, technological innovation, exclusion of vulnerable social groups, policy constrains and systematic analysis on climateinduced vulnerability, livelihood disruption due to the frequency and intensity of disasters, coping with the vulnerability of climate change with economic and social complexities, the coastal population remain barriers to sustainable adaptation strategies leading towards social and economic collapse . These distinctive characteristics make these coastal inhabitants more vulnerable in terms of social, economic or financial, physical, demographic, health, and environmental domains. Also, these sort of vulnerabilities lead them to become more susceptible to calamities, distress, disruption, and increases tangible & intangible loss . Nowadays, climate change is perhaps the most widely discussed of all the recent global environmental crisis-related topics, and several well-documented studies revealed that climate change creates an authentic link to natural disasters due to anthropogenic causes. That's why, little too intense alter in climatic The Dhaka University Journal of Earth and Environmental Sciences, Vol. 11 : 2022 Corresponding author: Nushrat Tashmin Email: [email protected] DOI: https://doi.org/10.3329/dujees.v11i2.68858 variability affects the social and economic wellbeing of populations . Climate change has influence over various interconnected issues of the nations of the world, such as: environmental, ecological, socio-economic, sociopolitical and so on . The ongoing threat of climate change imposes a severe impact on the lives and livelihood of coastal people, such as: physical infrastructure damage, lower level of health status, death of livestock, loss of income, physical injury, raising migration pattern, natural property damage and so on . It is now well known to all of us that among other developing countries, Bangladesh is recognized as the most climate-vulnerable country in the world . The frequency and intensity of disasters such as floods, cyclones, droughts, and landslides have changed due to the impacts of climate change. Moreover, the coastal area of Bangladesh is also facing the severe impacts of sea-level rise and salinity intrusion due to climate change . Over the last three consecutive decades, Bangladesh has experienced nearly $16 billion in GDP losses including property damage, livelihood disruption, communication destruction, infrastructure collapse, agricultural loss, and over thousands of death tolls due to about 200 climate-related disasters . Bangladesh has ranked 7th most vulnerable of the mostclimate-vulnerablee counties in the world, according to the IPCC 2021 report . Because of the flat topography and low adaptation capacity, the coastal communities of Bangladesh are highly exposed various natural disasters, such as: cyclone, storm surges, coastal flooding, sea level rise, and so on . The coastal areas of the Bangladesh are regarded as the frontline of climate change which are severely affected by increased flooding, increased intensity of cyclones, salinity intrusion and temperature extremes . Due to climate change the intensity and frequency of disasters increase and these climate induced disasters disrupt the social and economic wellbeing of populations specially in the coastal areas of Bangladesh . In this study, the researchers have chosen 4 unions of a coastal upazila of southern Bangladesh namely Koyra Upazila under the Khulna District of Bangladesh. The coastal areas of Bangladesh are affected by the impact of climate change and the Koyra Upazila is a remarkable disasterprone area where the impacts of these changes are notable and significant. This upazila is being frequently affected by cyclones and storm surges. This area was severely affected by the super cyclone Sidr and Aila . Moreover, research shows that in Koyra Upazial, salinity intrusion and waterlogging are the most common natural hazards . In this area the people are vulnerable due to the repetition of disasters before overcoming the impacts of previous disasters. Gradually vulnerable people living here are losing their capacity of disaster management and being trapped by the cycle of poverty . Therefore, this study focuses on the socio-economic vulnerability of the climate vulnerable Koyra Upazila. The objectives of this study are to explore the components and levels of vulnerability, explain the vulnerability scenario, and analyze the social perception of the climate-induced disasters of the selected coastal areas of southern Bangladesh. The present paper focuses on the socio-economic vulnerability of the selected coastal study areas of Bangladesh, and socio-economic impacts of climate induced disasters are also analyzed critically in this study. --- Background and Significance of the Study Climate-induced extrem force people to alter the course of life most of the time and ultimately engage tem in diversified occupation patterns for maintaining their lives & livelihoods. The result of extreme weather events such as floods, droughts, cyclones, and other weather-associated phenomena affect the wellbeing of affected people through the destruction of physical, human as well as social capital. Due to climate variability, displacement is also a common phenomenon especially in coastal areas of Bangladesh . Due to this sort of forced migration, the coastal residents become economically vulnerable . This is one of the most distinguished forms of vulnerability experienced by many households in this research's selected study areas. Climate change leads to negative consequences on every sector of our surroundings. Especially people who live in coastal regions suffer an immense crisis. The livelihood of the poor and marginalized people are disproportionately vulnerable. In coastal Bangladesh, the multiple stressors related to climate change make the life of the people more vulnerable than the rest of the country . Moreover, the study of vulnerability helps the authority to determine the actual condition of the communities and implement necessary plans for reducing the vulnerability. In this study, the socio-economic vulnerability of the community people at risk is assessed through the Socio-economic Vulnerability Index using a Composite Indicator Framework method. According to IPCC, the interrelationship between vulnerability and climate change depends on three components: exposure, sensitivity and adaptive capacity . These dimensions are further divided into five domains: physical, social, economic, demographic, and exposure to hazards . In this present study, the researchers used this SeVI index to assess the socioeconomic vulnerability of the study area with a specific focus on the social perception of the climate induced disasters. For the theoretical framework, this study used the PAR Model for developing the indicators for SeVI. The PAR model guided the researchers to determine the root causes and dynamic conditions of vulnerability. The most familiar conceptual model of vulnerability to disasters includes the Pressure and Release Model by Wisner. This Model emphasized vulnerability at micro-unit level. The PAR model shows the intersection of unsafe conditions and hazards which creates social vulnerability. According to this Model, disasters occur when vulnerable people are affected by hazards. The root causes of vulnerability are: social, economic and political conditions of community people. The vulnerable communities live under constant pressure due to their limited access to livelihood, education and healthcare, resulting in poor income and unsafe healthcare status and living conditions. To release this pressure, the vulnerability components of risk should be addressed properly by the concerned authorities . The study is about the assessment of socio-economic vulnerability of the people living in coastal Bangladesh . This study not only focuses on the socio-economic vulnerability but also it deals with social perception of the climate induced disasters of the selected coastal areas of Bangladesh. This index-based assessment with social perception can be an integral part of the contingency planning of the disaster-prone coastal areas of Bangladesh. Through qualitative and quantitative approaches, this study determines the changes of the holistic lifestyle of the community people at risk due to climate change. This study shows that the frequent climate change induced disasters have always been detrimental to the local vulnerable coastal communities. The socioeconomic vulnerability assessment is imperative for recognizing the impacts of climate induced disasters and formulating effective contingency plan. The vulnerability analysis by using the SeVI can be useful for assessing the coastal vulnerability of Bangladesh and identifying the priority areas for implementation of action-oriented community-based resilience planning for lessening the effects of climate change. --- METHODOLOGY Selection of the Study Area This Koyra Upazila is highly fragile and vulnerable in terms of climate change induced disasters. The available demographic sources reveal that about 13.3 million population may likely be homeless in the name of climate refugees as the victim of climate extremes in the southern coastal part of Bangladesh by 2050 . Koyra is a coastal upazila under the Khulna District of Bangladesh. It is about 100 km by road from Khulna City and located at the border of Sundarban Mangrove Forest. In 2009, Cyclone Aila created a huge damage and loss in Koyra and affected nearly a population of 152,496 over there. After this damage and loss, the government of Bangladesh and the relevant development partners regard this area as a priority for recovery and reconstruction . The Upazila is occupying an area of around 1800 km 2 . This area belongs to the immature delta slope which his hardly above the sea level. For this coastal location, the Koyra Upazila is susceptible to different disasters, such as: cyclone, flood, storm surges, salinity intrusion and water logging. The damage and loss scenario after the super cyclones reveals that the community people living in the study area are the worst sufferer of socio-economic vulnerable conditions . The researchers selected four unions among the seven unions of Koyra Upazila for this study. In this study, the primary data reveal the underlying and overlying causes of vulnerability in the study area. The data show the most vulnerable groups within these communities, such as: women, minor groups, low-income households, disable people, old-age people, displaced people etc. The researchers have chosen the four unions and tried to figure out the factors which are responsible for emerging crises. This study tried to reveal the socioeconomic vulnerability perspectives of the concerned unions under Koyra Upazila. --- Data Collection This research focused on a mixed-metho strategy by combining both quantitative and qualitative data. The researchers collected primary data conducting Key Informant Interviews , Focused Group Discussions , Questionnaire Survey and secondary data was collected through extensive literature review. By conducting systemic data analysis, the researchers tried to identify the socio-economic vulnerability of the people of selected communities by using the vulnerability index. Moreover, the researchers have also explored the adaptation strategies of the vulnerable frontline people by focusing on their livelihood patterns. The researchers have followed the non-probabilistic sampling method in this study. To have detailed insights into the life and livelihood of the climate vulnerable people, the researchers have selected 200 households on the basis of their climate-vulnerable situation. By conducting several FGDs with community people, the researchers have tried to collect data related to socioeconomic status of the households. To have detailed insights of the community lives and livelihood, the researchers have chosen some key informants from the selected study area with whom they maintained constant contact throughout out the fieldwork period. Moreover, some experts from the selected field of research interest were chosen for their valuable advice and opinions. --- Data Analysis After the data collection, the data was analyzed systematically in SPSS and Microsoft Excel for quantification. Furthermore, the qualitative data were analyzed thematically by chunking and coding. The researchers determined the indicators for vulnerability indicators by following the PAR Model. To determine the domain-wise vulnerability score, the researchers summed up the findings and organized each value according to the indicators of the vulnerability index: physical, social, economic, demographic and exposure to natural hazards . Here, the researchers took the score which varies between 0 to 1, where 0 means No Vulnerability and 1 means Extreme Vulnerability. In between 0 to 1, the researchers categorized the different levels of vulnerability including severe, moderate, partial, and fully vulnerable. The sub-components of vulnerability were standardized again between 0 and 1. Each subcomponent has an equal contribution to the SeVI index. A balanced weighted approach were used for calculating the SeVI : Weighted index score k = × The domain-wise vulnerability score was then calculated by averaging the weighted score of all sub-components for each domain category: Source: Sahana et al., 2021 --- RESULT The Nature of Climate-Induced Vulnerability in the Study Area For understanding and measuring the vulnerability of the selected coastal areas, the researchers have used the pressure and release model. In this research, the researchers have taken the livelihood pattern as an important indication along with other indicators, such as: Calorie intake, Life Style, Income, Expenditure, Savings, Debt, Housing Pattern, Literacy Rate, Access to Technology, Health Condition, Social Capital, Community Participation, Dependency on Relief and so on. According to the PAR Model, the social, economic and political situations make some people more vulnerable to natural disasters. Root causes, dynamic pressures and unsafe conditions makes the vulnerable population more sensitive to natural hazards. The Figure 3 shows the rooted character of vulnerability of the people who have limited opportunity as well as limited access to resources. --- Figure 4: The Social Causation of Disaster As per the PAR Model, the researchers have considered the vulnerability situation of the study area by combining two opposite forces: the mechanism of the progression of vulnerability and the occurrence of a hazard. The vulnerable situation of the community people is generated without the impacts of hazards initially then the occurrence of hazards creates stress in their life and complicates their situation . In the selected unions of the Koyra Upazila, the researchers have found that the climate induced disaster situations usually intensify the vulnerable situation of the community people. This situation gets more intense for the households who faces political, social and economic inequality. Researchers found that, the underlying economic and social problems of low-income households and socially marginalized groups affect livelihood opportunities and this situation gets worse during and after disasters. Moreover, the climate change situation exacerbates the damage and loss and the situation of the community people becomes more vulnerable. To have the insights about the vulnerability of the selected households of study areas, the researchers considered Community Based Disaster Risk Index and also took into account the number of dependent and independent variables based on Sensitivity , Exposure and Capacity . In this study, the index is comprised of relevant indicators which summarizes the overall climate vulnerability situation of the study area. Table 1 shows the domain wise vulnerability indicators and the associated general remarks on the findings. The level of vulnerability of each indicator is presented here with + and -signs . With these signs, the researchers were able to present the overall vulnerability of the selected households rapidly. The nature of vulnerability here is presented on the basis of the increase and decrease of the factors associated with the indicators. --- Financial or Economic 1. Number of employees per households 2. Sate of insurance debt or loan --- Availability of Lands Due to frequent climatic condition vulnerable coastal people loss their formal jobs and a vast majority of people are going under the poverty line . --- Social --- State of existing literacy rate 2. People attitude towards psychological trauma during disasters --- Number of marginalized people 4. State of vulnerable people access to Community decision making approach --- Social Capitals Majority of people loss social capital and being marginalized. Due to economic instability the holistic peace and stability in society are largely disrupted. Also, Social status created one type of caste system which is one of the most emerging issues for social transition of population. --- Demographic --- Population Density 2. Rate of migration towards urban area --- State of occupational pattern change Large portion of vulnerable people are migrating towards urban area for managing their livelihoods. --- Exposure to Natural Hazard 1. Not having shelter in cyclone centers 2. Willing to go to cyclone shelters --- Understanding National Waring System 4. Number of cyclones in last 10 years --- Number of floods in last 10 years People usually go to shelters homes during cyclones. Not all of them are willing to go to cyclone shelters but the ratio of willingness has increased after Aila. --- Vulnerability Scenario in Study Area The results of the study shoes that most of the households of the study area are vulnerable. The degree of exposure and sensitivity are high, and low adaptive capacity contributes to the high level of vulnerability. The physical, economic, social, demographic and exposure to natural hazards domains have contributed to different degrees of vulnerability . The selected households are found more exposed to coastal disasters and sensitive towards damages created by the disasters. Moreover, the low level of adaptive capacity of these households makes them more vulnerable. Most of the sampled households of the study area affected by the asset loss and home damages. The socio-economic vulnerability situation exposes their low access to healthcare services, less access to food and safe water, limited livelihood facilities, low economic condition and inadequate provision of infrastructure. The domains of SeVI are presented in Table 2. The researchers have also presented the domain wise vulnerability scenario of the study area. The researchers analyzed the areas of vulnerability in terms of exposure, sensitivity and capacity and categorized the unions on the basis of the category of of vulnerability. --- Social Perception about Climate induced Disasters The selected four unions are more disaster-prone areas among other remaining unions of Koyra Upazila. The residents of this area undergo a severe climate crisis. The extreme weather events due to climate change pose a great challenge towards their lives and livelihoods, During the field survey, the researchers asked local responders about existing weather patterns and rank the event that they faced in the last five years. Based on the emic perception of the local respondents, the disasters to which they are being exposed regularly are given below Therefore, almost a large section of the respondents said that cyclonic events make their vulnerable situation worse and these events are most responsible for civics distress and the researchers tried to figure out the recent devastating cyclone in concern unions in the following figure . It is well apparent from social vulnerability analysis thatthe Koyra Upazila is much more fragile in terms of social perspective, such as: in terms of poverty rate, availability, number of healthcare institutions, safe drinking water supply, rate of literacy, female participation on policy making perspective and so on. The community members do have their own socioeconomic and religious factors in this case which shape up their own perception about the climate-induced disasters. In their opinions, the frequency of disasters has increased in the recent past years and their direct and indirect experience within these situations have formatted their experiences regarding risk perception. Households which have socio-economic differences, do perceive the disaster risk differently. Some prefer to migrate from their community to near-by cities to find better options. To these households, migration is an adaptation strategy. However, the households which are economically vulnerable, are being trapped in their communities and cannot migrate due to their socio-economic dynamics. In addition to this, the male and female victims consider their situation differently due to their gender specific exposure and ethos. Female respondents consider the disastrous situations to carry more risks because they are more exposed their domestic roles and actives rather than their male counterparts. The respondents are less aware about the anthropogenic activities which leads the climate to change faster. Rather, they perceive these disasters according to local values, religious beliefs and cultural factors. Moreover, the community members do poses traditional strong family and social community values and during and after disasters they do play the roles of volunteer for communal wellbeing. As per the PAR Model, the socio-economic situation of a community determines their adaptive capacity towards disasters. The socio-economic development shapes a community's resilience capacity which sometimes depends on the non-structural mitigation measures. The social perception of a community provides us the overall mindset and social situation of local inhabitants. This study provides both the socio-economic vulnerability scenario and social perceptions of respondents about disasters. From this perspective, the policy makers and concerned authorities should take appropriate initiatives of awareness raising campaigns amid the community people. This sort of initiative will reshape their social perception about every stage of disaster cycle and it will have a positive impact on their socio-economic vulnerable situation. --- DISCUSSION Due to the unique geophysical characteristics of the coastal areas of Bangladesh, it is event from this research that, almost every year the entire study area is exposed to the increased frequency of extreme climatic events. Consequently, the households in this area are likely to become socioeconomically more vulnerable, butut the range of vulnerability varies between these unions. That's why the researchers have tried to calculate the index value of each domain from the perspectives of exposure, sensitivity and capacity. Based on SeVI index, each domain scores with a different value of vulnerability. The researchers tried explore the socio-economic vulnerability by studying the major components of the physical, social, economic, health and demographic . From the perspective of average socio-economic vulnerability , this study reveals that, among the selected unions, Bagali union is the most vulnerable from the perspective of almost all the domains of vulnerability index. During the fieldwork, it was quite clear that the reasons for which they are going through miserable conditions. The researchers tried to incorporate the social perception of community people about natural disasters to analyze the social process of the community. However, the other three unions are also vulnerable due to the climate change induced weather extremes. From the data analysis, it is evident that the selected four unions have high exposure to the climate vulnerability are at high risk of increased frequency of the disasters. This situation of exposure and sensitivity is rising the poverty rate and the living standard of the affected people is becoming low. During the fieldwork, the researchers have found some cases of child malnutrition, failure to repay debt, maternal and child motility, dropout from school, child labor and gender discrimination. The study area is highly exposed to climate induced disasters and extremely sensitive towards the damages caused by these disasters . Moreover, due to the socio-economic vulnerability the adaptation capacity of the marginalized households remains low. Among sample households, most of the households are affected by the economic vulnerability which limit their capacity to cope with the changing climate and disaster situations. So, these marginalized people are being trapped in the cycle of poverty. During the disasters, the damage and loss of their limited property exacerbate their situation and they become more sensitive to the extreme climatic events. Climate induced disasters also pose constant threat to the health of the vulnerable population living in the Upazila. Loss of agricultural productivity creates food insecurity and the salinity intrusion and drought in the dry season compel the safe drinking water to become scarce. The overall vulnerability scenario of the selected study area reveals the high socio-economic vulnerability among the communities of Koyra Upazila in terms of: low income, crop failure, vulnerable health status, scarce safe drinking water, inadequate infrastructure and so on. Moreover, the social perception of the respondents reveals the social systems and power relation of the communities. This social setup affects their adaptive capacity. The study shows that the social perception of community members is derived from the social processes which ultimately determine their access to opportunities and exposure to natural hazards. Also, the study shows that, during disasters, the local network of transportation becomes disrupted which leads towards a long-term communication disruption. Both the primary and secondary data sources also reveal that there is low level of literacy among the local people of selected unions. Moreover, the displacement during disasters also exacerbates the prevailing vulnerable situation of the sufferer households. Inadequate capacity and lack of amenities pose devastating impact on the lives and livelihoods of the local inhabitants and this situation is becoming more vulnerable due to the increased change in the climate. Those households who are at risk becoming marginalized and they require effective adaptation and mitigation measures to cope with this changing climate. --- CONCLUSIONS In this research, the researchers have tried to explore the existing vulnerability situation of the local communities of the southern coastal part of Bangladesh by assessing different domains of socio-economic perspectives which will be an important account for developing a contingency plan for the resilience of these localities. Moreover, the researchers also had some limitations while conducting the fieldwork, such as: time and fund constrains which restricted them to have the holistic vulnerable situation of more local people and emic insights of their way of life, their coping mechanisms and their real-life challenges during crisis events. However, the findings indicate that, the socio-economic vulnerability of the sampled households increases the poverty rate. The domains which the researchers selected in this research to study vulnerability have direct link with the disrupted livelihood and living standard of the local people. Due to the frequent climate extreme events, the damage and loss of the property, infrastructure and social capital make the inhabitants of the vulnerable coastal areas marginalized and they are constantly being deprived because of the climate injustice. Also, the findings of the study reveal that, in marginal households, the gender inequality persists and this often leads to gender-based violence so the women living these vulnerable unions are more vulnerable than their male counterparts. Though the vulnerable households sometimes perceive migration as an adaptation strategy but not all migration bring positive changes in the lives of the migrated households. Situation of the forced displaced households are far worst here. So, in conclusion it can be said that, the sampled households of the selected coastal areas are extremely exposed to climate change impacts and due to the lack of capacity they are sensitive towards the climate induced disasters. The selected unions of the Koyra Upazila are suffering from excessive socio-economic vulnerability because of the disrupted livelihood, low income, economic insecurity, physical vulnerability, low literacy rate, gender inequality, low health status, food insecurity, and problems regarding infrastructure, water and sanitation. These vulnerable coastal areas require efficient disaster risk reduction strategies, contingency plans, policy measures and implementation for resilience. Formulation of holistic vulnerability assessment framework and effective coastal risk management will be helpful in lessening the impacts of climate change. Engaging local community people in community-based risk assessment, initiating structural and non-structural mitigation measures, and promoting indigenous adaptive strategies are highly recommended by the researchers. To lessen the impact of vulnerability, the researchers suggest more costeffective solutions, awareness raising initiatives, capacity building programs, increased social security services, effective disaster response and increased coordination among stakeholders and community people. Moreover, strengthening the social system with increased economic opportunities is also an essential step. In this regard, the use of a comprehensive vulnerability index will escalate the feasibility of climate change induced vulnerability assessment. These provisions will be very much helpful for the frontline vulnerable people who requires immediate attention. Social safety net assistance can play an important role is this regard and the proper disaster risk management formwork will be necessary for reducing the vulnerability of the frontline people of the coastal areas.

The study intends to explore the socio-economic vulnerability of climate change in the coastal Koyra Upazila of Bangladesh. The present study is concerned with climate-induced disasters and associated vulnerabilities which adversely impact the social and economic aspects of coastal inhabitants of Bangladesh. The researchers selected four unions (Koyra, Amadi, Bagali, and Dakshin Bedkashi) among the seven unions of Koyra Upazila for this study. In this study, the socio-economic vulnerability of the community people at risk is assessed through the Socio-economic Vulnerability Index (SeVI). This research focused on a mixed-method strategy combining quantitative and qualitative data. By conducting systemic data analysis, the researchers tried to identify the socio-economic vulnerability of the people of selected communities by using the vulnerability index. This study reveals that the study area is highly exposed to climate-induced disasters and extremely sensitive towards the damages caused by these disasters. Moreover, due to the socio-economic vulnerability, the adaptation capacity of the marginalized households remains low. Among sample households, most are affected by the economic vulnerability which limits their capacity to cope with the changing climate and disaster situations. This study not only focuses on the socio-economic vulnerability but also it deals with social perception of the climate-induced disasters of the selected coastal areas of Bangladesh. This index-based assessment with social perception can be an integral part of the contingency planning of the disaster-prone coastal regions of Bangladesh.

INTRODUCTION The role of journalists on social media is no longer restricted to being a news promoter or a columnist remarking on the factualness of Twitter's Trending Topics. Journalists are constantly taking on a didactic function and endorsing discourse on geopolitical debates, social issues and even suggesting places, reading material, and restaurants for their followers on social media. Such characteristics reshape the journalists' participation within digital environments in such a way that the interaction seems more active and closer to the public -thus approaching a digital influencer persona. Brazil is the leading Latin American country in the growth of influencers and strategic uses of digital networks according to consumer behavior surveys. It is estimated that 71% of Brazilian social media users follow some kind of digital influencer . Based on this context, this article aims to discuss the construction of journalistic skills and the appropriation of influencers' practices in journalists' online communication. The empirical analysis was based on three Brazilian journalists, all of them related to the television industry and with a bold digital presence: Astrid Fontenelle , Guga Chacra , Evaristo Costa . The new form of journalists' presence on social media is leading the field towards a perspective that Bruns has already identified: "We follow individual journalists […] rather than trusting only in the imprints of news organisations, " . In Brazil, supposedly, these new communication actors are defining what is a successful online practice and even influencing journalists to build a digital presence. Therefore, this article argues that the authoritative speech of journalism is facing a change that could be understood based on two questions: 1) What are the similarities and differences between journalists' online practices and that of digital influencers'? 2) How have social media and the intimate perspective shortened the distances between journalists and their audiences? --- JOURNALISTS: ATTRIBUTES AND AUTHORITY The first academic and technical record of journalism is from the 1690s. At that time, Peucer identified relevant elements that underpin the journalistic practice supported by the pillars of truth, justice and ethics. The influence of journalism is under threat even when it is built through professional, ethical and deontological values. In order to understand this context, it is important to review some fundamental key values for legitimacy and journalistic authority. ISSN: 2763-8677 First, before the journalistic practice itself, there are deontological presuppositions that will together grant legitimacy and credibility to journalism. Ethics and morals are factors that guide the profession, relationships with colleagues, sources and the organization of the field. Abramo defends the public commitment of journalists and emphasizes that "There is not a specific ethics of the journalist: his ethics are the same as the citizen. " Other value confer legitimacy to journalistic practice its professional ethos: truthfulness, precision, inquiry into social problems, separation of information and opinion, verification, interpretations of reality . All of this is part of a general standard that Abramo classifies as journalistic ethical value. It may seem broad, but that plays a determining role in the structures of this professional practice. This paper aims not to discuss the ethical issue in depth. However, the understanding of a duty towards the citizens' interest is essential to journalism and, therefore, to this article. The journalistic practice relies on the ethics and morals inherent to the journalist and it is part of legitimacy and credibility -as a result of daily professional activity. Legitimacy derives from the fulfillment of some attributes that make journalism socially recognized as a profession. In other words, the recognition of journalistic practice by the public guarantees institutionalization. Therefore, recognition is professional assumption and a consequence of the professional practice based on ethics and morals, but it is not necessarily a characteristic for the existence of journalistic discourse . The same thing happens with the idea of journalistic credibility, which is a value that arises with a journalist's professional lifespan and its performance. Credibility arises from the bond of trust established between journalists and the public through the narrative and logic of the journalist's performance. When a performance is associated with truth, it establishes a relationship of belief and reliability. The existence of these essential principles to validate journalistic conduct only exist because the field has adhered to elementary characteristics to differentiate journalistic discourse from any other type of narrative. First of all, it is necessary to talk about the public interest that guides journalistic practice. In short, public interest comes from anything considered impactful on the life of a community and audience. Public interest anticipates the act of doing journalism and can be classified as collective motivation or social interest . In this context, Groth has put forth the possibility of understanding journalism as a science and its essence as a result of society interest . According to the author's conceptual and systematic model, journalistic activity can only be considered as such by analyzing the characteristics that form it. Thus, it is necessary to "disassemble and reassemble" the elements in order to understand their value and utility. Groth systematized some of them as elementary ones: periodicity, universality, timeliness and publicity. To build a logical analysis of the journalistic discourse it is necessary to add the characteristics of Groth's model to others such as plurality, uniqueness, objectivity, proximity, public interest and target audience. --- ISSN: 2763-8677 According to Gadini , plurality is a contemporary feature in journalism and it is greatly influenced by North American journalistic practice in the 1960s and 1970s. Plurality is the search for the opposition of a discourse, when journalists research different and even opposite perspectives and interview diverse profiles to ensure various voices in a report on the pretext of increasing the reliability and credibility of newspapers. Through plurality, the practice of "listening to both sides" became a key element in distinguishing journalistic discourse. Tuchman understands the plurality of sources as a fundamental attribute of the news and relates it directly to another founding element of journalistic discourse: objectivity. However, the debate about this element, as well as the notion of impartiality, is considered an ideal to be achieved, but it is impracticable in its totality. According to Tuchman , objectivity is a strategic rite in content production because it systematizes journalistic practice. Another category defined by the Brazilian professor Genro Filho, in "O Segredo da Pirâmide" , is singularity. The concept is based on premises such as the path of the journalistic narrative, the constant search for the new or unusual and the inherent potential of the practice to be a propellant for social debates. Although the author is criticized for the Aristotelian and Marxist thought that generated this category, it is from this theoretical viewpoint that the research in Brazilian journalism structures the discussion of the lead. Another reason to highlight this element is that journalism is a form of knowledge solidified in the singular, in the unique, in something neither universal nor particular . Another characteristic is proximity, related to the question "where?". It means the geographic space covered by journalists at the moment of news coverage. Through this characteristic one can evaluate the thematic scheduling given to subjects by journalism. Considering the degree of proximity and interest ISSN: 2763-8677 that the subject has from a certain community, the subject will have a greater repercussion . In journalism, speech also guarantees the confidence of the reader/viewer and the communication vehicle, therefore it is aggregated to the journalist's behavior and language expression. In addition, the journalistic technique through elements that guide the professional practice also guarantees authority to the professional. Therefore, journalistic authority brings together a community of social actors who primarily share an understanding of what journalism is and its main function, placing journalists as professionals engaged in proving themselves as observers who truthfully report reality. It would be possible to say that journalists claim their cultural authority . This kind of authority arises through the modus operandi of journalists, where the professionals themselves develop a professional identity/personality based on ethical and technical values. The result of the journalistic authority, beginning with the professional culture built over the years, is the recognition of the journalist as an authorized observer : someone in whom the population of a city can trust and that with time adapts to technologies and communication transformations. There are also other practices that legitimize journalists' work; professional orientation, development of specific news forms, and the personal narratives they circulate to support a privileged social place . Carlson argues that journalistic authority is always the product of complex and variable relationships, as the ones with the audiences, sources, technologies, and critics, they are all responsible for shaping journalistic authority in the contemporary media environment. --- DIGITAL INFLUENCERS: LEGITIMACY AND INTIMACY The term digital influencer became part of the academic, mediatic and commercial vocabulary in 2015 in Brazil and it refers to internet users producing content on social networks in various formats. Digital influencers are considered ordinary people, originally amateurs, that based on the participatory culture can produce content online and gain audiences. This possibility is a result of a landscape of participatory lurch where the tools of distribution and production are now available online and the production sphere is occupied by everyday people and no longer confined just to the traditional media. Digital influencers can also be understood as a type of internet celebrity , as members of a professionalized, commercialized and monetized circuit , and as opinion leaders . Becoming a digital influencer requires not just one's desire but also legitimation conceded by the public or followers on social media. Legitimation "is based on the necessity to create or to reinforce the ISSN: 2763-8677 position of legitimacy of the speaking subject" . Through a journalistic perspective, this process can be based on the speaker position of authority that can be conferred by an "institutional authority". In other words, it means that a journalist is commonly recognized as credible because of the mediatic institution where they work giving them the "right to the speech" . However, the internet has broken down some traditional media processes. Digital influencers have not ever had an "institutional authority" that could give them the possibility to speak and to gain audiences interested in what they share online. The legitimation process, then, depends fully on how followers recognize them as credible, rightful and worthy of their attention. Previously, fashion bloggers used to gain their audiences as a result of a democratic turn of fashion. Blogs were considered a space with a claim on the real: "one space where real fashion, fashion as worn by real people, can be seen" . This gave fashion blog readers the impression of a non-mediated relation as opposed to the traditional and mediated fashion media. As digital influencers are a developed and adapted form of their predecessors, the bloggers, they are still attached to this characteristic of realness and non-mediated relations. And as ordinary people, digital influencers cannot expect to have any kind of previous authority and if they do so it seems that any "offline" capitals are not automatically transferred to digital relations. Thus, legitimation needs to occur as followers acknowledge influence, credit, knowledge or some type of capital that is important in a social group, or a niche, that the digital influencer is trying to embrace. As part of a very specific legitimation process, digital influencers can be legitimated by traditional media and brands as a result of their followers' recognition of their legitimacy and reputation. And it can be identified in their potency to interfere in the consumption practices of their followers and the topics that they are able to circulate among an online community. It is important to emphasize that digital influencers are a broad definition of individuals that can also call themselves YouTubers, Instagrammers, TikTokers, bloggers or content creators. And despite the stereotypical tendency to group digital influencers into a uniform, homogenous category, they can be part of very different genres such as fashion, travel, food, games, books, humor and so on. This is a key aspect of digital influencers and is also related to multiple possibilities of reputation and legitimacy granting. In other terms, the right to speech is variable in the digital influencer economy. Another standpoint to understand digital influencers is in accordance with what Abidin has described as perceived interconnectedness . The author debates intimacy as a business strategy for influencers and also highlights the importance of an impression of authenticity. The same is identified by Reade , who describes how "social media influencers employ practices of authenticity to foster relatability and intimacy with their followers. " These strategies are based on the participatory ISSN: 2763-8677 culture that shares a sense of connectedness among users in different levels of visibility and production online. A prominent content genre among digital influencers is lifestyle. Abidin has identified that these influencers resort to disclosing intimacies as they showcase trivial parts of their lives, not only luxurious advertorial photos. It gives the illusion of commonality, horizontality and a less hierarchical relation . More specifically, Abidin draws on ethnographic evaluations and states that "Horton & Wohl's notion of parasocial relations […] builds a model of communication through which influencers convey intimacies" . The framework of perceived interconnectedness is presented as a parallel to parasocial relations -Horton & Wohl theorize on how television and radio personalities "produce onesided interpersonal connections and an illusion of intimacy with their audience through conversational small talk that appears informal, casual, and responsive" . Although digital influencers enact similar relations online, there are some differences that update the theory and make digital influencers' relationship with their followers more complex. The table below takes seven elements into consideration and characterizes each of them according to the traditional perspective and the digital one. Basically, the framework of perceived interconnectedness also clarifies the legitimation process of digital influencers as it should be grounded in a flat organization of actors, interactive dissemination of content, a discursive strategy based on perceived intimacy, and co-creation with their audiences. Reade describes three practices among digital influencers that could be related to what Abidin has identified: 1) Posting raw images , 2) Storying the everyday, 3) Moments of ISSN: 2763-8677 "real talk" about topics considered important to the audience. This process of impression management is a way to ensure authenticity and relatability between digital influencers and their public. Reade's findings also corroborate the notion of calibrated amateurism: "a practice and aesthetic in which actors in an attention economy labor specifically over crafting contrived authenticity that portrays the raw aesthetic of an amateur, whether or not they really are amateurs" . To do so, digital influencers can include behind-the-scenes content in their social media profiles, as well as "natural" photos and snaps of their daily lives aiming to promote a sense of realness, an impression of spontaneity and "unfilteredness despite the contrary reality with their followers" . This intimate disclosure strategy, however, is not present in journalism practices. Nevertheless, journalism is facing what authors consider a "reputational decline" mainly brought about by new technologies and postmodern disbelief in traditional institutions. Along with capital issues, journalistic reputational capital is also under threat: "Part of the historical authority of news institutions cannot be reduced to such easily quantifiable metrics as audience size, revenue, or even Pulitzer Prizes" . Now it is not possible to construct a good reputation online by that or by professional competences like "maintaining integrity, adding value to information for an audience, demonstrating knowledge, linking to sources and explaining methodologies" . Journalists are now required to do this "in a public, real-time realm" -a realm perfectly illustrated by digital influencers. --- THEORETICAL AND METHODOLOGICAL FRAMEWORK Based on the theoretical discussion, the authors identified some possible categories that could help analyzing and characterizing journalists and influencers practices online, and it resulted in the proposition of a framework of analysis . The theoretical framework describes the characteristics of journalists and digital influencers and it is strongly inspired by the proposition of Abidin and the model of Groth with the insertion of concepts of the theoreticians discussed above . The objective of this framework is to serve as a theoretical-methodological input throughout the analysis of the profiles of the journalists on social media platforms. In the elements' column, there are points shared by journalists and digital influencers as media actors. However, each of these elements reserves specifications of their professional practices, listed in columns 2 and 3. More than a checklist tool, the framework may be used to describe the journalists' profiles and then used as a qualitative perspective in order to more easily recognize major aspects of journalists and digital influencers. --- ISSN: 2763-8677 --- ASTRID, EVARISTO AND GUGA: JOURNALISTS ON INSTAGRAM As mentioned before, this article discusses the digital presence of three Brazilian journalists on Instagram. Their selection as an empirical object shows the relevance of their journalistic work and demonstrates how they are important personalities in the Brazilian media landscape. Also, this is not the first study conducted in Brazil about the trio of journalists Astrid Fontenelle , Evaristo Costa , Guga Chacra and their digital presence -what reinforces validity within the field as a recognized empirical object among peers. To recognize the similarities and differences between journalists' online practices and digital influencers and how this convergence is capable of affecting the idea of journalism value, this paper focuses on Instagram profiles since it is an important platform for digital influencers and it is slowly being adopted by television journalists . The first journalist analyzed is Astrid Fontenelle, a Brazilian journalist who is currently a TV host on GNT, a pay channel from Globo. She has hosted a female roundtable called Saia Justa since 2013. Astrid is known for soft journalism and variety programs. During the period of this research, she had 1 million followers on Instagram, and more than 5,576 publications2 . Evaristo Costa is a Brazilian journalist who is currently an anchor at CNN on a soft news program3 . From 1999 until 2017, Evaristo Costa worked at Globo, one of the major TV broadcasters in Brazil and Latin America. He ended his Globo career at the Jornal Hoje, a daily newscast. In February 2020, Evaristo Costa was considered the most influential personality in Brazil according to research from Ipsos 4 . During the period of this research, Evaristo had 7.5 million followers on Instagram, and more than 800 publications. Gustavo Cerello Chacra known as Guga Chacra. He is a television commentator specialized in the Middle East and holds a master's degree in international relations from Columbia University. Guga Chacra lived in Beirut, Lebanon, during 2008 and 2009 as an international correspondent for the newspaper O Estado de S. Paulo. Since 2009, he has been living in New York and is currently an international commentator at TV Globo on its pay channel dedicated to the news broadcast Globonews and on its ISSN: 2763-8677 radio station Rádio CBN. He is also an international politics columnist at the newspaper O Globo, from the same media conglomerate. During the period of this research, Guga Chacra had the lowest number of followers of our sample -294,000 -and more than 2,000 posts. Specifically, this article studied Astrid, Evaristo and Guga's feed posts and stories through a non-participatory observation between January and October 2020 by live archiving screenshots of publications. The research did only focus on journalists through the empirical observation since the authors have already a large history of investigating digital influencers practices. So, the influence logics were observed through theoretical and previous research and journalistic perspective was gained through empirical findings . In order to constitute a framework of analysis to get into the field, based on the discussions undertaken so far, the authors were strongly inspired by the propositions of Abidin and the model of Groth with the insertion of concepts of the theoreticians discussed above to recognize elements that would reveal attributes of digital influencers or journalists in the field. --- AUTHORITY AND FLOWS OF COMMUNICATION It is already a consensus that "online social networks constitute such sites of self-presentation and identity negotiation" and as a result construct a networked self. The three journalists analyzed dedicate the space of their Instagram bio for institutional labels: TV host, anchor and commentator, respectively . When Astrid, Evaristo and Guga show their journalistic ties in their bios, there is evidence of impression management. The name of the media outlet also works as a sign of their position of authority -an institutional authority . The journalists have some credits and capital that they have accumulated from their "offline career, " as opposed to an influencer that starts from digital. The way in which they enunciate their ethos reveals that the trio is in a process of digitalization or adaption of their content and works to a digital sphere. During the observation period of this research, there was a lack of evidence of a bi-directional flow of communication between the journalists and their followers. Astrid Fontenelle does not respond to all of the comments left on her Instagram posts, but she constantly likes all of them even without answering back. But there are still some occasions when Astrid answers a few messages. Guga Chacra also likes almost every message from his followers even when not answering them. Evaristo Costa, however, seems to be the least involved in bi-directional flows of communication since he rarely responds to, comments on, or likes the messages on his posts. This element can be read through different perspectives: career digital influencers are in a moment of professionalization where a team is required to handle social media, interact and respond to messages. Considering that the journalists are not fully dedicated to social media as a career, the amount of comments on their posts does not have the expected interactive flow of dialogue. It is also possible to infer that their digital presence is not attached to a flat interaction but rather to a top-down distribution of content, a replication of traditional media flows. A journalistic approach is also acknowledged throughout the different content genres on Instagram feed. There is a balance of posts about private life; professional life and social issues. Private life posts are marked by the appearance of her son and husband; pictures with friends; moments of leisure; and even moments of disclosure concerning her faith 5 . Professional life posts are the ones where the journalist invites her followers to watch Saia Justa, tells what happened on the show and shares institutional videos 6 . Transmedia expertise is an element present in digital influencer practices but crucial to journalists . This convergence strategy, spreading what was once on TV to social media, also reinforces her professional links and a "right to speech" that is tacitly and constantly reminded to followers. If Astrid were a digital influencer, it would be possible to describe her as a niche influencer, or a minority influencer -a digital influencer who is part of a social minority and produces content based on personal issues, according to Abidin . But she is a journalist, so the commitment to some specific discussions can be part of her practice as an opinion leader. On Instagram, racism is a constant agenda; from personal posts about her son during #blacklivesmatter; pictures of a Brazilian Black boy, João Pedro, killed at home by a stray bullet in Rio de Janeiro; posts about Thelma Assis, the young Black woman who won Big Brother Brasil 2020 7 and a wide variety of discussions. But other agendas are also considered like antifascist posts; or an outfit picture with a T-shirt that says habibi , a tribute to Lebanese people after the explosion in Beirut 8 . 5 Posts published on 07/03/2020, 10/18/2020, 08/06/2020, 05/25/2020, respectively. 6 The posts were published on 09/30/2020, 09/24/2020, 07/01/2020, respectively. 7 Posts published on 07/07/2020, 05/19/2020, 04/28/2020, respectively. 8 Posts published on 07/01/2020, 08/06/2020, respectively. The publications about racism, social minorities and even politics put Astrid in an interesting context. The journalist combines her own opinion with credible sources like data from social institutes and historical facts. And even when there is a personal approach, there is a clear element of discursive authority granted by journalistic expertise and referencing sources. This is in contrast to digital influencer practices that constantly present an amalgamation of opinion, information and even commercial content . --- Content strategies: humor, proximity and relatability Evaristo Costa's Instagram posts are not defined by a journalistic periodicity nor a digital influencer logic of publication determined by the speed imposed by algorithms . The journalist does not post daily and his Instagram has been left without publications for long periods . There is also no editorial parameter or a clear thematic distinction among the posts, as opposed to Astrid's. The lack of consistency in Evaristo's posts reveals how Instagram is secondary to his practice. This is an important point, as journalists tend to use social media as an extension of their profession and not ISSN: 2763-8677 as a medium where journalism itself is made. Contrarily, digital influencers use social media as a medium and as a labor locus . At the same time, Evaristo has more followers than Astrid Fontenelle and Guga Chacra. One of the main reasons can be credited to his career at TV Globo and as an anchor of one of the major daily newscasts in Brazil, Jornal Hoje. Another assumption has to do with one of his unique characteristics online, which is the humor employed in his posts that seems to catch his followers' attention. The first image of Figure 3 is a post for Journalists' Day where Evaristo appears wearing unexpected bottoms for a hard news journalist. In a different publication , a smiling Evaristo is posing in front of an ancient building in London, wearing a suit and the caption reads "My belly is shriveling up to button the suit. Laughing in despair". The humoristic characteristic moves Evaristo Costa away from a stereotype of a journalist and approaches a digital influencer perspective: he is not connected to his audience because of his role as a public representative or as a result of a sense of informational integration, but instead is based on a proximity element and a communal sense . Concurrently, humor is capable of a discursive generalization and a possibility of amplifying Evaristo's reach online, as it does not depend on a niche audience or specific interests -an element of universality. Evaristo does not employ many transmedia strategies. There are some videos preparing the audience for his TV show debut on CNN but this is not constant on his profile. Furthermore, professional posts are also presented in a humorous way. Despite the announcement of the TV show with an ISSN: 2763-8677 institutional and more formal video, Evaristo publishes playful professional posts. Figure 8 is a behindthe-scenes image, a common type of publication among digital influencers who work with an aesthetic of amateurism . In the photo Evaristo appears on a rainy day in London recording a part of his TV show and the caption reads: "This is what CNN doesn't show!" . The caption is a joke about the journalist's former outlet media, TV Globo. "This is not shown by Globo!" was commonly said by Brazilians as a complaint for more democratic media coverage but it is now used even as a meme to talk about controversial issues. Another example is the third image of Figure 2 where Evaristo is back in his suit in a professional pose with a CNN background . The captions, however, are a part of an axé song by the Brazilian singer Ivete Sangalo, known as the Queen of Carnival in Brazil. The line says: "There's going to be a party!". Humor, and especially its Brazilian component, plays an important role on Evaristo's Instagram, working as a connector between him and his followers: a supposed distant TV journalist shares ordinary jokes online, an element of proximity and relatability. These elements could characterize Evaristo Costa as a digital influencer, since authenticity is a crucial element to their practice. --- Content strategies: professional life and public interest Lastly, Guga Chacra's Instagram profile tends to a more impersonal perspective. Differently from Astrid and Evaristo, a significant part of Guga's publications is not personal pictures of himself. Guga is a frequent publisher; there are new posts every other day and sometimes twice a day, although his Instagram does not have a daily periodicity. His content strategy is possible to be grouped into four clusters: 1. Landscapes; 2. Professional posts; 3. Personal and family posts; 4. Facts and public interest posts. Landscape posts are frequent on Guga Chacra's Instagram feed. As a Brazilian living in New York, he is constantly photographing the city. These posts do not even have a caption: the whole attention is on the view. Guga Chacra is not known for his photography so this practice could be considered a type of amateurism and an entrance into participatory culture . At the same time, it works as another possibility of self-presentation that is not restricted to an Instagram bio. The international lifestyle can be recognized also as a professional identity negotiation among Guga and his followers. Professional posts replicate a unidirectional distribution flow because of the brevity of their captions that are not always stimulating conversations with followers, like Astrid Fontenelle's, or causing laughs, like Evaristo Costa's. In this category, Instagram posts are used as transmedia strategy: the propagation of TV moments online and the presentation of articles or interviews published in print media9 . It would be possible to characterize Instagram as a repository of Guga Chacra's achievements at work and his institutional authority, even when using Instagram for some behind-the-scenes moments. Among professional posts there are screenshots of his own tweets. Guga Chacra has 1.1 million followers on Twitter, a telling number that is higher than the number of followers on Instagram. The chosen tweets addressed the COVID-19 situation and Brazilian political issues emphasizing his role as a commentator, where he is able and free to make more opinionated expressions. The use of social media assumes some type of disclosure content and intimate relations. This is not exclusive to digital influencers and it would not be different on Guga Chacra's profile. However, his personal posts seem to have a double intention: proximity with followers and legitimacy of his trajectory. Disclosive posts like the ones of his parents grants not only relatability but credit and authority, as the posts are also linked to some of Guga Chacra's professional abilities or skills from his youth. His career is often intertwined with his personal life: when he publishes a picture from 1997 of a road trip in Lebanon and other countries from the Middle East with his family or when he posts a "throwback" of him in Jordan in 2004 10 . But simultaneously there is a reinforcement of his work as a commentator specialized in the Middle East or his role as an international correspondent who is expected to have visited the world. 12 Posts published on 09/30/2020, 09/05/2020, 06/03/2020, respectively. ISSN: 2763-8677 --- TRANSMEDIA EXPERTISE The last important issue about the journalists' use of social media and appropriation of digital influencer practices is related to some specific tools. The first one is Instagram Stories. This tool is known for its ephemerality, a post on Stories remains there only for 24 hours so, contrary to the feed, it is not an archive but a space for conversation. It recalls an "oral paradigm of communication, where the visual is now a disappearing utterance" . From the trio, Astrid Fontenelle is the only one who uses Stories as a consistent content strategy. Though Astrid publishes every day on her feed -sometimes even twice a day -on Stories she posts up to 70 posts daily consisting of self-videos or photos. Benefiting from the ephemeral nature of Stories, Astrid begins her day reading the newspaper with her followers. The nature of these publications is sharing the news, but she also stands out as a type of media critic, questioning the headlines and approaches to some topics . This group of Stories postulates Astrid's commitment to accuracy, information and impartiality -journalistic expertise and ethics. At the same time, Astrid's use of Stories reveals a grasp of Instagram tools, putting her in a category of users who can vary content and formats online. This also brings her closer to digital influencer cultures where being a heavy user of social media and an expert in new tools is a prerogative of the job . ISSN: 2763-8677 Stories are also a place for authenticity allowed by the disappearance of content in 24 hours. So, it is not uncommon to see Astrid sharing her meals of the day, home issues, TV shows that she is watching , a new filter being used 13 . Evaristo Costa and Guga Chacra post more infrequently on Instagram Stories. Evaristo uses the tool for shooting landscapes from London using, once again, humor. The third image of Figure 6 is a Story where the journalist chooses the song "Arco-íris" by Xuxa, a former singer and TV host known as "Queen of the little ones" because of her career on TV shows for kids. The song gives a funny contrast as it has childish lyrics and musical arrangements. Conversely, Guga Chacra uses Stories mainly for reposts related to his participation in Globonews. The second tool is Instagram Reels, a feature launched in August 2020 that allows users to post short and funny videos, very similar to TikTok. Only Astrid Fontenelle and Evaristo Costa have experimented with Reels but neither has used it for a long time. However, appropriation of tools is not just using what the platform offers but also recognizing some digital conventions. Evaristo Costa does that when he posts a variety of formats on his feed. There are TikTok videos where he dubs himself presenting his former TV show Jornal Hoje and funny voiceover audio clips about relationships from TikTok 14 . Evaristo Costa also publishes memes, photos edited by himself with a comic vein like with a picture of him beside Kim Kardashian when he was nominated the most influential personality in Brazil; or a photo with the British Royal Family at the time Prince Harry announced leaving his royal duties 15 . 13 Stories published on 10/28/2020, 07/09/2020, 03/02/2020, respectively. 14 Posts published on 04/29/2020, 05/19/2020, respectively. 15 Posts published on 02/17/2020, 01/08/2020, respectively Lastly, with digital influencers there is a business model that assumes partnering with brands . An attachment to brands is not allowed for journalists that work at some specific media outlets in Brazil. Guga Chacra is likely under these contractual rules because he has no commercial publications on Instagram nor any association with brands or sponsored posts. Astrid and Evaristo seem to be in more flexible contracts. Evaristo Costa is an ambassador for an investment agency and the brand appears on his feed alongside other brands. Astrid Fontenelle occasionally shows brands, mainly on Instagram stories and some partnerships on her feed16 . --- JOURNALISTS OR DIGITAL INFLUENCERS? As individuals who have made a career out of social media, digital influencers can be considered experts in the uses of social networks like Instagram, YouTube and Twitter. Based on that, it is natural to predict that anyone using social media according to some standards of success is a digital influencer or is trying to become one. But considering that these individuals are a specific type of media actors, professionals or celebrities, good use of social media is not confined to them. That is why it is important to assume that journalists can be influential as part of their work, but they are not necessarily digital influencers as a career. --- ISSN: 2763-8677 But when the journalist "becomes" a digital influencer, another part of their practice is in the game, like the possibility of being a columnist or even a commentator. Journalists who keep a distance from controversies and opinions can enunciate them outside the institutional space on their social and personal profiles. Astrid Fontenelle and Guga Chacra are good examples of that. Both of them gain visibility and strength in a space dominated by the dynamics of digital influencers because of differentiated knowledge, an academic degree, the mastery of several discursive types, experience with travel and global cultures, and other deontological assumptions of the profession. In consonance with Vasconcellos , it would be possible to affirm that more than digital influencers, journalists are information professionals -and that commitment is irrefutable. On the other hand, social media can also work as a leisure moment, disassociated from institutional codes or ethical duty. Evaristo Costa uses Instagram as a possibility to create new bonds with the public and disclose a more intimate personality that has nothing to do with his professional one. As a result, he tends to also be considered for brands and other types of partnerships since he uses more personal and detached strategies online. The authority granted to journalists partly relies on the public perception of ethical conduct and a posture of a representative of citizens. In other words, ethical, social and moral principles have a huge impact on journalists' process of legitimation. Because of that, proximity to social media is an important source of credibility among the audiences, as the impacts of relatability can change perceptions of professional value. Conversely, the legitimation process for digital influencers varies according to the influencer niche but it has little impact on how ethical the digital influencer is. This is a crucial difference between the two professions. Besides that, based on the theoretical framework proposed and the empirical analysis it is possible to derive a practical tool: a scale of how close a journalist is to digital influencer practices or vice versa. The scale is one possibility for quantifying which elements move a journalist closer to and which move them away from digital influencer practices. It makes it possible to identify if a journalist's behavior is nearer the behavior of digital influencers, if it mainly follows journalistic practices, or if it is still in the balance of both. ISSN: 2763-8677 The 13 elements serve for both professional categories with their own specifications. Astrid scored 8 points in the journalist degree and 5 points in the midpoint, balanced degree. This illustrates the qualitative findings: Astrid Fontenelle is constantly employing her ethical values, as well as her textual and transmedia expertise to create new forms of sociability online and even expand her work. Astrid is a watchdog of citizens issues, a close friend and a proud mom for her audience. Evaristo Costa, in turn, scored 5 points in the digital influencer degree, 4 points in the journalist degree and 4 points in the midpoint, balanced degree. The metric reveals a more balanced score and also a behavior that is closer to digital influencers. Perhaps it is not a coincidence, then, that he is the journalist with the largest number of followers. But the score also confirms the qualitative analysis: Evaristo is the only one who uses Instagram as a tool for entertainment and a possibility to define a new performance of himself online. A more generic discourse and less attachment to social issues seem to guarantee another possibility of bonding with his followers and a different locus of work. Guga Chacra scored 11 points in the journalist degree and 2 points in the midpoint, balanced degree. Again, it corroborates the finding: Guga does not use Instagram at its full potential and does not exploit features or even captions. Social media is a space to extend his institutional credibility and academic/ 17 A high-quality image is available at: https://drive.google.com/file/d/1WaKGqUtlaHxiSF6R8T3BWZBhqf9kMCnu/ view?usp=sharing ISSN: 2763-8677 specialized authority with equal portions of proximity and intimacy more related to social media in general than the practice of a digital influencer. --- CONCLUSIONS This paper had two guide questions: 1. What are the similarities and differences between journalists' online practices and that of digital influencers'? 2. How have social media and the intimate perspective shortened the distances between journalists and their audience? The first question has already been answered, but an extra observation is required: everyone is confined to a visibility society. We are all, on different levels, part of this culture and so are journalists. The appropriation of some practices of digital influencers is natural for journalists, doctors, teachers and different liberal arts professionals since digital influencers seem to master social media rules and possibilities. They can all be influential people in their own professional or social field -influence is an attribute not restricted to social media -, but being a digital influencer as a career is different and demands specific abilities and competencies. Considering the other questions, it is be possible to define a thesis: 1) Journalists are recognized from broadcast and not native to social media like digital influencers are. As a consequence, some social credit needs to be redefined and even granted again on social media, and sharing intimacy is a way to do that; 2) Journalists use Instagram because traditional media is facing a loss of prestige and a need to be more convergent. So, journalists themselves use their profiles on social media to promote their own TV shows: a strategy to move audiences from digital to television; 3) On different levels, journalists recognize the importance of digital influencers and take some advantage of their narratives and content formats to try to seek the same kind of visibility and success/ fame online. Inspired by digital influencers' performance, journalists guarantee their own authority on social media. It is not enough for journalists to control knowledge through the news, it is also necessary to establish communication networks with the public online. This reconfigures the professional prestige that was intertwined with the types of authority historically held by journalists but now lies within social media. The impacts of that on institutional media and its credibility could be investigated in future works. But what Astrid, Evaristo and Guga highlight is a reconfiguration of journalism and its prestige. Their examples map out how to enter fields where journalists have not yet occupied but are, as usual, extremely necessary.

This article answers two main questions: What are the similarities and differences between journalists' online practices and that of digital influencers'? How have social media and the intimate perspective shortened the distances between journalists and their audience? To understand this scenario, a trio of Brazilian television journalists was investigated to comprehend their digital presence. The authors elaborated an exploratory methodology for qualitative analysis of Instagram publications. Based on elements that characterize the digital influencers and journalistic practices, a score was also created to understand the potentials of their discourse, performances and content strategies online.

Background Depressive symptoms are a well-recognized public health problem throughout the globe, and older adults are more likely to suffer from depressive symptoms [1]. The data from World Health Organization have recognized DS as the most common mental disorder affecting approximately 7% of the population worldwide [2]. However, mental health among older adults has not been a priority area in many low-income countries [3]. Epidemiological studies from South Asian countries reported a high prevalence of depression among older people, ranging from 34.4% in India [4], 40.6% in Pakistan [5], and 44.2% in Bangladesh [6]. A recent study from Nepal showed, 60.6% of older adults had DS [7]. Despite its higher prevalence, there is poor access to early diagnosis and treatment of depression is generally held back because of social stigma with mental health issues and insufficient screening and treatment services in primary health care [8,9]. Symptoms of depression are not usually recognized by families, caregivers or by older adults themselves because of a lack of understanding of the condition, shame or denial that these conditions need medical care and a perception that DS are incurable in Nepal. Various short and long-term factors across the life cycle influence the development of DS. The presence of chronic disease and socio-demographic factors such as age, gender, illiteracy, income level, marital status, mistreatment and lifestyle behaviors play a role in the development of DS [5,[10][11][12]. Early screening and management of depression can improve the quality of life of older adults [1]. However, healthcare systems in low income countries like Nepal are not able to deal with mental health problems including DS among this target population. Nepal is a signatory to the Comprehensive Mental Health Action Plan and to the Sustainable Development Goals , which aim to reduce mental health problems through implementing strategies for promotion and prevention, and providing comprehensive and social care in community-based settings [13,14]. However, performance falls well short of these objectives. To date, there have been no systematically conducted studies in rural parts of south-eastern Nepal where most of the marginalised communities resides. In this line, this study could provide evidence to guide policies and programs to tackle this overwhelming problem. Hence, the present study aimed to address a critical knowledge gap by assessing the level of DS and its relationship with socio-demographic characteristics, lifestyle factors and multi-morbid conditions among community dwelling older adults of Morang and Sunsari district , Nepal. --- Methods --- Study design and participants A community-based cross-sectional study among older adults aged 60 or above was conducted between January and April 2018. A multi-stage cluster sampling approach was adopted to select study subjects in the rural setting of the Morang and Sunsari districts. Primary data was collected from 794 older Nepalese people adults' aged ≥60 years. In the first stage, eight Rural Municipalities were randomly selected from the list of RMs within each of the Morang and Sunsari districts . Second, five wards in each of the selected RMs were randomly selected. Finally, samples were selected randomly from the list of eligible subjects and were interviewed by the trained interviewers. The detailed methodology of this study has been published previously [15]. --- Measures --- Socio-demographics and lifestyle factors A semi-structured questionnaire was used to collect information on socio-demographic profiles and behavior and lifestyle characteristics from the study participants. Socio-demographic and lifestyle variables are described in our previous work that is published by Yadav et.al [16]. --- Chronic disease history Data was collected on the presence of Diabetes, Cardiovascular disease, Arthritis and Chronic Obstructive Pulmonary disease from respondents . Information on self-reported chronic disease was then verified either from medical records or by asking to see any prescribed medicines which patient was taking for the relevant condition. --- Depressive symptoms The 15-item Geriatric Depression Scale was used to assess depression symptoms among our study population [17]. This instrument has previously been validated and used in Nepalese older population [18]. Scores range from 0 to 15. A test score from 0 to 5 is considered normal and patients with test scores greater than 5 were assessed as having depressive symptoms. Factor analysis was performed and the Cronbach's alpha for GDS was found to be 0.81, indicating high consistency. --- Data collection The study used both self-constructed semi-structured questionnaire and a structured validated short form of the Geriatric Depression Scale to collect the information. The semi-structured tool was constructed through a rigorous literature review and with input from an expert already working in the field of geriatric health. The English version of the questionnaire was first translated into Nepali and back translated into English to check the consistency. Pretesting of the questionnaire was carried on a similar group of participants to assess its acceptability and validity before the study began. The data was collected by four trained enumerators via face to face interviews with the subjects in the community setting. The study received ethics approval from Nepal Health Research Council, Nepal. A written informed consent form was obtained from each study participants prior to interview. --- Statistical analysis Descriptive analysis was performed to assess the distribution of the variables. The χ2 test was employed to compare the prevalence of DS within different categories of a variable with 5% level of significance. Since the survey data were nested in nature with variations among clusters , we used a mixed-effect logistic regression model to determine the true association between DS and associated factors. We considered sociodemographic, lifestyle and chronic disease variables as fixed effects and cluster variation as random effects. The parameters of the model were estimated through the generalized estimating equation approach by considering exchangeable correlation structure among clusters [19]. The full model was run with those variables showing P < 0.25 in the unadjusted analysis. We used 2 models for multivariable modeling: in model 1, we analyzed the association between DS and socio-economic as well as lifestyle variables, whereas in model 2 we analyzed the association between DS and chronic diseases adjusted for both socio-demographic and lifestyle factors. Both unadjusted and adjusted odds ratios were reported with 95% confidence intervals . All analyses were performed using the statistical software Stata . --- Results --- Background characteristics A total of 794 people aged 60 years or above participated in the study with a mean age of 69.9 years, where 55.4% of the participants were aged between 60 and 69 years and 15% aged 80 years or above. As summarized in Table 1, the male-female ratio was similar among the study population . A majority of the participants were Hindu and illiterate . Nearly 38% of the participants were of Indigenous origin, and nearly 20% were Dalit. Only around half of the older adults reported being married at the time of the survey. In terms of occupational status, 54.2% of the people reported not being involved in any type of income generating activities. As such, around half of the participants had a family income of 5000 NRs or less. Of the participant's, 62.2% reported smoking history and 48.2% used to consume smokeless tobacco. In addition, 36.5% of the participants reported alcohol consumption, and 77.1% did not take part in any physical exercise . Moreover, the prevalence of Osteoarthritis, CVD, Diabetes and COPD was 41.7, 2.39, 5.3, 14.6% respectively and 14.6% of the participants were suffering from at least two of these comorbidities . --- Prevalence of depressive symptoms Overall, 55.8% of the participants were suffering from DS. Univariate analysis revealed that, DS were significantly higher among female, participants from Sunsari district, Buddhist, Dalit Madhesi ethnicity, illiterate, unemployed and participants with low family income . Prevalence was also higher among those who smoke, do not exercise, and suffer from Osteoarthritis, CVD, COPD, multiple morbidities . --- Association of risk factors with depressive symptoms In the unadjusted analyses, age, gender, religion, ethnicity, marital status, literacy, occupation, family income, smoking history, physical exercise and having multiple morbidities were shown as moderately to highly significant in association with DS . However, after adjusting all the potential covariates in the multiple logistic regression model, only gender, religion, ethnicity, occupation, family income, smoking history and having multiple morbidities remained as significant risk factors for DS at 5% level of significance. It was found that, female participants had 61% higher odds of suffering from DS compared to their male counterparts . An individual of the Buddhist community had around 1.9 times higher odds of suffering from DS compared to a Hindu person , and compared to people of Brahmin/Chettri/Thakur origin, an individual from Dalit community had 2.6 times higher odds of suffering from DS . Notably, an unemployed person had nearly two times higher odds of suffering from DS compared to a person engaged with any income generating activities . Participants with a smoking history had around 50% higher odds of suffering from DS than those who never consumed tobacco . Economic status was also identified as a prime risk factor as the participants with a family income of 5000-10,000 NRs had 77% higher odds of suffering from DS than those with a family income of 10,000 NRs or high . Also, a person suffering from multiple morbidities had 67% higher odds of suffering from DS than those with no multiple morbidities . --- Discussion This study was an investigation that demonstrates an association between socio-demographic variables, lifestyle characteristics and chronic conditions with DS among the older adults in rural setting of Nepal. Our results showed that being female, Buddhist, Dalits and Madhesi ethnicity, unemployed, low family income, the presence of multiple morbidity and smoking are significant risk factors for DS among older adults in Nepal. Nearly 55.8% of the older adults were found to be suffering from depressive symptoms. This is lower than reported from other studies in Nepal [7,18]. A possible explanation for the heterogeneity of prevalence in Nepal could be that different study settings were used, or samples were drawn from diverse ethnicities and castes from the southern plain region were included. The prevalence was also higher than that of a community-based study conducted in rural setting of India and Bangladesh [4,20]. In support of our findings, research from low and middle income countries noted a higher proportion of depression among the older population than that of high income countries such as USA, Canada and Australia [21][22][23][24]. Our study found that, social determinants were associated with DS and females were more likely to have DS. This is consistent with findings reported by Chalise and Rai from Nepal where authors reported females were more depressed than their male counterparts [25]. This is also supported by the findings from a cohort study that showed a clear link between depressive mood and female gender [26].A possible explanation for this could be that females in a patriarchal society in Nepal may have low selfesteem, low social status and empowerment, feeling helplessness, low health literacy, a longer life expectancy and limited access to health services compared to the males. This is supported by the low position of Nepal in the gender inequality index 0.476 [115th position] in the world, which depicts the disparity of health across genders [27]. In the present study, the subjects who ascribed to Buddhism religion was also found to be more depressed, which could be because of the poor socio-economic status of this minority group in the study setting. However, we are cautious in interpreting this result because of the small number of participants in this group. Moreover, we could not find any other studies confirming this association. In contrast to another study from Nepal [7], we found smoking was associated with DS in our study population. Our finding is in line with the findings to that from China and Japan [10,28]. The underlying mechanism linking smoking and DS are complex: Nicotine has antidepressant properties that release dopamine in the mesolimbic reward pathway, which in turn elevate the mood and alleviate stress [29]. However, evidence suggests that smoking plays an important role in changing neurophysiology that increases a smoker's risk of DS [30]. It is also evidenced that smoking is the vascular risk factor for vascular depression [31]. Additionally, one author mentioned that association between smoking and mental disorders are a result of shared environmental and genetic factors [30]. Literacy, an important socioeconomic factor, was not found to be significant with the occurrence of DS. In contrast, the studies from various settings showed illiteracy as a strong predictor of DS [5,7,32]. The underlying reason for this discrepancy is not clearly understood. Unemployment and insufficient family income were both associated with DS. The possible explanation could be that older adults engaged in low income occupation might have led to financial constraints in their life, eventually leading to depression. Adding to this, 2019 Human Development Index has revealed that 34% of the Nepalese population is multi-dimensionally poor and 23% is vulnerable to multidimensional poverty which shows that the country has a deep chasm to fill when it comes to addressing health equity [27]. Meanwhile, in the Nepalese context, most of the older adults depends on family members/caregivers for their daily needs, where families tend to provide better care if the older adults have some economic resources. This situation may turn to increased conflict at the family level, which in turn puts older populations at risk of mistreatment, which may increase the risk of depression [16,33]. However, in this current study we haven't accessed the association between elder mistreatment and depression, and we suggest the need of future studies to check this hypothesis. In the Nepalese context, caste/ethnicity has been a central feature to describe the level of poverty, poor health literacy and health status. In this light, our study demonstrated that the risk of DS was two times higher in Dalits compared to those of higher caste. Emerging evidence shows that Dalits experience a wide range of social and economic discrimination at various levels and this results in a severe form of health inequalities [15,33,34]. Stigmatization is pervasive and worsens psychological stress for those in the Dalit communities [35]. The prevalence of depression among the older population suffering from chronic multi-morbidities was significantly higher than those with one or fewer conditions. This is consistent with findings from other studies [36,37]. This is the first study from Nepal to report an association between multi-morbidity and DS among Nepalese older adults. This finding is consistent with a meta-analysis that showed chronic illness to be a major risk for depression among older adults [38]. Evidence has shown that people among whom depression co-existed with multi-morbidity may have more functional impairment, poorer quality of life and increased mortality [39]. Depression impairs independence in the older population and, over time, worsens functional outcomes among the multi-morbid group. The decline in functional status may make them more dependent and vulnerable to mistreatment. Our findings underscore the need for programs to detect, prevent and manage DS in these groups. More specifically, programs need to adopt a population-based approach that includes screening and diagnosis of DS among older adults, mental health literacy of patients and use of evidence-based practices to manage the DS in the community setting particularly focusing on women and marginalized ethnic minorities. In addition, the Government of Nepal is in the process of scaling up the "Package of Essential Non-Communicable disease " throughout Nepal. This is a great opportunity to consider the establishment of health and wellness centers under this program to encourage participation of older adults in yoga, relaxation techniques and meditation, which may be useful for good psychological health. Furthermore, peripheral health care professionals, community health workers, psychologists and psychiatrists need to work together to reach and treat older adults with depression in rural Nepal. Strengths of our study include its large sample size and high response rate . Another strength of this study includes the strong methodology and use of locally trained enumerators for data collection. However, our study is subjected to certain limitations too. All the associations in this study were cross sectional and cannot evaluate causality. There is a need for longitudinal studies to establish this casual pathway and evaluation of interventions that could address the impact multi-morbidity. As this study only involves the participants from two districts of Nepal, findings cannot be generalized to other setting of Nepal. Another limitation was the use of self-reported data, where socialdesirability bias may have occurred. --- Conclusions A large proportion of the Nepalese older populations were found to have DS, and these were associated with socio-demographic characteristics, lifestyle factors and the presence of multi-morbid conditions. Our findings emphasize the need for programs to detect, prevent and manage depression among the older adults in the community setting. We also suggest the need of targeted screening for DS in the clinical setting focusing on high risks groups such as older adults with multiple morbidities or any chronic conditions, stressful environments or social isolation. This may help health professionals to intervene early in order to avert worsening of the condition. --- --- --- --- Competing interests The authors declare that they have no competing interests. ---

Background: Depressive symptoms (DS) are a well-recognized public health problem across the world. There is limited evidence with regard to DS and its associates, such as socio-demographic characteristics, lifestyle factors and chronic conditions in low-income countries like Nepal. In this study, we aimed to assess the level of DS and its relationship with socio-demographic characteristics, lifestyle factors and chronic disease conditions among community dwelling older people in Nepal. Methods: We conducted a cross-sectional study of 794 older adults aged 60 or above residing in the rural setting of the Sunsari and Morang districts of eastern Nepal between January and April 2018. Multi-stage cluster sampling was adopted to select the study participants. Data included socio-demographics, lifestyle factors, self-reported chronic disease conditions and the Geriatric depression scale. On Geriatric depression scale, an older adult with a test score greater than five were defined as having depressive symptoms. Determinants of DS were estimated through the generalized estimating equation (GEE) approach by considering exchangeable correlation structure among clusters. Results: In our study samples, nearly 55.8% of the older adults were found to be suffering from DS. We found a significant association between DS and being female (aOR: 1.25, 95% CI: 0.89-2.09), Buddhism (aOR: 1.95, 95% CI: 1.58-2.42), Dalits (aOR: 2.60, 95% CI: 1.19-5.65), unemployed, low family income (aOR: 1.77, 95% CI: 1.07-2.92), smokers (aOR: 1.49, 95% CI: 1.01-2.20) and having chronic multi-morbid conditions (aOR: 1.67, 95% CI: 1.09-2.55). Conclusions: The prevalence of DS was high among community-dwelling older adults in eastern Nepal. Our findings suggest the need for mental health prevention and management programs targeting the older population in rural Nepal.

the different groups belong. Given that biology and genes clearly play important roles in people's mental and physical health , there is certainly some merit to this explanation. The second explanation is that the differences reflect a social problem, health disparities. Health disparities refer to systematic variations in the mental or physical well-being of members of different social groups that specifically result from inequitable economic, political, social, and psychological processes . This article only considers health disparities. However, these two kinds of explanations are not mutually exclusive; indeed, recent research strongly suggests that genetic and social factors have a reciprocal relationship as they affect health and a host of other outcomes . One influential contributor to health status disparities is healthcare disparities, which are inequalities in access to and/or the quality of medical care among different social groups. In other words, it is posited that members of socially disadvantaged groups experience poorer health than their counterparts from socially advantaged groups because they receive poorer healthcare. In this article, to limit the scope of this review, we primarily focus on how social psychological processes, specifically those related to intergroup relations and intergroup bias, can produce healthcare disparities. However, we note that there are other literatures on how some of the same kinds of social psychological variables may more directly affect people's health. For example, there is a large body of research on how perceived and actual discrimination directly affect the physical and mental health of members of target groups . Until relatively recently, healthcare disparities have been primarily studied by social epidemiologists, public health researchers, health economists, and political scientists. These scholars have mainly focused primarily on societal and structural causes of this social problem, such as differences in health insurance coverage and access to comparable medical facilities. These remain important and useful approaches to understanding healthcare disparities. Nevertheless, in the last decade or so reports in the United States and Europe on the role of bias and discrimination in healthcare disparities have brought increased attention to the relevance of social psychology and related disciplines to the study of healthcare disparities. This review presents an overview of our own and other researchers' social psychological analyses of healthcare disparities. In the review, we focus on healthcare disparities among members of racial/ethnic minorities because this is the focus of the clear majority of research on healthcare disparities . However, race/ethnicity is not the only group characteristic that has been linked to healthcare disparities. For example, patients who have lower socioeconomic status have developmental, mental or physical disabilities, are elderly, are members of certain religions, and/or are gay, lesbian, bisexual, or transgendered all experience healthcare disparities . Because we base our analysis on generalizable psychological processes and intergroup dynamics, we believe that most of the models and theories we present can be readily applied to these other forms of bias and healthcare disparities beyond those involving race/ethnicity. Our review also primarily presents research conducted in the US. This is because in the US: substantial differences in the health status of Whites and of racial/ethnic minorities, especially those who self-identify as Black and/or African/Afro-Caribbean American have been very well documented ; racial healthcare disparities have received relatively more attention in both the medical and psychological literatures than in other countries; and there has been a long-standing interest in racial bias in social psychological research . The attention given to healthcare disparities in the US has led some health researchers to suggest that serious healthcare disparities are largely confined to countries in which healthcare systems are primarily privately financed, such as the US. There is, indeed, ample evidence that structural differences in the healthcare systems and how they are financed affect patients' access to healthcare and the quality of care provided . However, research clearly shows that throughout the world racial/ethnic minorities typically receive poorer healthcare than members of racial/ethnic majority groups. For example, racial/ethnic healthcare disparities are found in countries with government supported healthcare systems, such as Canada , Israel , New Zealand , Sweden , the United Kingdom , Serbia , and in many Latin American countries . In summary, healthcare disparities among racial/ethnic minorities and majorities represent a ubiquitous multinational problem. In our review of healthcare disparities, we begin by considering some factors other than racial bias and discrimination that might covary with membership in a racial minority and affect the quality of healthcare people receive. We readily acknowledge the importance of these other factors, but a core argument of this article is that even when these factors are controlled, racial-minority patients' and non-minority physicians' race-related thoughts, feelings, and behaviors can each influence the quality of the healthcare members of racial minorities receive and thus produce disparities in health status. This argument is presented in a three-level model of the causes of healthcare disparities that can lead to health status disparities. The levels of analysis --societal, intrapersonal, and interpersonal --are clearly interrelated and overlap with one another, but for the purpose of organizational clarity we consider them separately. After presenting the model, we suggest interventions at each of these levels for addressing racial disparities in healthcare. --- Causes of Racial Healthcare Disparities As Bhopal rightly pointed out in his discussion of racial health and healthcare disparities in the UK and Europe, "[t]he causes of the inequalities in health status and quality of healthcare that are so easily demonstrable by ethnicity and race are complex and difficult to disentangle" . However, several factors have been identified as major contributors to healthcare disparities. --- Socioeconomic Status One major potential source of racial 1 healthcare disparities is socioeconomic status . In many countries SES strongly covaries with race, such that racial minorities on average have lower SES than racial majorities. Low SES is an environmental stressor and has direct, deleterious affects on people's health , but here we are concerned with SES and healthcare disparities. Research has shown that healthcare providers often provide poorer care to low SES individuals perhaps in part because, relative to high SES patients, low SES patients may be seen as having poorer self-control, and less likely to adhere to treatment recommendations . Other potential causes of poorer healthcare for low SES individuals may be that many people with low SES are not able to take time off work to seek medical help or afford health insurance and/or copays. However, racial healthcare disparities exist across all levels of SES and are evident even among the most affluent groups . Statistically controlling for differences in SES reduces but does not eliminate healthcare disparities between racial groups . Thus, SES is an important cause of healthcare disparities but, by itself, is not a sufficient explanation of the racial healthcare disparities reported throughout the world. --- Language Proficiency Research in several different countries has shown that language barriers faced by immigrants often play a role in healthcare disparities among racial minorities. For instance, in the US, limited English language proficiency is associated with decreased utilization of preventive healthcare services , longer hospital stays , poorer health control , and increased medical errors . However, the fact remains that in many countries, healthcare disparities commonly occur among members of racial minorities comprised almost exclusively of native language speakers , suggesting that language barriers only explain a portion of healthcare disparities. --- Health Literacy Health literacy is the degree to which individuals have the capacity to obtain, process, and understand basic health information and services they need to make appropriate health decisions . Health information provided to the lay public may include complex and unfamiliar terms; individuals with limited health literacy may experience difficulty processing such information and become so discouraged or intimidated that they may become reluctant to seek healthcare. In the US, there is a strong association 1 Because so much of the research to be reviewed focuses on healthcare disparities among Black patients, we will primarily use the term, racial minority, for the remainder of this paper. Racial minority, as it is used here, describes a social construction of race as a meaningful category rather than a description of a group's genetic characteristics. between level of health literacy and race ; thus some health researchers argue that health literacy rather than social variables is a major cause of racial healthcare disparities. Indeed, studies in the US sometimes find that when differences in health literacy are controlled, the size of healthcare disparities between racial minorities and Whites in the US is reduced . However, evidence of healthcare disparities tyoically remains even after controlling for health literacy . Although health literacy may be a unique cause of healthcare disparities, in practice it may be hard to disentangle the role of health literacy in racial healthcare disparities from the effects of racial attitudes and beliefs. For example, physicians may be less accurate in judging the health literacy of Black than White patients , suggesting that health literacy's impact on medical interactions may, in part, be moderated by physicians' racial attitudes or cultural competence. Also, health literacy levels among Blacks may be influenced by the fact that they have a much higher level of distrust of physicians and the medical system than Whites , which in turn may cause them to seek medical information less often than Whites and to be less accepting of the information they obtain. Such reactions would serve to reduce a person's health literacy. Taken together, there is substantial evidence that SES, language proficiency and health literacy contribute to inequalities in the relative quality of healthcare different racial groups receive. However, these explanations, which primarily focus just on the attributes of racial minority-group members, do not provide a sufficient explanation of racial healthcare disparities. Social psychological processes also play an important role. An increasing number of public health researchers and government officials have begun to focus on how race-related thoughts, feelings, and actions may relate to racial healthcare disparities. For example, in their article in the European Journal of Public Health, Agyemang, Seeleman, Suurmond, and Stronks argued that "the time is right to take discrimination in health and welfare services more seriously in Europe" . Research reports in the UK, New Zealand, the US, and Sweden have made similar arguments . In the next section, we provide explanations of how and why racial-related phenomena contribute to healthcare disparities and then some possible solutions to the problem. --- Multilevel Processes of Healthcare Disparities The model presented in Figure 1 presents causes of healthcare disparities . The model considers three separate but clearly interrelated processes that might be associated with healthcare disparities: societal, intrapersonal, and interpersonal processes. For purposes of clarity, the three levels in the model are discussed separately, but as the bi-directional arrows in Figure 1 indicate, there are interrelationships and complementary influences among the levels. In this section, we briefly describe each level of analysis. Then, in the next section, we more fully examine specific processes within the intrapersonal and interpersonal levels. In the exposition of these parts of the model, we use social psychological theory and research on explicit and implicit social cognitions, social identity and categorization, and related intergroup processes to help us better understand how racial bias and discrimination may produce healthcare disparities and thus disparities in the health of members of minority groups. Because our focus is on social psychological processes rather than societal/structural processes responsible for racial healthcare disparities, we give much more emphasis to the latter two levels than to the societal/ structural one. --- Societal-Level Processes The first level of analysis, the societal or structural level, focuses on how race-related social environments, institutional practices, and attitudes that characterize a particular society, region, or specific community affect healthcare disparities. At this level of analysis, the presumed causes of healthcare disparities are not primarily psychological processes occurring within and between individuals. Rather, healthcare disparities are presumed to be manifestations of widespread racial bias and/or discrimination. More specifically, they are race-related historical, societal, economic, and structural factors that result in certain racial minority groups experiencing persistent and widespread unfair treatment and institutional discrimination in a variety of realms. One consequence of societal-level processes is that racial minorities may live in difficult social environments, characterized by, for example, less mobility, greater isolation from other segments of the population, fewer treatment options available to them, less exposure to treatment innovations, and fewer high quality medical facilities within close proximity . These challenging social environments may prevent racial minorities from receiving the same quality of healthcare as people living in better-resourced neighborhoods. Bias and discrimination at a societal or structural level may also be related to healthcare disparities in other, less direct, but important ways. Recent research by Reid, Dovidio, Ballester, and Johnson illustrates this process. Reid et al. examined how community level stigma may affect the efficacy of health interventions directed at minority group members. The interventions, conducted in various communities around the US, were intended to increase condom usage among Black Americans. In their meta-analysis of 70 separate studies, Reid et al. found that interventions were most effective when Whites in the community had positive attitudes toward Blacks and residential segregation was low. Reid and colleagues posit that Blacks living in communities with widespread negative attitudes toward Blacks and high levels of residential segregation may feel they are the targets of discrimination and stigmatization and as a result, trust members of outgroups less. When outgroup members offer Black community members health-related interventions, they may be less likely to accept them. Finally, as already noted, feelings of being the target of discrimination may serve as a life stressor for racial minorities, increasing the physiological load on their bodies and thus their susceptibility to various diseases. For example, Sims et al. found in large sample of Black adults a significant association between exposure to discrimination and the prevalence of hypertension . Our model therefore posits that at the societal level, racial/ethnic bias and discriminatory norms, policies, and practices can create social environments and conditions that can directly lead to healthcare disparities. Further, it also appears that bias and discrimination at a societal level may create psychological environments that can also produce healthcare disparities among members of a stigmatized target group. There is considerable overlap between the next two levels of the model, intrapersonal and interpersonal processes; however as noted, we believe that to best understand the psychological processes that underlie healthcare disparities we need to consider them separately. For the purposes of this article, the intrapersonal level concerns what is "within" a person -that is, thoughts and feelings that operate at both explicit and implicit levels and affect people's reactions to a socially defined group in the absence of direct contact with a member of that group and/or in situations where they actually encounter a person perceived to be a member of that group. Specifically, the intrapersonal level of the model concerns how the race-related thoughts and feelings of minority patients and of non-minority physicians affect healthcare related thoughts, feelings, and actions. For example, Black patients' feelings about past racial discrimination may affect things such as the kind of medical care they seek, what they think of the care they receive, the health behaviors in which they engage and perhaps most importantly their reactions to interactions with healthcare providers. In a similar fashion, racial bias and stereotyping among physicians may affect diagnosis and treatment decisions, and perhaps most importantly their reactions to racial minority patients. The interpersonal level in the model focuses on medical interactions between patients from one ethnic/racial group and physicians from another racial group. At this level, the analysis concerns what patients and physicians say and do during these interactions and how each reacts to the other's words and actions. This is an important aspect of racial healthcare disparities in the US because racial minority patients are much more likely than racial majority patients to engage in racially discordant medical interactions . For example, in the US approximately 75% of Black patients see non-Black physicians when they seek medical help . This statistic also reflects the dramatic under representation of racial minorities in medical school and medical practice . For instance, in the US, where Blacks are about 12% of the population only about 4% of all physicians self-identify as Black; this percentage has been the same for about the last 30 years . Although we use the term interpersonal to describe this level of the model, one might quite reasonably also use the term intergroup contact to describe racially discordant medical interactions. In primary care interactions, patients and physicians may often be, at most, only minimally acquainted with one another prior to their interaction. Thus, while the physician and patient are in a face-to-face interaction that involves considerable dialogue, in a racially discordant medical interaction they may react to one another more as representatives of a group than as individuals. From this perspective, such interactions may represent one of the kinds of intergroup contact described by Brown and Hewstone . Specifically, a racially discordant medical interaction may often be an instance of intergroup contact in which intergroup differences are salient and important, but it is also interpersonal because the patient may provide sensitive, personalized information to the physician. As already noted, the race-related thoughts and feelings patients and physicians bring with them to racially discordant medical interactions have an extremely important influence on what transpires during the interaction. Specifically, these intrapersonal processes influence what each party says and does during the interaction, which, in turn, shapes what the other says/does during the interaction and, thus, the quality of communication in the medical interaction. However, as we discuss, there are other interpersonal processes that can affect communication between the participants and we shall consider them as well. Whatever the cause, the quality of communication can affect the quality of treatment decisions each party makes and, in the case of the patients, also influence how likely it is that they will adhere to these treatment decisions. --- Social Psychological Theory and Racial Healthcare Disparities In this section, we review social psychological research on how the ways people think about and respond emotionally to others and how people interact with each other can result in healthcare disparities. --- Intrapersonal-Level Processes A core assumption of our approach to health disparities is that, as the IOM report posited, race-related attitudes and beliefs play a critical role in healthcare disparities and thus are responsible for disparities in health status. Race-related stereotyping and attitudes can be conscious or explicit; that is, people are aware of their thoughts and feelings and can manage them consciously. However, as the result of over-learned racial associations, stereotypes and attitudes can also be automatic, nonconscious or implicit ; that is people may not be aware of these feelings and thoughts . The intrapersonal level of the model concerns processes through which an individual's own racerelated thoughts and feelings affect their healthcare-related thoughts, feelings, and behaviors. While these intrapersonal processes obviously have social origins and consequences, in this section we consider instances where their impact on healthcare disparities is substantially independent of the thoughts, feelings and actions of another person. We consider these intrapersonal processes in patients and physicians separately. Racial minority patients-Black patients prefer to see Black physicians and report higher quality care and greater satisfaction with their medical care when the physician is Black than when the physician is White . Importantly, racial minority individuals' negative perceptions of physicians are directly linked to their healthcare-related behaviors. For example, Black patients are more likely to schedule and keep appointments with Black than White physicians . Additionally, racial minorities are more likely to delay or fail to seek mental healthcare and terminate treatment prematurely , compared to racial majorities. These racial minority patients' negative perceptions of physicians may be at least partially due to lack of trust. When compared to members of majority racial groups, members of racial minority groups are more likely to experience mistrust of and dissatisfaction with their medical care and to believe they are mistreated by healthcare providers . Furthermore, individuals who are less trusting of physicians are less likely to use preventive services and adhere to physicians' recommendations . This relationship is illustrated in data that we collected from a primary care clinic in a large US city . This study is one of a series of studies conducted in this setting, which is described in more detail in a later section. Immediately before Black clinic patients had their appointment with a non-Black physician we measured how much they trusted physicians in general, using two items taken from Dugan, Trachtenberg, and Hall's Interpersonal Trust in a Physician short form: "I completely trust the doctors' decisions about which medical treatments are best," and "All in all, I trust doctors completely." Sixteen weeks after their appointment with the doctor in the clinic, these patients received a mail survey about their adherence to their own physician's treatment recommendations . Figure 2 presents a scatter plot of the relationship between the Black patients' trust of physicians prior to racially discordant medical interactions and their adherence to their own physician's recommendations 16 weeks after the interactions. There was a significant positive relationship = .43, p=.003) between general trust and specific adherence. Although our emphasis is on social psychological explanations of these thoughts and feelings, one cannot consider the attitudes and feelings of racial minority patients without first placing them in a social and historical context. A substantial portion of racial minorities' mistrust of medical care and preference for racially concordant physicians very likely has its origins in the general levels of racism and racial discrimination that are part of the past history of many western industrialized countries. Embedded in this history is the dismal legacy of racist beliefs and practices in the medical profession . For example, in the US and Europe, polygenism-the theory that human races were separate biological species-dominated scientific theory from the early 17 th century until perhaps as recently as the early 20 th century . Blacks, Latinos, and other socially disadvantaged racial minority groups were also frequently used as participants in dangerous medical experiments without their willing consent and with little regard for their welfare . Thus, it cannot be overlooked that some of racial minorities' mistrust of general medical care is rooted in harsh and potentially traumatic historical realities . This mistrust may represent a rational response to a system that has ill-served racial minority patients in the past. However, we believe there are other more basic social psychological processes at work as well. Social categorization and social identity also play important roles in what racial minority patients think and feel about their medical care and their healthcare related behavior. Because human beings are constantly required to process large amounts of complex information, they develop strategies to reduce cognitive effort and preserve cognitive resources by using categorization and generalization of information . Categorization and application of social categories to evaluations, perceptions, and treatment of others clearly help people function more effectively in everyday life. However, the process of social categorization inevitably makes a person's social identity more salient and provides the basis for intergroup bias . In addition, according to Social Identity Theory , when people's group membership is salient, their feelings of esteem are closely tied to their group's status and they respond in ways that promote their group's distinctiveness and enhance its status. People generally assume that ingroup members share their attitudes and beliefs , and they expect outgroup members to have a contrasting perspective . Minority-group members are particularly vigilant to cues of bias from outgroup members . These psychological processes systematically influence the perceptions and behavior of minority patients. In this context, their identity as a member of a racial minority group that has been the target of prejudice and discrimination is highly salient. Expectations of prejudice/discrimination among minority patients play an important role in intergroup encounters generally, and in medical interactions, specifically. Richeson and Shelton demonstrated that when Blacks interact with Whites, they often worry they will be the target of prejudice and/or confirm negative stereotypes associated with their racial groups. Thus, one source of racial minority patients' possible negative reactions to the healthcare system may be expectations they will experience stereotyping, prejudice, or discrimination in their medical care. Related to this, Burgess, Warren, Phelan, Dovidio, and van Ryn argue that racial minority patients in racially discordant medical interactions may experience stereotype threat, which involves the activation of cultural stereotypes about one's group and triggers stereotype-confirming behavior. This may have particularly detrimental effects on the responses of racial minorities to healthcare interventions that might reinforce these stereotypes. For example, some Black patients could be reluctant to participate in classes that provide health information because they may fear this would reinforce stereotypes about their intellectual abilities and educational achievements. However, such feelings are probably not uniform across racial minority group members; there are clearly individual differences in recognition and expectations of racial bias. Theories such as stigma consciousness or race-based rejection sensitivity concern the chronic affective and cognitive consequences of being the target of stereotyping, prejudice, and discrimination. According to these theories, individuals who score high on stigma consciousness or race-based rejection sensitivity often expect to be the target of prejudice. Research has shown that individuals who expect to be the target of prejudice evaluate intergroup interactions more negatively and make efforts to avoid them . In addition, it has been shown that when individuals expect to be a target of prejudice and discrimination, they become vigilant for signs of prejudice and discrimination , resulting in a confirmation of their negative expectations. In the healthcare context, heightened sensitivity to discrimination and prejudice due to past experiences may influence how racial minority patients perceive their physicians and thus healthcare related decisions. For example, Greer found that Black hypertension patients who perceived their physicians to be racially biased were less likely to make future appointments than patients who did not perceive their physicians in this way. Importantly, general expectations of prejudice and discrimination that are not directly related to healthcare or specific healthcare facilities/personnel can also affect patients' health-related attitudes and behaviors. Benkert, Peters, Clark and Keves-Foster reported that the more racism Blacks perceive in the world around them, the less trust they have in the healthcare system. Perceptions of general discrimination -including experiences of bias outside the medical encounter and in everyday life -can also affect reactions to individual medical interactions. We used a larger sample of Black patients at the same innercity clinic employed in the Hagiwara et al. study to examine the impact of perceived discrimination. Immediately prior to the meeting with the doctor, we assessed patients' general experiences with discrimination. To do this, we used items from Brown's Self-Perceived Racial and Ethnic Discrimination Scale, in which respondents were asked whether they had ever experienced unfair treatment in each of seven social domains: jobs, education, medical treatments , job applications, police encounters, housing, and dealing with neighbors. In our sample, the majority of patients reported they had experienced discrimination in the past at least in one domain. In mail surveys, sent 4 and 16 weeks after their appointment with the doctor, patients reported on their adherence to the doctor's recommendations , as well as their self-reported health using the 20-Item Short Form Health Survey , which assesses physical functioning, role functioning, social functioning, mental health, pain, and current general health status. We found that, when compared to Black patients who reported experiencing relatively little past discrimination, those who reported experiencing high levels of past discrimination expressed significantly less satisfaction with their medical interactions and closeness with the physician they had just seen. Black patients who experienced higher levels of discrimination were also less likely to adhere to their physician's recommendations four weeks after the visit, and this lower adherence, in turn, was associated with poorer health status among these patients 16 weeks after the visit. Casagrande, Gary, LaVeist, Gaskin, and Cooper also found that experiences with discrimination were associated with delays in seeking medical care and poor medical adherence even when controlling for medical mistrust. Racial minority patients' negative health-related attitudes and behaviors may not always be the result of negative perceptions of physicians and medical care. They may also sometimes be the result of strong identification with their own racial group and consequent identitybased motivations. According to the Identity-Based Motivation Model , in order to maintain their social identity, people tend to engage in behaviors that are perceived to be what other ingroup members typically do, regardless of whether the behaviors have objectively positive or negative consequences for them . Thus, if individuals perceive that exercising and maintaining a healthy diet is the norm for their social group, they are apt to engage in regular exercise and healthy eating habits. In contrast, if individuals believe that many fellow ingroup members smoke and/or value smoking, they are likely to smoke even if they recognize the negative consequences of smoking. Oyserman et al. have shown that racial minorities viewed health promotion behaviors, such as exercising, getting enough sleep, and eating fruits and vegetables, as outgroup-defining behaviors and unhealthy behaviors as ingroup-defining behaviors. Oyserman et al. also investigated the effects of increasing the salience of racial identity among racial minorities. When the minority participants' social identities were made salient , they expressed more feelings of fatalism about improving their health and had more difficulty correctly identifying health-related knowledge. --- Non-minority physicians- There is more than ample evidence that racial minorities receive less appropriate and aggressive healthcare than members of racial majorities. Disparities have been found in the general level of care Black and White patients receive and in the treatment of specific diseases. These disparities cut across specific mental and physical diseases, and healthcare settings . Regarding mental health, racial minorities are less likely to receive mental healthcare and receive poorer quality care when they are treated . For example, Mallinger and Lamberti studied Black and White patients with schizophrenia who were receiving antipsychotic medications. They found that Black patients were less than likely than White patients to receive other medications to control the ancillary symptoms of schizophrenia. Mental healthcare disparities remain even when SES and relevant demographic factors are controlled . Turning to physical health, racial disparities are also found in the treatment of cardiac disease and many different cancers, including breast, lung, prostate, colorectal and other gastrointestinal cancers . One might expect that in situations in which swift and decisive action is required racial healthcare disparities might disappear, but treatment disparities also exist in emergency room settings. For example, racial minorities with symptoms of a heart condition are less likely than White patients to receive analgesics, and their myocardial infarctions are more likely to be missed. These disparities exist even when one controls for levels of disease severity and factors such as SES, demographic characteristics and insurance availability . Finally, healthcare disparities can be found in pediatric practices as well. In a study of over one million medical encounters, Gerber et al. found that physicians were significantly less likely to give antibiotics to their Black than their White patients, and if antibiotics were given to Black children they were less likely to be broad spectrum drugs. It would be an oversimplification to attribute these and a host of other treatment disparities simply to physician race-related thoughts and feelings. Treatment disparities are a complex social problem with multiple causes, many of which are likely at the societal level of explanation, and of course, patient preferences and decisions play important roles in what treatments are provided. In this section, we only consider those treatment disparities that theory or empirical evidence suggest may be due to race-related intrapersonal processes among physicians. As was the case with patient effects, we rely heavily on theories that have their origins in social categorization and social identity processes. These include stereotype activation, contemporary racial bias, and cognitive load. Stereotype activation, stemming from social categorization among physicians, may at least partially explain the apparent contradiction between the widely expressed condemnation of bias in medical care by healthcare professionals and the well-documented treatment disparities that continue to exist. van Ryn and her colleagues have proposed a social-cognitive model that uses social categorization processes to explain how patients' race influences physicians' diagnosis and treatment decisions. According to this model, physicians more or less automatically categorize their patients into social groups based on race, which is often the most salient social group cue. Upon categorization, stereotypes and prejudice associated with racial groups in which the patient is categorized are activated. These activated stereotypes then influence physicians' perceptions of patients and expectations about patient adherence, which in turn affect physicians' treatment decisions. For instance, it has been shown that physicians stereotype Black patients as poorer, less educated, less intelligent, less trustworthy, more likely to engage in risky health behaviors, and less likely to comply with medical recommendations than White patients . These negative stereotypes and expectations about Blacks may further influence physicians' treatment decisions, such that they may be less likely to recommend an appropriate treatment for Black than White patients . For example, van Ryn et al. examined physicians' decisions to recommend coronary bypass surgery for Black and White patients with the same cardiac conditions. They found that physicians were more likely to recommend this procedure to Whites than Blacks, but this effect was fully mediated by the physicians' perceptions that the Black patients were less educated and had a less active lifestyle. Moskowitz et al. have also shown that physicians trust non-White patients less than White patients when prescribing opioid analgesics; this occurs despite the fact that rates of illicit use and opioid analgesic misuse between these two groups are similar. Stereotypes activated by social categorization also influence how people interpret others' behaviors and intentions . Explicit and implicit stereotypes associated with social categories may therefore also influence physicians' diagnostic assessments. As an example, Hirsh, Jensen, and Robinson found that nurses evaluate pain expressed by patients differently based on their beliefs and expectations as to how individuals with certain social characteristics would react to pain. More recently, Moskowitz, Stone, and Childs used a priming paradigm and found that White doctors were faster to recognize diseases and conditions such as HIV, drug abuse, obesity, and stroke after being subliminally primed with a Black face . Moskowitz et al. thus concluded that physicians' initial reactions to patients may be influenced by racial/ethnic stereotypes about the frequency of certain diseases among Blacks relative to Whites. This stereotyping may lead to misdiagnoses and inappropriate treatments. Importantly, in neither the Hirsch et al. or Moskowitz et al. study did the participants have any conscious awareness of their biases. These findings suggest that healthcare professionals may be biased "even in the absence of the practitioners' intent or awareness" . No study to date has systematically examined how accurate physicians' negative stereotypes about racial minority patients are. However, as noted above, racial minorities are indeed less likely to adhere to physicians' medical recommendations and keep appointments . Thus, as with any other stereotypes, there may be some basis to some of physicians' expectations about racial minority patients. Nevertheless, the major problem with stereotype activation is that negative expectations based on stereotypes are generalized to the entire social group regardless of whether particular individuals actually possess stereotypical characteristics. Furthermore, when physicians act on their negative expectations about racial minority patients, the patients may ultimately confirm the negative expectations by acting in accord with these negative stereotypes , thus resulting in a vicious circle. For example, physicians who expect their minority patients to be less adherent to their medical recommendations may devote less time to discussing the condition and treatment with minority patients, which then leads minority patients to understand the recommendation less well. This lack of understanding the doctor's recommendation, in turn, may produce less adherence -creating the self-fulfilling prophecy in the medical encounter . Within the medical profession, explicit expressions of racial bias are especially rare, and behavior that might represent such bias in medical care is widely and vigorously condemned . Despite this, there is evidence of more subtle, perhaps often unintentional, contemporary racial bias. The majority of physicians show a moderate to strong implicit preference for Whites relative to Blacks . Moreover, such biases do affect treatment decisions. For example, in a vignette study, Green et al. found that implicit racial bias affected physician recommendations for Black patients with acute coronary syndromes. As implicit racial bias increased, so did the likelihood that physicians would treat a White patient more aggressively and appropriately than a Black patient. More recently, Sabin and Greenwald used vignettes involving pediatric patients and found that as physicians' pro-White bias increased, their willingness to prescribe narcotics for Black children with pain due to surgery decreased. Theories of contemporary racial bias, such as Aversive Racism Theory , Symbolic Racism , and the Justification-Suppression Model , would lead us to expect that such discriminatory actions are most likely to occur when a situation allows people to rationalize that their actions toward racial minorities are not due to racism but due to something else. In the health context, one situation in which this may occur is when the guidelines or cues for appropriate medical decisions are somewhat ambiguous. For instance, few guidelines for treating pain exist because many patients' reports of pain are diffuse, nonspecific, and difficult to empirically verify and quantify . Studies have shown that, as predicted by these theories, across a variety of different kinds of pain and medical settings, racial minority patients are more likely to be undertreated for pain than Whites . Burgess et al. argued that implicit stereotypes about Blacks may cause physicians to either underestimate the seriousness of the pain or overestimate the likelihood a patient might abuse pain medications. However, importantly, in this situation, physicians could also rationalize the under-treatment of Blacks by using other factors that are much more justifiable and apparently nondiscriminatory than their bias against Blacks. Trawalter, Hoffman, and Waytz found that people believe that Blacks and individuals from lower SES feel less pain when injured than White upper SES individuals. Such biased perceptions may also be held by physicians. Finally, we consider some circumstances in which physicians may be more likely to make less appropriate diagnosis and treatment decisions for racial minority patients than racial majority patients. The cognitive load that physicians experience when making their decisions may be a significant factor in their responses to minority, relative to majority, patients. A depletion of cognitive resources due to increased cognitive load is likely to increase the probability that people will engage in automatic rather than controlled cognitive processes because they need to prioritize multiple cognitive processes. When one encounters a member of an outgroup, the activation and application of social categories are rather automatic and require minimum cognitive efforts, whereas individuation of that person requires conscious efforts . Therefore, when people are under high cognitive load, they are more likely to rely on social categories and stereotypes as opposed to individuating information . As mentioned earlier, Richeson and Trawalter showed that racially discordant interactions often activate Whites' need to self-regulate their behaviors in order to avoid appearing to be racist. This engagement in self-regulation depletes their cognitive resources. Because contemporary physicians are encouraged to act and treat all patients in the same way, they are likely to engage in self-regulation which in turn results in depletion of cognitive resources. Additionally, among physicians, such effects may be exacerbated by their workload. In most industrialized countries, caseloads and work demands on physicians have increased dramatically . Thus, physicians may be more inclined to rely on stereotypes associated with their patients' racial groups as they make diagnostic and treatment decisions in higher-stress clinical situations. In summary, intrapersonal processes in the form of explicit and implicit race-related thoughts and feelings affect what minority patients and their physicians think and do in regard to healthcare. Although some of the theories and models we have presented have not been directly tested in healthcare settings, those that have provided substantial evidence for the utility of social psychological theory for an understanding of how intrapersonal processes can help us understand racial healthcare disparities. --- Interpersonal-Level Processes As already discussed, there is a high probability that when a member of a racial minority group sees a physician, the physician will not be a member of this group . These racially discordant medical interactions are usually qualitatively different from racially concordant medical interactions. It has been shown, for example, that, relative to racially concordant medical interactions, racially discordant interactions are shorter in length , less patient-centered , less positive , and involve fewer attempts at relationship building . Furthermore, White physicians spent significantly less time planning treatment, providing health education, engaging in informal conversation, and answering questions with Black than White patients . We have explored interpersonal-level processes in medical interactions in cancer treatment settings. We collected data over five years at two large comprehensive cancer centers in different regions of the US, one in Detroit, Michigan and the other in Tampa, Florida. At the center of this research are analyses of 235 video recordings of discussions between cancer patients eligible for clinical trials and their oncologists . Although, one might think the video recording of these interactions would affect the behavior of patients and physicians, in fact, we have found that after the first minute or so of the discussions, they pay almost no attention to the camera A portion of these discussions were racially discordant medical interactions between Black patients and their non-Black physician . Our analyses of these interactions focus primarily on communication between the parties. The studies we describe here do not contain data on either patient or physician race-related thoughts or affect. They do, however, speak to healthcare disparities. For example, we have conducted two separate kinds of analyses to compare how much information physicians provide to Black and White patients when they discuss clinical trials. In one study coders watched the video and used a checklist to record how many of a list of potential drug side effects were discussed with patients. Physicians were almost twice as likely to mention any side effects with White than Black patients and, if side effects were discussed, Whites had more of them described to them than did Blacks. Eggly, Barton, Winckles, Penner, and Albrecht conducted a discourse analysis of discussions of clinical trials with two matched small samples of Black and White patients. In the discourse analysis, transcripts were made of all the conversations. Topic areas were identified from the transcripts, and then the words spoken by the physician about each of these topics were counted. Overall the physicians used more words in the discussions with Whites than with Blacks, and said more words specifically about clinical trials = 1.06). Furthermore, physicians spoke more about the study purpose and about risks to White than Black patients . We have also found that Black cancer patients are less likely than White cancer patients to receive important general information in oncology settings. Eggly et al. studied the frequency of direct questions and indirect questions . Eggly et al. found that Black cancer patients, as compared to White cancer patients, asked fewer questions, and a smaller proportion of these were direct questions . These findings point to the fundamentally different nature of the interpersonal exchange that occurs in racially discordant than racially concordant medical interactions. We now consider some of the social psychological processes that may explain these effects. Because the behavior of patients and physicians is highly interdependent in medical interactions, we discuss the influence of social psychological processes on the patientphysician interaction rather than on patients and physicians separately. Social identity-As previously discussed, the process of social categorization, which conserves cognitive resources, makes social identity more salient and provides the basis for intergroup bias . According to Social Identity Theory , when group membership becomes salient, as often occurs in racially discordant medical interactions, people derive their self-image from the social categories to which they perceive themselves as belonging. Thus, they strive to achieve or maintain positive social group images. Group membership is very likely to be salient in racially discordant medical interactions as racial membership is one of the most frequently used features for categorizing people into different social groups. Consistent with this, Street, O'Malley, Cooper and Haidet reported that patients feel more personal and ethnic shared identity with their physician in racially concordant than racially discordant medical interactions. Several processes are associated with social identity and attempts to achieve or maintain positive social group images. We posit that these processes may contribute to racial disparities in the quality of medical interactions. Communication across group lines is often less effective than within groups not only because people are generally less accurate at perceiving expressions of emotion displayed by outgroup than ingroup members , but also because they are biased in their misperceptions. For instance, people are more likely to perceive a hostile face as belonging to an outgroup member and to misinterpret neutral facial expressions as conveying anger for outgroup than ingroup members . One way to achieve or maintain positive images of one's own social group is to perceive, evaluate, and treat ingroup members more positively than outgroup members and/or outgroup members more negatively than ingroup members . At the same time, ingroup members tend to see members of outgroups as homogeneous and are less trusting of them than members of their own group . In addition, they also expect that outgroup members will be biased against them as well . These differential responses to ingroup versus outgroup members are likely to affect the quality of racially discordant medical interactions. Because both patients and physicians are likely to strive to achieve or maintain their positive racial group images, they may engage in ingroup favoritism, outgroup derogation, or both. As a consequence, healthcare disparities based on ingroup favoritism would likely reflect particularly positive affect toward ingroup patients/physicians, whereas healthcare disparities based on outgroup derogation would reflect negative affect toward outgroup patients/physicians. Divergent interaction goals-Another way to achieve or maintain positive social group images is to present oneself in a positive manner. Shelton, Richeson and their colleagues propose that in interracial interactions, individuals' salient social identities are each associated with different or even conflicting interaction motives or goals. Specifically, in such interactions, participants from racial majority groups may be concerned with not appearing to be prejudiced, while racial minority participants may be equally concerned with not being the target of prejudice or discrimination and/or not confirming racial stereotypes . In support of this latter proposal, Bergsieker et al. found that Latinos and Blacks, but not Whites, were much more concerned with being respected and seen as competent when they were in racially discordant dyads than in racially concordant ones. These divergent interaction goals of majority and of minority participants may undermine effective communication between them. For example, in a recent analysis of physician-patient interdependence in racially discordant medical interactions Hagiwara, Kashy and Penner found that non-Black physicians' perceptions of the extent which they and their Black patients were working as a team during a medical interaction were essentially orthogonal to the patients' perceptions of "teamness." This kind of lack of convergence in perceptions of the interaction is seen by many researchers as a major cause of disparities in minority use of healthcare services . Additionally, when individuals are concerned with self-presentation, they often engage in self-regulation. As described above, self-regulation requires cognitive resources as individuals try to monitor for undesirable behaviors while executing more desirable behaviors during medical interactions. Because cognitive resources are limited, individuals engaging in self-regulation often are unable to attend to other tasks . This self-focus may result in individuals appearing to lack interest in their interaction partners. For instance, Shelton et al. found that when Whites self-regulated their behaviors, apparently to avoid appearing biased, Blacks interpreted this excessive self-regulation as a lack of interest in them. These findings suggest that divergent interaction goals of minority patients and majority physicians and their attempts at self-regulation to attain these goals could result in ineffective communication during medical interactions. Poor communication during the medical interaction, in turn, could also result in more negative feelings after a medical interaction, even if these feelings did not exist prior to the interaction. Some empirical support for this conjecture comes from Gordon et al. , who studied trust among Black and White lung cancer patients. Gordon and colleagues reported that while the two groups did not differ on trust of the oncologists prior to the interaction, Black patients had lower post-visit trust than White patients. Patient perceptions of how the physician communicated with them predicted post-visit trust. Intergroup anxiety and stress-Intergroup anxiety and stress is another process that may also shape racially discordant medical interactions. The theory of intergroup anxiety proposed by Stephan and Stephan also provides a possible explanation for why the quality of racially discordant medical interactions is often poorer than that of racially concordant ones even when physicians and patients may not hold negative attitudes toward and/or perceptions of one another. According to the intergroup anxiety theory, people's expectations that their interactions with outgroup members can go wrong can often result in intergroup anxiety. This general negative arousal, in turn, may result in a desire to minimize interactions with outgroup members or even in hostility toward outgroup members . However, as already noted, physicians and patients often cannot avoid racially discordant medical interactions. Thus, racially discordant interactions may often be awkward, threatening, and distressing . Indeed, Trawalter et al. propose that people's behavior in racially discordant interactions can be understood as "initial stress reactions and subsequent coping responses" . Under such circumstances, the choice of how to cope with this stressor depends on whether the person sees the situation as a challenge or a threat . If people feel they have enough resources to cope with the stressor, they are likely to perceive the stressor as a challenge and respond in an effective, productive manner. By contrast, if people believe they do not have enough resources to cope with the stressor, they see it as a threat and their response is likely to lead to a negative and unproductive interaction. Threatened individuals may, depending on perceived resources, antagonize their partner, psychologically withdraw from the interaction, or as suggested above, overcompensate . All of these responses are consistent with the descriptions of racially discordant medical interactions provided earlier in this section. Explicit and implicit attitudes-Much of our own recent research has focused on the influence of physicians' implicit attitudes, both directly and in combination with their explicit attitudes, on the interpersonal-level dynamics of medical encounters. As we have already discussed, race-related thoughts and feelings can affect non-Black physician and Black patient healthcare-related thoughts, feelings, and actions independent of actual interactions between them. In this section, we consider how race-related affect and cognitions can affect the quality of racially discordant medical interactions. We illustrate several of these processes with our own research in this area, which highlights the distinction between explicit and implicit racial bias. As already noted, although there is typically an association between explicit and implicit processes, this association is often rather small. Furthermore, explicit and implicit racial bias can influence behaviors in different ways . More specifically, explicit bias generally predicts deliberative, controllable behaviors, whereas implicit bias generally predicts spontaneous behaviors that are hard to monitor and control . Thus, individuals may not be able to self-regulate undesirable behaviors stemming from implicit bias, which would affect their interactions with people who are the object of this bias. The impact of implicit physician bias on racially discordant medical interactions has been the focus of much of our research on health disparities. Earlier, in the section on intrapersonal-level processes, we briefly discussed two findings from our research at the primary care clinic showing that if Black patients report experiencing more racial discrimination in their everyday life and/or feeling less trust in physicians , they are less likely to adhere to a non-Black doctor's medical recommendations. In this section, we discuss a series of studies related directly to interpersonal-level processes that draw on various measures in the longitudinal database collected on this sample of Black patients and their physicians at the clinic. Thus, we offer more details of the social environment and research procedures to provide a context for the work. The clinic is in the inner city of Detroit, Michigan. The great majority of Detroit residents self-identify as Black/African American, and Detroit is one of the most racially segregated cities in the US . Over two years, we recruited 156 Black patients at the clinic. . The sample was predominantly female , and all patients had health insurance. Participants were of quite low SES: 45% had annual family incomes of less than $20,000 USD, which is below the official US government "poverty level" for a family of four . Another 30% had family incomes of less than $40,000, and almost 30% had less than 12 years of formal education . There were no systematic effects for patient SES across our studies, perhaps in part due to the limited range of income in the patient sample. The physicians were 19 medical residents. Most were East Asian, which is quite common in the US for primary care clinics in low SES minority communities . Ten of the physicians were women. Patients were approached by a Black research coordinator when they arrived for a scheduled medical appointment; 73% agreed to participate. After being consented, patients provided demographic information and completed the measures that have already been described before they saw their physician. Patients then interacted with the physician. One hundred and thirty-six of these interactions were video recorded. Immediately following the interaction, patients completed several measures, including a 14 item measure of the extent to which they believed the physician was patient-centered , and their perceptions of physician warmth/coldness, and their satisfaction with the encounter . The patients also answered items about trust in the physician they had just seen and two items that measured team perceptions As already noted, four and sixteen weeks after the interaction, we sent patients mail questionnaires asking them about their health and their adherence to their own physician's recommendations . Patients also received the measures of physician-specific and general trust . and controlling for attrition did not change any findings that we report. When the physicians enrolled in the study, they provided demographic information, and 15 of them completed measures of explicit racial bias and a race-based Implicit Association Test . Immediately after the interactions, physicians completed two team items parallel to those completed by the patients. To examine the dynamics of the interpersonal-level processes in these medical interactions, we used the video recorded interactions to examine the amount of time physicians and patients each talked during the interactions; then they computed ratios of physician to patient talk time . Relative talk time was chosen for several reasons. First, relative talk time is considered a valid measure of participants' motivation to dominate an interaction . Second, there is clinical significance to the relative amount of time patients and physicians talk during medical interactions because it is believed to be related to health care outcomes, such as patient satisfaction and adherence. For example, according to the Patient-Centered Communication Model , greater physician talk time is associated with less patient satisfaction and adherence, while greater patient talk time is associated with more satisfaction adherence . And third, people are better at self-regulating how explicit racial bias is manifested in more planned behaviors than they are at regulating how implicit racial bias is manifested in more spontaneous behaviors . Thus, how much time physicians talk may systematically reflect physicians' implicit racial bias. Overall physicians talked more than patients. In addition, as expected, we found that physicians' implicit, but not explicit, levels of racial bias influenced the physician-patient talk time ratio. The higher the level of physicians' implicit racial bias, the more they talked relative to physicians with lower levels of implicit bias . That is, more implicitly prejudiced physicians dominated the medical interaction more verbally. Cooper et al. reported convergent findings for a separate sample of medical interactions. Hagiwara et al. speculated that physicians with higher levels of implicit bias harbor negative attitudes toward and stereotypes about Black patients , which might lead them to want to exert more control during interactions with them. This is manifested in the association between their implicit bias and how much they talked relative to their patients. In the same study, we also examined the impact of patients' reports of previous experiences of discrimination by Whites on physician-patient talk time ratio. As seen in Figure 3b, the greater perceived past discrimination patients reported, the smaller was the physician-patient talk time ratio; that is, participants who reported experiencing high levels of previous discrimination talked more than those who reported experiencing low levels of discrimination. The same pattern was observed for trust-the less the trust, the more the patients talked. At first glance, these findings seem inconsistent with a patient-centered perspective on medical interactions . In attempting to explain this inconsistency, Hagiwara et al. argued that greater patient talk time may not always reflect positive patient reactions to medical interactions. Specifically, we proposed that in racially discordant medical interactions Black patients may fear that, based on their prior experiences with bias and mistrust of the physician, they may become the victims of discrimination. They therefore attempt to assert more control in their medical interactions to achieve higher quality care. Indeed, consistent with this interpretation of greater talk-time among Black patients, we found that the smaller the ratios , the less they subsequently adhered to physician recommendations . We acknowledge that patient-centeredness is a valid and desired goal for all medical interactions, but propose that the behaviors associated with patient-centeredness may differ between racially concordant and racially discordant medical interactions. As, Shelton, West, and Trail have shown, sometimes exactly the same behaviors may be viewed in different, frequently opposite, ways in same-race interactions and different-race interactions. Thus, our research further highlights the importance of recognizing and understanding how the distinctive dynamics of racially concordant and discordant interactions can affect medical encounters and outcomes. Physicians' racial prejudice, implicit and explicit, can have a direct negative impact on the experience of Blacks in medical encounters, and ultimately on the quality of care that Blacks receive. For example Cooper et al. found that higher implicit physician bias was associated with less positive affect and poorer ratings of interpersonal care from Black patients. Similarly, Blair et al. found that higher physician implicit bias was significantly associated with lower patient-centered care as reported by Black patients. In our own work, we have examined how the interaction between explicit and implicit bias may affect Black patients' reactions in racially discordant medical interactions. This study again uses data from participants in the primary care clinic, specifically the 15 physicians who completed the measures of explicit and implicit racial bais and the patients they saw. This work was based on Aversive Racism Theory . Aversive racism theory proposes that racial majorities often hold positive explicit, conscious thoughts about and feelings toward racial minorities that are consistent with their egalitarian values. However, at the same time they also hold negative implicit, nonconscious thoughts about and feelings toward these minorities, resulting in discomfort from the dissociation between their explicit egalitarian values and their implicit negative feelings toward racial minorities and their fear of appearing to be a racist. In order to alleviate their discomfort and fear, individuals characterized by aversive racism tend to avoid or disengage from racially discordant interactions, often without conscious awareness of the reasons for their behavior . However, people often cannot avoid certain racially discordant interactions. Once engaged in these interactions, people who fit the profile of an aversive racist may display inconsistent behaviors because they can selfregulate deliberate behaviors that are usually associated with explicit attitudes while they have difficulties self-regulating automatic behaviors that are usually associated with implicit attitudes . These inconsistent behaviors may elicit negative reactions from members of racial minority groups . Although the phenomenon of aversive racism was first demonstrated in the context of White Americans' response to Blacks, it has subsequently been found with other target minorities in countries such as Canada, the UK, and the Netherlands . The negative consequences of aversive racism may be especially relevant to racially discordant medical interactions, because for legal and ethical reasons, physicians cannot typically avoid racially discordant medical interactions. Research on Blacks' reactions to racially discordant social interactions led us to examine whether physicians who fit the profile of an aversive racist may elicit negative reactions from racial minority patients. In this study , we used the scores on the explicit and implicit racial bias measures described above to identify physicians who, relative to other physicians, showed the aversive racism profile -that is, low explicit and high implicit bias. We also computed a composite measure of patients' perceptions of the patient-centeredness of the interaction, including the items about patient satisfaction and perceived doctor warmth. We then compared differences in patients' experiences in the interactions with physicians reflecting the four configurations of explicit and implicit physician bias: high explicit/high implicit , high explicit/low implicit, low explicit/high implicit , and low explicit/low implicit . We found that physician explicit bias affected their own perceptions of being on a team with the patients , but had no impact on patient reactions. However, as can be seen in Figure 4, Black patients who interacted with the aversive racist physicians were less satisfied with the interaction and felt less positive about and close to their physicians than Black patients who interacted with other physicians, including physicians who were high on both explicit and implicit bias. We have also found, using just the patients who completed the trust measures, that patients trusted the "aversive racist" physicians significantly less than the other physicians. Turning to the issue of the role of patient attitudes, we know that patient behavior affects how physicians communicate with them; for example, contentious patients elicit contentious communication from their physician . We know of no published work on how patient race-related attitudes/beliefs may affect physician reactions to them; however a subsequent analysis of our original patient-physician data suggests there may be such a relationship. In this analysis of physician-patient interdependence in the racially discordant medical interactions, we examined the impact of patients' experience with discrimination on their physician's reactions to the interaction. We found the more past discrimination patients reported, the less likely their physicians were see the two of them as a team. To summarize, at the interpersonal level, racially discordant, relative to racially concordant, medical interactions are typically characterized by poorer communication and more negative affect. Social identity theory may provide a general theoretical framework in which to place many of these negative outcomes. Social identity processes can influence the quality of racially discordant medical interactions through several paths. People tend to engage in ingroup favoritism and/or outgroup derogation, which may explain why racial majorities are more likely than racial minorities to have positive medical interactions. Theories of how individuals self-regulate their behaviors to achieve interaction goals as well as how individuals cope with the stress associated with interracial interactions may provide further insight into why these interactions often have less than positive outcomes even in the absence of obvious intergroup bias. Contemporary research on explicit and implicit racial attitudes suggests that these attitudes affect the quality of these interactions and, indeed, our own work provides substantial support for this. Perhaps the most intriguing finding in this regard is that although we currently do not know exactly what behaviors they are reacting to, Black patients are very good at detecting implicit bias among their physicians and react negatively to it. Their dissatisfaction and distrust are directly negatively linked to the likelihood they will adhere to the treatment recommendations their physicians make and thus how much they may benefit from a medical interaction. Thus, as was true with intrapersonal processes, social psychological theories of processes that occur at the interpersonal level seem to be quite useful in understanding racial healthcare disparities. In the final section of this article we consider how these theories may be used to develop effective interventions to reduce these disparities. --- Social Psychological Approaches to the Reduction of Healthcare --- Disparities In this section we consider ways to address healthcare disparities at the intrapersonal and interpersonal levels. We acknowledge that interventions at the societal level are critical and necessary to fully address healthcare disparities. However, discussion of these societal-level interventions is beyond this review's scope. Following a key distinction in our model , this section is divided into interventions at the intrapersonal and interpersonal levels. In other words, we consider interventions based on social psychological theory and research that may effectively improve the quality of racially discordant medical interactions. At the outset, we must acknowledge the relative dearth of interventions that specifically target racially discordant medical interactions. We shall present these and where appropriate also discuss possible interventions that draw on social psychological theory and research. We discuss two general kinds of interventions. One kind is interventions that target negative stereotypes and attitudes towards all members of some racial minority. The goals of such interventions we describe are to produce general changes in physicians and patients that might affect healthcare disparities. The other kind is interventions that target a specific racially discordant interaction. Their goal is to improve the outcome of these interactions by reducing the impact of negative race-related thoughts and feelings of the specific patient and/or physician participating in the interaction. One might hope that these effects might generalize beyond this interaction, but our primary goal is much more modest. We simply want to reduce healthcare disparities in that interaction. --- Intrapersonal Level Processes In this section we focus on potential interventions that target the race-related thoughts and feelings patients and physicians may bring to racially discordant medical interactions. First we consider the racial minority patients. Racial minority patients-Interventions with racial minority patients have focused on both cognitive and motivational processes. Specifically, these interventions have addressed patients' expectations of bias and the health consequences of identity-based motivations. Members of both racial majority and minority groups approach interracial interactions guardedly and with negative expectations . When racial minority members expect to be treated unfairly by a majority group member in racially discordant interactions, they experience negative moods and/or interaction anxiety and may become defensive during the interactions . Therefore, one target for interventions is to reduce the impact of these expectations on racially discordant medical interactions. As noted earlier, Burgess and colleagues proposed that one manifestation of racial minority patients expecting to be the target of prejudice and discrimination is the activation of stereotype threat, wherein minority group patients may worry that their physicians may negatively stereotype them. Thus, minority patients may not fully and actively participate in the interaction, lest they engage in behaviors that might reinforce these stereotypes. Recently, Havranek et al. tested the impact of an intervention intended to reduce stereotype threat among minority patients on the patients' reactions to racially discordant medical interactions. In this study, patients were randomly assigned to either a treatment group, in which they engaged in a values-affirmation exercise shortly before they met with their physician, or to a control group. Values-affirmation exercises encourage people to identify personal values that are important to them. Such exercises have been shown to reduce stereotype threat, presumably due to the positive impact on the self . The interactions were audio recorded. Patients who engaged in the values affirmation exercise asked for more information about their medical condition during the interactions and reported a less negative mood after the interactions, as compared to patients in the control group. Additionally, independent coders rated these patients as more friendly and responsive during the interactions. The goals of medical interactions extend beyond accurate diagnosis and the best treatments. Today, most medical facilities also emphasize steps that people can take to prevent diseases. Indeed, there is already a very large public health and communication literature on persuasive messages that target and attempt to change health-related beliefs and behavior . Recently, communication researchers have proposed that health messages should be tailored based on the target's social identity . This approach is based primarily on classic models of persuasive communication . We would suggest another approach based on Oyserman and colleagues' report of their efforts to change identity-based motivations in middle school students. Oyersman, Bybee, and Terry targeted racial minority students' self-concepts, social identity, and academic achievement. They developed classroom exercises aimed at connecting students' "possible selves" and their social identity. Specifically, the classroom exercises encouraged minority students to see their "possible selves" as congruent with a social identity that was associated with academic success. The primary means to do this was to have students think about another student and/or adult with whom they identified, who provided a model of their academically successful possible self. This intervention was successful and had substantial long-term academic consequences for the students. For example, relative to comparable students who did not receive the intervention, the targeted students had fewer school absences and higher grades and scores on standardized tests. Moreover, these effects persisted over time. It would seem useful to explore whether similar interventions that target self-concept and social identity might be developed for health-related behaviors among racial minorities. Physicians-At both explicit and implicit levels, non-minority physicians tend to harbor negative feelings towards members of racial minorities and hold negative stereotypes about minority patients. Thus, interventions that target these attitudes and beliefs might serve to reduce healthcare disparities. These interventions target people's general stereotypes and attitudes that affect reactions to some target group, including their thoughts, feelings, and actions when they interact with members of that group. Burgess and colleagues argued that changes in physician training will reduce the impact of ingroup bias and racerelated stereotypes and attitudes on healthcare. Most medical schools now require training in "cultural competency"-"a set of attitudes, skills, behaviors and policies enabling individuals to establish effective interpersonal and working relationships that supersede cultural differences" . The term cultural competency training covers a host of different approaches, but in the training models that have been empirically evaluated, do not directly address intergroup relations or racial bias . Burgess et al. recommended teaching medical trainees about the nature of stereotyping, thus increasing their understanding that even negative stereotypes are a pervasive social phenomenon that can unintentionally shape medical decision making and interactions. This type of training should help them to devise strategies to counteract stereotypes rather than try to deny or suppress them. At a conceptual level, we would agree that such an approach has value. In fact, this approach is consistent with findings from the social psychological literature showing that, with some practice, people can develop strategies to counteract the influence of implicit racial bias on their explicit thoughts, feelings, and actions . However, given the other demands on medical education to prepare trainees for the current rapidly-changing healthcare environment, these kinds of programs will need to be integrated into already crowded medical curricula, which may make them less feasible or effective. Although it may be difficult to directly address medical trainees' race-related thoughts and feelings through special training, it might be possible to modify the training experience in ways that reduce racial stereotypes and the anxiety and uncertainty that often accompanies interracial interactions. Allport's classic contact hypothesis suggests some strategies. Repeated positive and constructive contact between ingroup and outgroup members can often reduce implicit and explicit negative attitudes and beliefs about members of an outgroup, as well as negative affective reactions that often accompany interacting with members of the outgroup; and thus improve intergroup relations . So, one way to reduce physicians' negative race-related feelings during racially discordant medical interactions is to provide them with multiple opportunities to interact with minority patients during their training. This could be done in the context of a widely used training tool in medical schools--interactions with "standardized patients" . That is, one possible way to give medical students experience with minority patients is simply to increase the diversity of the standardized patients they see, and provide feedback that specifically addresses those aspects of the interaction that concern intergroup relations. It might also be useful to provide students with behavioral scripts before these interactions. This approach is suggested by a study by Avery, Richeson, Hebl, and Ambady that found providing Whites with behavioral scripts to follow before their interactions with Blacks significantly reduced their anxiety during the interactions. This might serve to desensitize non-minority physicians to minority patients early in their medical training. It would also be beneficial to provide non-minority medical students with naturally occurring contact with equal or higher status individuals from racial minorities, such as minority medical students or medical school faculty. However, only a small portion of medical students are from racial minorities, and minorities are also substantially under represented on medical school faculties . One possible reason is for this is that, at least in the US, minority medical school faculty members have significantly lower promotion rates than White faculty . This limits the opportunities for White students and faculty to have contact with equal or higher status minority group members. Other interventions might attempt to reduce physicians' reliance on racial stereotypes when they interact with racial minority patients. For example, one might attempt to reduce the cognitive load of physicians whose practices involve a high percentage of racial minority patients. Laboratory studies of racially discordant interactions suggest that Whites under low cognitive load are less likely to engage in social categorization and more able to individuate their interaction partner than those under high cognitive load . Thus, interventions might be developed to increase physicians' cognitive resources by reducing the stress and anxiety associated with their clinical practice. This might involve organizational interventions, such as reducing patient caseloads through the use of nurses or other physician "extenders." Of course, given the fiscal realities of medical practice in the 21 st century in almost all countries, it seems that such changes in organizational practice will not be easy to achieve. --- Interpersonal Level Processes We have argued that one major reason for the less productive and positive communication in racially discordant medical interactions involves the social biases that physicians bring, often without conscious awareness, to these formal encounters. Basic social psychological research and theory can help guide the development of interpersonal-level interventions to reduce healthcare disparities. In this section, we consider strategies for structuring racially discordant medical interactions and reshaping social categorization processes to enhance the effectiveness of racially discordant medical interactions. Creating equal status in racially discordant medical interactions-A metaanalysis conducted by Pettigrew and Tropp testing Contact Theory reveals that intergroup contact, and particularly under conditions of equal status, common goals, acquaintance potential, and support of authorities, has beneficial effects on intergroup attitudes. Within medical contexts, the Patient-Centered Communication Model of physician communication , which is widely taught in North American and European medical schools , may serve to promote these facilitating conditions, especially equal-status relations. Whereas physician-patient interactions have traditionally reflected marked status distinctions, the Patient-Centered Communication Model stresses the need for physicians to understand the patient's perspective, consider the patient's psychological and social context, reach a shared understanding of the patient's problem and its treatment, and offer patients meaningful involvement in treatment decisions . A central goal of patientcentered physician communication is to empower patients to participate actively in medical interactions. Research shows that patient-centered medical interactions produce more patient involvement and more positive outcomes . However, as we have already noted, the effects of patient-centered communication have not been systematically examined in the context of racially discordant medical interactions. Moreover, our own research suggests that greater patient participation may not always reflect physicians empowering patients. Recall that Hagiwara et al., found that Black patients who perceived greater past discrimination and reported less trust talked more in their medical interactions with non-Black physicians. This suggests that this behavior may have reflected negative rather than positive affect. As a consequence, in these raciallydiscordant interactions, greater talk-time by patients predicted less subsequent adherence to the doctor's recommendations. Thus, while patient-centered communication may have value in improving the outcomes of racially discordant medical interactions, it needs to be empirically evaluated before being implemented as a remedy for the problems associated with these interactions. One intriguing, but as yet untested, idea for an intervention to create optimal conditions for racially discordant medical interactions is to try interventions modeled on Aronson's Jigsaw Technique . The Jigsaw Technique was designed to reduce intergroup conflict in small groups and emphasizes the importance of each individual's participation and of the information each has in the solution of a problem of mutual interest. This technique can create equal status among the participants. Although this technique was originally developed to improve diverse classroom environments, it might be meaningfully modified for medical interactions. For example, framing the medical interaction as one in which each party has valuable information that the other one needs to solve the "problem" might engender a more equal status, cooperative interaction than is the norm in racially discordant interactions. It would seem worthwhile to explore this approach to improving racially discordant medical interactions. The optimal conditions identified by Contact Theory and the dynamics of the Jigsaw Technique operate to reduce intergroup bias by altering the ways people in these interactions socially categorize others 2 . As already discussed, because of the importance of race and ethnicity in social relations generally, people automatically and often unconsciously respond to others primarily based on their race or ethnicity, which can negatively affect racially discordant medical interactions. Thus, one possible avenue to prevent such outcomes in interactions is to reduce the likelihood that participants automatically categorize themselves as members of two different groups. Specifically, positive intergroup contact and cooperative interdependence, as identified by Contact Theory and the Jigsaw Technique, can alter physician and patient perceptions of membership in an ingroup and in an outgroup based on race/ethnicity to more individuated perceptions of one another and/or to recognition of a shared identity within the context of a specific medical interaction. To the extent that social categorization processes are a critical factor biasing perceptions and interpersonal-level processes in racially discordant medical encounters, interventions targeted more directly at changing the nature of social categorization can help eliminate healthcare disparities. In the remainder of this section, we discuss interventions designed to create more individuated impressions or common-group representations in medical encounters. --- Individuation of participants in racially discordant medical interactions- Several models of intergroup relations posit that people do not categorize others into specific social categories when they are provided with rich individuating information about target individuals. For instance, Turner, Hewstone, and Voci have shown that self-disclosure among members of different groups is associated with an increase in positive attitudes and more heterogeneous perceptions of members of the outgroup. Consistent with these research findings, Burgess and colleagues specifically argued that individuation of a minority patient is an effective way to reduce the impact of race-related thoughts/feelings on physician reactions to that patient. We are currently testing an intervention that is partially intended to help individuate racial minority cancer patients in the eyes of their oncologists. Prior to meeting the oncologist, minority patients are provided with a list of questions they might ask their non-minority oncologists during discussions of their treatment and given advice on the best way to ask them. We believe that patients asking more questions about their specific condition and treatment will activate more individuated impressions of minority patients during the interactions. In addition, asking these questions provides an example of an engaged, interested patient that contrasts with documented physician expectations that minority patients are unintelligent, uninterested or uncooperative . The patient may be seen as an exception to the social category, thus diminishing the impact of pre-existing stereotypes on the oncologists' views of the patient. As a result, treatment decisions will be more likely to be based on the particular characteristics of this patient than stereotypes about the group to which he/she belongs. 2 We thank the editor for suggesting the potential usefulness of the Jigsaw Technique in interventions that address healthcare disparities. In the same vein, previous research reveals that minority cancer patients typically receive less information from their oncologist , which may reinforce minority patients' stereotypes about nonminority physicians as not interested in or biased against them . Thus, we posit that this intervention will have cascading positive influences on racially discordant medical interactions. Helping patients to ask more questions is expected to elicit greater responsiveness from oncologists, which, in turn, should undermine patients' negative stereotypic perceptions, at least with respect to this particular healthcare provider. That is, greater responsiveness in the encounter by oncologists may serve to individuate the oncologists to the patients and thus reduce the impact of the patients' race-related thoughts and feelings on the interaction itself and the treatment decisions patients make. Note that this intervention would not really address the general attitudes and stereotypes that either oncologists or patients bring to the interactions, but it might reduce the relevance of these feelings and thoughts to communication during their immediate interactions and the decisions each party makes. Of course, one could simply provide physicians and patients with large amounts of information about one another prior to the interactions, which might also serve to individuate them. However, promoting individuation by providing patients and physicians detailed information about one another may not be practical. Physicians may not be willing or able to invest great amounts of their very limited time exchanging individuating information with their patients; similarly, patients may not be willing or able to learn about their physicians as individuals. Furthermore, the stressful aspects of racially discordant interactions and the time pressures and heavy workload for many physicians who treat minority patients may increase the cognitive load on such physicians. Under these conditions attempts at individuation via increased information may actually have counterproductive effects. For example, Phillips, Northcraft and Neale studied homogeneous and diverse groups working on a complex problem. They found that individuating members of diverse groups with information about similar attitudes and values did not increase attraction among them and, in fact, actually resulted in poorer performance relative to diverse groups that did not receive this information3 . These findings suggest that one might be cautious in providing large amounts of individuating information during already cognitively taxing racially discordant medical interactions because this information may further increase the cognitive load. Under such circumstances, physicians may actually fall back onto racial stereotypes to reduce the cognitive load. Creating a common identity in medical interactions-Another, possibly more productive, way to improve the outcomes of racially discordant medical interactions might be to try to change patients' and physicians' perceptions of one another as "us" versus "them." Social psychological research has shown that it possible to do this by creating a common group identity, in which members of disparate groups are induced to focus on their common interests and goals they share in common . The creation of a common group identity has been found to reduce intergroup bias and increase intergroup trust. In a recent study , we attempted to improve the outcomes of racially discordant medical interactions by creating an intervention to promote common group identity between the low SES Black patients and their non-Black physicians in our samples in the Detroit clinic. Prior to their appointment with their primary care physicians, 72 Black patients completed all the baseline measures described above and then were randomly assigned to a treatment group or control group. The 42 patients in the treatment group received written instructions that introduced the concept of being on a team with their doctor. They were asked to sign a contract stating that they would be "partners in a team that is working to solve your medical problems." To reinforce this message, treatment-group patients were given buttons and pens with the team name and team colors. Patients also received 10 suggestions for being part of a team, which were posted on the wall of the room in which the treatment-condition patients saw their physician. The 30 patients in the control group received their usual standard of care. The 7 physicians randomly assigned to the treatment condition received similar instructions at the beginning of the study and wore the same kind of buttons during the interactions. Each physician interacted with multiple patients across the course of the study. Patients completed the measure of physician specific trust and the two team items immediately after the interaction; 4 and 16 weeks later they completed the trust measures again and the adherence measures. In all the tests of treatment-control differences, we controlled for baseline trust and adherence measures. As illustrated in Figure 5, we found that relative to the control group, patients who received the common identity manipulation showed increased trust in their physicians and in physicians in general at four and 16 weeks after the interactions. Perhaps more importantly, 16 weeks after the interactions, members of the intervention group adhered to physician recommendations significantly more than did members of the control group. These findings reveal the important role of changing the way physicians and patients think of each other during their medical interactions -from "us" and "them" based on their racial or ethnic group memberships to "we" in terms of their relationship in this medical context -on healthcare success. Understanding the psychological mechanisms that underlie intergroup bias can thus help guide the development of interventions to improve the effectiveness of medical encounters in ways that particularly benefit minority-group patients, thereby reducing healthcare disparities. As we close this section, we might consider which of the two kinds of interventions holds the greatest promise for reducing healthcare disparities. Ultimately, of course, it is those interventions that target general negative race-related thoughts and feelings. However, for the immediate future it may well be most practical and efficacious to target specific medical interactions and focus attention on improving interactions between racial minority patients and their physicians. Thus, the most practical and realistic goal is to reduce the impact of race-related thoughts and feelings on how patients and physicians interact with one another and, thus improve the outcome of these interactions. --- Summary and conclusions Around the world, members of racial/ethnic minorities almost invariably experience poorer mental and physical health than do racial/ethnic majorities. Racial/ethnic disparities in healthcare play an important role in these disparities in the health of different racial/ethnic groups. Societal, intrapersonal, and interpersonal level processes involving race-related thoughts, feelings, and actions play important roles in healthcare disparities. We presented a social psychological analysis in the exposition of the healthcare disparities model. To do this, we focused more on processes at intrapersonal and interpersonal levels than on the societal level, and primarily considered the impact of these processes on racially discordant medical interactions. At both levels, we saw that a social psychological analysis provides considerable insight into the reasons why these interactions are typically less positive and productive than racially concordant interactions. While healthcare disparities have a long history and are pervasive throughout the world, they are not inevitable and can be reduced, if not eliminated. Social psychological theories have much to offer in this regard. Some healthcare disparity interventions, directly derived from social psychological theories, have already been tested and seem to hold great promise of improving both communication during and the outcomes of racially discordant medical interactions. Moreover, there are other untested interventions, also derived from social psychological theory that hold promise as means to address healthcare disparities. We believe there is a special benefit to developing such theory-driven interventions. Presently, many of the interventions directed at healthcare disparities are largely atheoretical and may often simply apply an intervention that was shown to be effective in one context to another context, without much consideration of why it was successful. One important practical drawback of such approaches is that if the intervention succeeds , it is very difficult to identify the mechanisms responsible for the outcomes. Thus, the investigator has little guidance on how to improve the intervention or apply it to other settings. However, theory-driven interventions which specify and test the presumed causes of the expected outcome and the mediational processes associated with the outcomes, are likely to provide valuable heuristic and practical information for future efforts, whether or not the specific intervention is successful. We note that there are two excellent reasons for bridging social psychological theory and racial and ethnic disparities in health and healthcare. First, this pervasive social problem provides an exciting natural laboratory in which to test theories of intergroup relations and racial bias. At the same time, psychologists have a potentially important role to play in applied, practical research intended to reduce health disparities. Thus, this article represents an argument for a melding of basic and applied research. Thus, addressing the problem of healthcare disparities may be another example of Kurt Lewin's well-known axiom, "Nothing is as practical as a good theory." A multi-level model of the causes of healthcare and health status disparities Copyright © 2013 Elsevier. Racial attitudes, social control, and adherence in racially discordant medical interactions. Hagiwara,N. et al.,Social Science and Medicine,87,[123][124][125][126][127][128][129][130][131] Figure 3 Caption Overall physicians talked more tan patients ; the larger this ratio, the more physicians talked relative to patients.

Around the world, members of racial/ethnic minority groups typically experience poorer health than members of racial/ethnic majority groups. The core premise of this article is that thoughts, feelings, and behaviors related to race and ethnicity play a critical role in healthcare disparities. Social psychological theories of the origins and consequences of these thoughts, feelings, and behaviors offer critical insights into the processes responsible for these disparities and suggest interventions to address them. We present a multilevel model that explains how societal, intrapersonal, and interpersonal factors can influence ethnic/racial health disparities. We focus our literature review, including our own research, and conceptual analysis at the intrapersonal (the race-related thoughts and feelings of minority patients and non-minority physicians) and interpersonal levels (intergroup processes that affect medical interactions between minority patients and non-minority physicians). At both levels of analysis, we use theories of social categorization, social identity, contemporary forms of racial bias, stereotype activation, stigma, and other social psychological processes to identify and understand potential causes and processes of health and healthcare disparities. In the final section, we identify theory-based interventions that might reduce ethnic/racial disparities in health and healthcare.ethnicity/race; health and healthcare; disparities; explicit and implicit bias; intergroup processes Across a broad spectrum of mental and physical illnesses, members of socially disadvantaged groups typically experience poorer physical and mental health than members of socially advantaged groups. This is a persistent social problem in at least 126 countries, which include 94.4% of the world's population (Dorling, Mitchell, & Pearce, 2007). There are two general classes of explanations of this pervasive public health problem. The first is that these differences reflect biological and genetic differences in the populations to which

Introduction Intimate relationships are linked with better mental and physical well-being across the lifespan [1][2][3]. For people with chronic illness or disease in particular, intimate relationships can greatly impact their ability to self-manage their condition, illness-related quality of life, coping responses, post-traumatic stress, and growth, among many other factors [4][5][6][7]. The downstream impact of intimate relationship quality can thus be substantial, especially for partners struggling with the day-to-day manifestations of their condition or disease [8]. When the condition or disease specifically impacts sexuality, intimate partner support is especially critical. Sexual intimacy and sexual satisfaction are crucial components of relationship quality [9] and are directly linked to relationship stability [10]. Longitudinal studies of married couples have shown that sexual satisfaction at time of assessment is related to later relationship satisfaction, even in studies spanning more than a decade [11]. In fact, a recent study employing machine learning to the study of over 11,000 romantic couples found that sexual satisfaction was one of the top five predictors of higher relationship quality [12]. Although scholars are in disagreement over the true direction of the association between relationship satisfaction and sexual satisfaction-meaning which distinctly drives the other-these concepts have consistently been highly correlated in cross-sectional and longitudinal research [13]. Unsurprisingly, changes to sexual intimacy are among the most predominant concerns for people undergoing treatment for breast, colon, or reproductive cancers [5,14,15]. Such changes could leave one or both partners with unfulfilled sexual needs and potentially pose a threat to relationship longevity. In the current study, we examined sexual satisfaction and distress in the context of chronic vulvovaginal pain . CVVP is vulvar and/or vaginal pain that is unexplained by underlying physical trauma, medications, or any known medical conditions [16,17]. This pain can occur during day-to-day mundane activities, such as sitting for long periods of time or wearing tight clothing, that can cause prolonged vulvar pressure. CVVP can also occur during menstruation with the use of menstrual products, during gynecological pelvic exams, and during sexual activity both with and without vaginal penetration [18,19]. Although prevalence estimates are based on incomplete data, around 13 million people are likely experiencing CVVP in the United States at the time of this writing [16]. Research on CVVP has shown that the pursuit of diagnosis and treatment is an arduous journey. Many participants report that healthcare providers have insisted that their pain is psychosomatic and thus not "real" [20][21][22]. Other studies have shown that people with CVVP typically have to see multiple healthcare providers before they reach one who has the expertise needed [23]. Prolonged health-related uncertainty and navigating multiple healthcare environments is likely to produce high levels of stress for people with CVVP; as with any other chronic condition, social support is essential [24]. Social support is defined as "information leading one to believe that he/she is cared for and loved, esteemed, and a member of a network of mutual obligations" . Social support generally includes the exchange of encouragement, validation, and resources [26] and is intended as a catalyst, enhancer, or buffer [27]. In particular, emotional support is defined as providing care, love, empathy, and support [26,28,29]. The receipt of emotional support has been shown to buffer stress, particularly in situations beyond a person's control [30]. Likewise, the receipt of emotional support reduced the odds of post-traumatic stress disorder symptom emergence in people who had experienced a physical assault [31]. In the context of chronic diseases, including cancer, diabetes, and lupus, emotional support contributes substantially to better psychological adjustment and subsequently to better overall well-being [32][33][34]. Furthermore, for women experiencing chronic illness, emotional support received from a romantic partner has been associated with better adjustment to the illness, less occurrence of depressive symptoms, and higher levels of marital satisfaction [24]. Unlike many other chronic conditions, however, CVVP impacts not only psychological and physical well-being but romantic relationship functioning. While intimate partners are typically among the main sources of social support for adults [35], people may be reluctant to seek support from their partners when it relates to CVVP. This is because the presence of CVVP can complicate the maintenance of sexual intimacy. As previously detailed, sexual intimacy and resulting sexual satisfaction are especially important for perceptions of relationship quality but CVVP may prevent typical engagement in sexual behaviors [36]. This may lead some people who have a CVVP condition to choose not to disclose their condition to their partner, in order to avoid their partner viewing them differently. Studies have shown that women with CVVP feel shame because their pain interferes with their ability to sexually please their partner-something that many women feel makes them incapable of being the "perfect girlfriend" or worthy romantic partner [37]. Thus, women and others may hesitate in seeking support for their CVVP from intimate partners, because doing so would communicate their ongoing discomfort with sexual activity and highlight their perceived shortcomings, which could in turn harm the relationship. In the current descriptive study, we examined the prevalence of disclosure to romantic partners in a sample of over 300 women experiencing CVVP. For women who had disclosed their condition to their romantic partners, we further investigated the extent to which they perceived their partners as providing relevant emotional support and how that support was associated with sexual dissatisfaction and distress. Longitudinal studies involving control groups of women without CVVP are needed to provide the most rigorous and causally informative understanding of how CVVP impacts romantic and sexual intimacy and overall well-being; however, the literature in this area is still emerging, and the existing library is small. Here, we conducted a descriptive and correlational investigation to provide needed foundational information for future hypothesis-driven work [38][39][40]. A better understanding of health-related disclosure and partner support in the context of chronic pain conditions like CVVP can help to advance research and therapeutic intervention on physical, psychological, and social well-being. Although there is a small but active literature on the experiences of women with CVVP, the great majority of work in this area has been conducted with samples mostly comprising White women. A clear limitation of the existing knowledge base is in the experiences of people of color who have CVVP conditions. Studies examining lifetime prevalence of chronic vulvar pain in racially diverse communities have found no difference in lifetime rates for Black versus White women, but findings have suggested that Hispanic women are between 1.4 and 1.8 times more likely than women of other races/ethnicities to experience chronic vulvar pain [18,23]. Rates of lifetime prevalence may be relatively stable across racial groups, but studies have indicated that women of color have unique struggles with CVVP. One study examining racial differences in the experience of provoked vestibulodynia-pain localized at the vulvar vestibule [41] found no difference in Black versus White women's pain ratings, but Black women had higher rates of functional impairment caused by their vestibulodynia [42]. The disparity in impairment is especially disconcerting considering the nearly twenty percent difference in diagnostic rates for Black versus White women with chronic vulvar pain [18]. We know of no research on CVVP in the context of romantic relationships that focused or specifically mentioned people who identify as a race or ethnicity other than White. However, work conducted with patients navigating other chronic diseases has shown the significance of sexual intimacy in the lives of women of color. For example, in a qualitative study of Black women breast cancer survivors under the age of 45, Lewis and colleagues reported that half of the sample believed that their cancer negatively affected their relationships and particularly their partner's emotions [43]. In this study, participants identified sexual side effects as a paramount issue in their cancer prognosis. Relatedly, another study found similarly negative romantic and sexual relationship outcomes for Black, Asian, and White breast cancer survivors, but Latina survivors experienced additional challenges due to cultural conceptions of femininity and reconciling those conceptions with the mastectomies they had received [44]. Though sourced only from a very small review of the literature, these findings make clear the likely impact of sexually-or reproductively relevant health conditions on sexual intimacy for women of color. Outside of the illness context, recent work has further demonstrated the importance of sexual intimacy on these women's romantic relationships. Townes and colleagues conducted a study of Black women's perceptions of sex satisfaction within the context of relationships and found that nearly 80% of the sample considered aspects of sexual intimacy to be essential for relationship functioning [45]. In the current study, we expand the existing literature, focusing specifically on the experiences of women of color in our investigation of CVVP, partner support, sexual satisfaction, and distress. The current study focused on social support, specifically aspects of emotional support, and its impact on individuals with chronic vulvovaginal pain. This study also specifically examined people of color-predominantly Black women-which was unique from previous studies on this topic that either lacked participants of color or failed to report on racial/ethnic categories at all [46][47][48][49]. In addition, this study sought only to examine the proposed associations for individuals who were partnered in long-term, committed romantic relationships. Based on previous literature, to analyze the current sample, our research questions were: how emotionally supportive are the intimate partners of the participants who have disclosed their chronic vulvovaginal pain?; is more support from partners associated with lower sexual dissatisfaction and/or distress?; and aside from emotional support from one's partner, what sources are the most helpful in providing emotional support to participants who experience chronic vulvovaginal pain? --- Materials and Methods --- --- Measures Demographics. Demographic questions were asked to obtain information about the participants' age, gender identification, relationship status, and sexual orientation. Partner support. Emotional supportiveness from the participant's partner was measured with the item "How supportive was your partner's response when you told them that you have chronic vulvovaginal pain?", with choices ranging from very unsupportive to very supportive . Dissatisfaction about sex life. Dissatisfaction was measured with the item "How often do you feel dissatisfied with your sex life?" from the Female Sexual Distress Scale-Revised [50], and participants answered on a 5-point Likert type scale ranging from "never" to "always" . Distress about sex life. Stress was measured with the item "How often do you feel distressed about your sex life?" from the Female Sexual Distress Scale-Revised [50], and participants answered on a 5-point Likert type scale ranging from "never" to "always" . Emotional support sources. Participants were asked "What information source is the most helpful for emotional support?" and were allowed to select up to three choices from the following options: medical professional, general medical information website , vulvar/vaginal pain-specific medical information website , health-related forums and social media , friends or family members, newspapers or magazines, government health agencies , radio or podcasts, religious organizations or leaders, lifestyle websites or blogs , or none of the options. They were then asked, "What information source is the least helpful for emotional support?" and were allowed to choose up to three choices from the same list provided above. --- Procedure Participants were recruited via the Internet. Standardized messages, including a recruitment flyer, were posted on various social media sites and were subsequently shared widely by other social media users. All research procedures were approved by the university's Institutional Review Board; furthermore, this study was deemed exempt, as it presented no more than minimal risk to subjects and followed the guidelines for ethical research presented in the Belmont Report [51]. Recruitment for this study included several eligibility requirements: age of 19 to 55, no experience of perimenopause or menopause, identifying as a person of color , and having experienced vulvar, vaginal, and/or pelvic pain for at least three to six months. Participants who failed the survey attention check items or the reCAPTCHA item and participants with duplicate IP addresses were also removed from the dataset to ensure data quality. In addition, this report examined only the data of eligible participants who reported being in a form of long-term, committed relationship. After data cleaning, 333 participants were included for statistical analysis. Eligible participants reported demographics and whether they had been formally diagnosed with a chronic vulvovaginal or pelvic pain condition . Participants who did not have a formal diagnosis reported whether they suspected they had a chronic vulvovaginal or pelvic pain condition to capture individuals who had not yet sought care or who had not yet found an appropriate healthcare specialist that could provide a diagnosis. Participants also responded to ad hoc survey items regarding their information-seeking behaviors and their experiences in deciding to disclose their chronic pain to relationship partners. These data were part of a larger study on the multifaceted experiences of people of color living with chronic vulvovaginal and/or pelvic pain. The larger study included data on a host of health and behavioral aspects of CVVP unrelated to the current study. Here we discuss only variables related to participant demographics, partnered social support, sexual dissatisfaction, distress, and other sources of support identified by participants as helpful to their experiences with CVVP. --- Data Analysis SPSS version 28 was used to analyze the data. Statistical significance was determined by p-values less than 0.05. Descriptive statistics were run in order to answer the research questions posed in the study to determine percentages of participants. Two separate linear regression models were used to determine if partner supportiveness was associated with less sexual distress and/or less sexual dissatisfaction. --- Results To answer our proposed research questions, we examined descriptive statistics to determine how supportive intimate partners were after their partner disclosed their chronic vulvovaginal pain. Our results showed that partners were rated as largely supportive. On a seven-point scale, only 2.5% of the sample selected the midpoint or lower, corresponding with "neither supportive or unsupportive" or worse. Furthermore, no one reported that their partner was "very unsupportive". One in four participants reported that their partner was very supportive; 38.6% reported that their partner was supportive ; and 34.4% reported that their partner was somewhat supportive. Next, a linear regression was run to examine the relationship between partner supportiveness and dissatisfaction about the participant's sex life. The results yielded a significant negative relationship between partner supportiveness and dissatisfaction , indicating that increased partner support was associated with lower levels of dissatisfaction. Therefore, H1 was supported by the data. Next, a linear regression was run to examine the relationship between partner supportiveness and distress about the participant's sex life. The results yielded a significant negative relationship between partner supportiveness and distress , indicating that increased partner support was associated with lower levels of distress. Therefore, H2 was supported by the data. Lastly, descriptive statistics were run in order to determine what sources of emotional support participants found to be most and least helpful as related to their chronic vulvovaginal pain. The most helpful emotional support sources were medical professionals , followed by health-related forums and social media , family members or friends , and vulvar/vaginal pain-specific medical information websites . The least helpful emotional support sources were none of the options provided , followed by vulvar/vaginal pain-specific medical information websites and medical professionals . All of the values for each source for both least and most helpful are reported in Tables 1 and2. As participants were allowed to select up to three of the choices provided, percentages did add up to over 100%. --- Discussion In the current study, we investigated the presence of social support from romantic partners for participants experiencing chronic vulvovaginal pain . We also examined associations between that partnered support, sexual dissatisfaction, and distress in the context of this type of pain. Our results provided evidence-albeit of the cross-sectional variety-for the positive role that social support plays for individuals who experience CVVP. Although the role of a romantic partner's social support has been explored in several other contexts, this was the first study to examine its role in individuals who experience CVVP. The current study should help provide encouragement to individuals who experience CVVP to disclose their diagnosis, as very few participants received unsupportive responses from their partners, and most experienced very positive, supportive responses. Additionally, higher levels of partner supportiveness were associated with lower sexual dissatisfaction for the participant as well as less distress surrounding their sex life. Because of the intimate relationship that participants shared with their partners, and the difficulties CVVP may cause in the experience of sexual intimacy, it is reasonable to believe that having one's partner being supportive would be of the utmost importance in how the participant views their sex life and sexual experiences. This finding also suggests that additional research should aim to examine disclosure strategies for individuals with CVVP, which could assist in finding the best way to discuss this sensitive topic with sexual partners. Within future studies, researchers should examine how these relationships between emotional support and sexual dissatisfaction play out for individuals who are single or in the early stages of a relationship, as singles may be more hesitant to reveal their diagnoses to sexual partners who are not consistent or new because of the lack of trust, which takes time to establish. Beyond the role of partner support in CVVP, we also endeavored to identify the most helpful sources of emotional support for people experiencing CVVP. Two choices emerged when examining what sources of emotional support participants found the most and least helpful. These choices, for both the most and least helpful respectively, were medical professionals and vulvar/vaginal pain-specific medical information websites. This suggests that medical professionals play a key role in not only providing their patients with medical information and diagnoses but being people that patients can turn to for emotional support, specifically for individuals who experience CVVP and may not feel that others fully understand what they are experiencing. Imploring medical professionals to listen to their patients and provide patient-centered care, particularly for patients who are experiencing CVVP, may assist with their patients' overall sexual and mental well-being. Previous research has shown that patients with chronic pain have better outcomes when their medical providers are able to use person-centered messages and develop a supportive connection with their patients [52]. In addition to medical providers, vulvar/vaginal pain-specific medical information websites were also identified as a most and least helpful source of support for many of the participants. There may be more nuanced reasons as to why some individuals with CVVP find these websites more or less helpful, such as a lack of specific information that individuals are searching for or a lack of certain desired features. Further research is needed to analyze the content of these websites in order to identify what can be improved in order to assist individuals with CVVP. --- Limitations Though we have expanded the literature on the understanding of social support and its impact on racial and ethnic minority people with chronic vulvovaginal pain , the current research was not without limitations. First, we relied exclusively on self-report data collected from the internet. Though it would be difficult to collect perceptions of support in the context of CVVP beyond the self-report, future research could employ more innovative research designs to ensure consistency in findings. Furthermore, online studies have their own limitations, including sample restriction due to accessibility of the internet and associated technology . Online surveys also remove the researcher's ability to control the environment in which people answer survey questions. The survey environment has the potential to affect how willing people are to be truthful in their responses to sensitive questions. For example, people may spend less time on-and subsequently, provide lower quality data for-questions related to sexuality if they are in the presence of other people who may be able to see the survey items. Second, our sample for this particular inquiry was restricted. The overall sample from the larger study included a wider array of participants , but our specific research inquiry necessitated a focus on people in established romantic relationships. However, people with CVVP who are single or not in stable romantic relationships need to be included in future research. Work with these populations would provide insight into how their sexual and romantic desires are impacted by pain, how they pursue partners, and where they derive support from with regards to their CVVP . Those transitioning into a romantic relationship would also provide valuable insight into how couples navigate disclosure of their condition and subsequent support dynamics. Additionally, our participants all resided in the United States, but we did not have data on where our participants were raised . This limited our knowledge of potential cultural differences. We could not ascertain the extent to which partners were willing to discuss gynecological issues with their partners-including vaginal pain-or if their cultural expectations drove their likelihood to share, or withhold, this type of information from their partners. Future work focusing on a larger population should collect information on social-environmental factors that may impact the magnitude of observable effects. We had intended to compare participants who had been formally diagnosed with a CVVP condition with those who had not received a formal diagnosis. This type of comparison would potentially uncover differences in sexual dissatisfaction, distress, or partner support based on whether a healthcare authority had declared one's symptoms to be reflective of an "authentic" condition. Unfortunately, only 16% of the sample reported not having a formal diagnosis; 68% of the sample did report a formal diagnosis, and an additional 16% did not answer this survey question. The distinct imbalance between diagnosed and undiagnosed groups made comparisons less informative than ideal. We recommend that future researchers work to collect adequate samples of both groups to be able to examine indirect or unexpected benefits associated with receiving a formal CVVP diagnosis. We had little variance to examine with regard to differences between participants who had and had not disclosed their pain condition to their romantic partner. The great majority of our participants had disclosed their CVVP to their romantic partner. This finding was in line with prior work on the role of romantic partners in the experience of endometriosis. Although this work found a largely negative impact of endometriosis on the romantic relationship, over 90% of women in the study reported that their partners were interested in being informed of their health and the status of their condition [53], suggesting that disclosure would likely be appreciated by most partners. However, it should be noted that good communication skills are important for navigating long-term impacts of any chronic condition. An investigation of Canadian and American women with provoked vestibulodynia-a type of chronic vulvar pain-found no consistent pattern in women's Dyadic Sexual Communication scale scores [54]. Thus, having a condition that directly impacts one's partnered sexuality does not inherently make one better at working with a partner toward constructive solutions. Future research on partnered communication in the context of CVVP would be useful for therapeutic interventions. Future researchers may also wish to delve into the present topic with more specificity, examining particular facets of distress or satisfaction in the realm of sexual intimacy. In particular, studies focused on partner communication in relation to feelings of shame or of guilt around not being able to "normally" engage in sex would be highly valuable for therapeutic interventions involving couples. Other covariates such as number of children, desire to have more children, and frequency of sexual intercourse before and after CVVP diagnoses may be of use to include in future studies. Alternatively, research examining nonvaginal forms of sex that couples with CVVP may utilize to navigate sexual pain while enhancing mutual satisfaction is an important avenue of future study needed to further understand the multifaceted nature of these dynamics. Finally, the present research was cross-sectional, giving us only a snapshot of support dynamics in the context of CVVP and an inability to prove causal relationships. Future research should examine the evolution of intimacy, support, and sexual expression as it unfolds over time within the confines of romantic relationships and beyond, perhaps through the use of longitudinal studies. --- Conclusions In the present paper, we investigated the role of social support-specifically emotional support-and its impact on 333 people of color with chronic vulvovaginal pain . Our results suggested that social support is essential for those who are navigating CVVP and that disclosing their CVVP diagnosis to partners frequently yielded positive and inherently supportive responses. Additional research could benefit from finding suitable strategies to disclose and discuss CVVP with sexual partners to encourage positive interpersonal and sexual health outcomes. Taken together, these findings highlight the importance of emotional support in the context of navigating a sexual partnership while managing CVVP and indicate potential routes to more satisfaction and less distress in the realm of sexual connection. --- Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available because of privacy concerns. --- Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. ---

Within the social support literature, individuals who experience chronic pain have shown many positive outcomes and benefits when receiving the appropriate level of emotional support. In the current study, individuals who experience chronic vulvovaginal pain (CVVP) were asked about their partner's supportiveness, other sources of emotional support, and their satisfaction and stress surrounding sexual activity. The participants (n = 333) also identified as people of color, with a majority identifying as African American or Black (n = 227). The participants indicated that their partners were overall supportive of their diagnoses and found other emotional support sources through medical professionals, vulvar/vaginal pain-specific medical information websites, and family or friends. After conducting linear regressions, results showed the partner supportiveness was associated with less distress and less dissatisfaction surrounding sexual activity. Future research is suggested to further examine social support's role for minority patients who experience chronic vulvovaginal pain.

Introduction The thalassaemias are recessively inherited disorders of haemoglobin synthesis common in people of Mediterranean, Middle East, Indian subcontinent and Asian origins. Thalassaemia is considered the most common genetic condition worldwide, and about 3 % of the world's population carry the β thalassaemia genes . In Malaysia, approximately 4.5 % of the population are carriers of beta thalassaemia , and 3-40 % are HbE carriers . Recent online data from the Malaysian Thalassaemia Registry as of September 2011 showed a total of 5,115 registered patients of which 2,207 have β thalassaemia major and 1,594 have HbE-β thalassaemia. Patients with thalassaemia major require lifelong regular blood transfusions which necessitate burdensome and expensive chelation agents. In Malaysia, patients have free access to blood transfusions in government hospitals nationwide. From the year 2005, chelation agents in the form of subcutaneous deferoxamine and oral deferiprone are also offered free to patients. The newer oral chelator, deferasirox, will become accessible especially for the younger patients from the year 2012. The only curative option is by allogeneic haemopoietic stem cell transplant. In Malaysia, from the year 1987 to 2008, a total of 144 transplants were performed for haemoglobinopathies with an overall disease-free survival of 87.5 % . However, the majority of Malaysian thalassaemia patients will still undergo treatment with regular blood transfusions and iron chelation in the form of subcutaneous deferoxamine. The huge economic and emotional burdens of thalassaemia are important public health issues leading to the need for an effective prevention programme. Many of the thalassaemia prevention programmes that were successful in various countries employed population screening, prenatal diagnosis and selective termination of pregnancy. Consequently, the birth rate of children born with thalassaemia major fell dramatically by 96 % in Cyprus , 94 % in Sardinia and 70 % in Iran Those countries were, however, homogenous with regard to ethnic and religious composition, and employing similar preventive strategies in other countries and cultures may not yield equivalent success due to various dissimilar socio-cultural, religious, economic, technical and ethical perspectives. Additional factors may need to be accounted for in a multi-ethnic country such as Malaysia. Malaysia has a multi-ethnic population of 28.3 million consisting of 50.1 % Malays, 22.6 % Chinese, 11.8 % indigenous peoples and 6.7 % Indians, and the remainder are noncitizens and minority groups . Data from the Department of Statistics in Malaysia also revealed the following-the Malays are all Muslims whereas 87 % of the Chinese population practises a mix of beliefs with influences from traditional Chinese religions such as Buddhism, Confucianism and Taoism, the majority of Indians practise Hinduism and Christianity is embraced by 46 % of indigenous people especially those from the states of Sabah and Sarawak as well as 11 % of Chinese, 6 % of Indians and other minority groups. Based on data from the Malaysian Thalassaemia Registry in 2009, the main ethnic background of Malaysian thalassaemia patients was recorded as Malay , Chinese and Kadazan-Dusun . The Kadazan-Dusuns are the indigenous people of the state of Sabah. Compared to the background population demographics, thalassaemia appeared to affect more Kadazan-Dusuns and Malays compared to the Chinese and Indians . This may be contributed by a higher prevalence of consanguineous marriages as well as carrier rates among the Kadazan-Dusuns where a recent study revealed carrier rates of 33.6 and 12.8 % for alpha and beta thalassaemia, respectively . About 4.5 % of Malays and Chinese are beta thalassaemia carriers , but more Malays are affected by the homozygous states, and failure of preventive strategies among the Malays could have contributed to the ratio seen. The success of preventive strategies depends largely on the acceptance of the targeted communities. Most of the successful thalassaemia preventive strategies incorporated prenatal diagnosis and selective termination of affected foetuses into their programmes. Studies addressing the acceptance of these measures have been done in Muslim-majority countries such as Iran , Pakistan , Lebanon and Saudi Arabia but are relatively scarce in Malaysia. A Malaysian survey among the general lay public documented not only a low level of knowledge about thalassaemia but it revealed that only 36.6 % of participants accepted selective termination of affected foetuses with the Malays less likely to accept abortion compared to the Chinese and Indians. Although the reasons were not explored further, it was perceived that cultural and religious restrictions contributed to the Malay participants' rejection of abortion. Our study addressed the attitudes of Malaysian parents who already have children affected by thalassaemia. These data represent the first such study in the country, and we hope it will provide insights that can contribute towards successful prevention in a multi-ethnic country. --- Methodology This cross-sectional study was conducted over 4 months in the state of Johor which has the fourth largest number of thalassaemia patients among the 14 states in Malaysia. As there were no published questionnaire surveys in this area, we conducted a focus group discussion with five parents of thalassaemic children to assist the development of survey items. A structured questionnaire using layman's terms was developed in English and then translated to Malay and Mandarin Chinese which is the commonest form of Chinese language used by the Malaysian Chinese. Each version of the questionnaire then underwent rigorous pilot testing before it was finalized. The self-administered questionnaires were then distributed to parents of children with thalassaemia major . Participation was voluntary, and all responses were anonymous and returned in sealed envelopes to the attending staff that was also trained to clarify survey items should parents encounter difficulties comprehending them. We used a convenience sample of parents attending the day-care centres in three major hospitals in Johor as well as parents attending a state-level thalassaemia seminar organised in June 2011 in Johor Bahru, the state capital. Where both partners were approached, the questionnaires were administered separately. Demographic details were recorded such as sex, age, religion, ethnic group, highest education attained and number of children with thalassaemia. As many Chinese practise a mixture of Buddhism, Taoism and Confucianism, this is categorised under "Buddhism", and Catholics were included in the "Christians" category. Each parent was asked if he/she was in favour of antenatal diagnosis at the early stage of pregnancy. Participants who favoured antenatal diagnosis were then asked if they would choose to terminate the pregnancy if the foetus was confirmed to have thalassaemia major at an early stage of pregnancy. Reasons for their responses to both items were further explored, where the participants were asked to choose the single most important reason out of a list of predefined choices. A space for free-text comments was also allocated if the reason was not among the pre-defined choices. All the data were analysed using SPSS , and results were presented using descriptive statistics. We compared the responses of our participants using chi-square tests via cross-tabulation for categorical data or t test for independent groups for continuous data. This study was reviewed and approved by the Malaysian Ministry of Health Research and Ethics Committee and performed in accordance with the ethical standards in the 1964 Declaration of Helsinki. --- Results All 117 parents approached agreed to participate in the survey. One participant did not respond to questions on antenatal diagnosis and termination of pregnancy and was excluded from the analysis; thus, we achieved a 99.1 % response rate of completed surveys. The demographic details of the participants are depicted in Table 2. Out of 116 parents who completed the survey, 83 were in favour of antenatal diagnosis. No significant factors were associated with a preference for antenatal diagnosis . The 33 who declined antenatal diagnosis gave the main reasons as "condition not serious enough to warrant a termination of pregnancy" , "emotional reasons" and "termination of pregnancy is forbidden by my religion" . Out of 83 parents who were in favour of antenatal diagnosis, only 33 were agreeable to terminate a pregnancy carrying a foetus with thalassaemia major . For the remainder 50 respondents who declined, "termination of pregnancy is forbidden by my religion" , "condition not serious enough to warrant a termination of pregnancy" and "emotional reason" were the main reasons given . Out of the 38 parents who refused to terminate the pregnancy based on religious grounds, 36 were Malay Muslims and the remainder 2 participants were Chinese . Religion was significantly associated with a decision to terminate an affected pregnancy among parents who agreed for antenatal diagnosis , with 73.4 % of Muslim participants against termination compared to 25 % of Christians and 13.3 % of Buddhists . Comparing different ethnic groups, a greater proportion of Malays compared to Chinese declined termination . There were no difference among the respondents who were in favour of termination versus those who were against termination in terms of gender, age, highest education level and number of children affected with thalassaemia . Chinese: •Buddhist, 17 •Christian, 2 •Muslim, 1 Indian: •Hindu, 2 a The total response is less than 116 due to missing data --- Discussion In a cross-sectional survey among 3,723 members of the general public in Malaysia , only 36.6 % were supportive of selective termination of foetuses diagnosed with thalassaemia major. As these households have no affected children, many may assume that parents with thalassaemic children will be more inclined to terminate affected pregnancies having personally raised a thalassaemic child which leads to greater awareness of the suffering that this condition can where most women wanted prenatal diagnosis as they wish to know their child's condition, but they did not necessarily agree to the termination of affected foetuses. In our study, religious consideration was the main reason offered for declining selective abortion , and this mainly affects the Malays who are all Muslims. In Malaysia, the Muslims' family and religious practices are guided by authorities like the National Council of Islamic Religious Affairs which oversees religious decrees at the national level. However, in states where there are reigning sultans, religious affairs by legislature are under the purview of the various state Islamic religious councils. The national council decrees are legally non-binding and serve as guidance to be adopted upon consent of the sultan. In relation to thalassaemia, the Fatwa Committee, at the Council's 52nd Muzakarah in July 2002 has decreed that "termination of pregnancy before 120 days is permissible if the foetus is disfigured, ill and can harm the life of the mother". As a foetus affected with thalassaemia major does not fully qualify the above prerequisites, this affects the Malay Muslims' full acceptance of termination of thalassaemic foetuses. With regard to the Islamic stance on termination of a foetus with a serious disorder, a number of Islamic countries like Kuwait and Saudi Arabia have ruled that termination is permissible before ensoulment at 120 days of gestation . In Pakistan, two renowned Islamic scholars ruled that a pregnancy can be terminated if the foetus is affected by a serious genetic disorder and if termination is before 120 days of gestation . In Iran, abortion was illegal when pre-marital screening was first introduced, but based on systematic feedback by the population screened, intensive and widespread ethical discussions were held which led to a Fatwa issued in 1998 which allowed pregnancy termination up to 16 weeks for some genetic disabling disorders including beta thalassaemia . Most of the religious edicts do not permit abortion after 120 days unless the mother's life is harmed. Muslims are committed to comply with religious teachings, and religious permissibility in these countries has led to acceptance of antenatal diagnosis and termination of thalassaemic foetuses. The acceptance rate was 90.3 % in Iran and 75.2 % in Pakistan . Recent studies on Muslim populations revealed that education about the religion's stance on termination of pregnancy improved uptake of prenatal diagnosis and termination of pregnancies affected by thalassaemia and sickle cell anaemia . For example, in Saudi Arabia, Alkuraya and Kilani found that 28 out of 32 Muslim participants refused abortion, but 13 changed their minds after they were given the Fatwa on abortion. In Egypt, where previously there was a poor uptake of termination among pregnant mothers with affected foetuses, a recent study showed that with comprehensive in-depth counselling addressing the Islamic aspects of termination, uptake of selective abortion increased significantly . These studies highlight the importance of religious permissibility as well as education of the target population to enable a successful prevention programme. Besides religious teachings, other factors which may affect a Muslim woman's decision on abortion are the perceived severity of the disease and family support . In our study, 86.7 % of the Chinese in Malaysia accepted termination. Most of the Malaysian Chinese practise a mixture of Buddhism, Taoism and Confucianism, and a smaller minority embraces Christianity. The lack of strict religious guidelines on abortion issues, less compulsion to follow religious guidelines within the community as well as a more pragmatic approach in life most likely led to the higher acceptance rate of abortion. As for the Christians who made up 6.8 % of our study population, 25 % were against termination, and those who were Roman Catholics appeared to decline abortion as this is forbidden among them. In our cohort of 116 parents, 33 parents have declined prenatal diagnosis, and the most cited reason was that "the condition is not serious enough". Even amongst parents who accepted antenatal diagnosis but declined abortion, this was the second most cited reason after religious restrictions. This viewpoint reflects the improvement in the Malaysian health care services for patients with thalassaemia. Prior to 2005, patients were only provided with free access to blood transfusions, and the majority of them succumbed to severe iron overload in their teens. However, since 2005, the government provides free iron chelation, and survival rates together with quality of life have improved leading parents to view that thalassaemia is not a "serious condition". We did not study the age of the affected children amongst parents surveyed, but most of the children were likely to be relatively young as the mean age of their parents in the survey was only 40.4 years old. As most complications tended to set in during adolescence or adulthood, parents of these young children might have regarded thalassaemia major as a "condition not serious enough to warrant abortion". It is likely that these parents have not been adequately informed of the long-term complications, management and the effects on quality of life of this condition; hence, the importance of providing this information during counselling sessions to atrisk couples should be emphasized in the future Since year 2005, the Malaysian Ministry of Health has launched a national programme to prevent and control thalassaemia nationwide. Best possible care in the form of free iron chelation as well as public awareness and health education campaigns are carried out. Screening, however, is offered only to interested parties who present to the medical facilities on a voluntary basis, and pre-marital carrier screening in general is still not mandatory. After an affected child is diagnosed, retrospective genetic counselling is offered as well as cascade screening to relatives. Prenatal diagnosis and termination of pregnancy for thalassaemia is available since 20 years ago in Malaysia . This approach which has led to successful prevention in many countries has yet to be actively incorporated into our Malaysian national programme partly due to the current conservative and cautious stand on abortion by the consensus decision of the Fatwa council members. However, as exemplified in other Muslim countries and precedents from other medical-related decrees, the consensus opinion may change to accommodate new evidences or other mitigating circumstances that would allow termination of pregnancy in selected cases. Pre-implantation genetic diagnosis which may be more acceptable is still not feasible due to financial and technical reasons. We recognise the following strength and limitations in our study. The high response rate we achieved to our survey was most likely due to the convenience sample gathered in a single sitting, the anonymous manner in which it was conducted as well as the items which were concise and easily understood by our participants who received at least a primary school education. As the survey was conducted in settings where their children's health care providers were present, the parents may have been compelled to accede to the request to participate in the survey. Furthermore, in such settings, participants' responses might be biassed towards giving socially desirable responses despite the anonymised nature of our survey. For instance, the opinions on prenatal diagnosis and abortion might have been skewed towards what they perceived the health care providers were in favour of. When exploring the reasons parents chose a certain reproductive option, we provided pre-defined choices and limited the response to a single selection. This approach was useful in standardizing responses, but it might not have encompassed the breadth and depth of possible responses which might be obtained had we chosen to explore this in focus group discussions. Further clarification should have been sought for parents who chose "emotional reasons" so that their values, beliefs or feelings could be identified. We used "early stage of pregnancy" in our survey without specifying the gestational age, and this might have led to some uncertainties in the participants' responses. Additionally, our study was conducted in the day-care centres of three referral hospitals where two thirds of the patients in Johor state sought treatment as well as among parents who attended a thalassaemia seminar in the state capital. Therefore, parents staying in the smaller districts or those who lacked the motivation or opportunity to attend the seminar may be under-represented. Lastly, the findings in the state of Johor cannot be generalised to the Borneo state of Sabah which carries the highest burden of thalassaemia among all the states in Malaysia. Most of the Sabahan patients are from the Kadazan-Dusun ethnic group where a sizeable proportion of them are Catholics who may reject abortion for thalassaemia major based on religious restrictions. --- Conclusion This study highlighted a number of important observations with far-reaching implications for the prevention of thalassaemia in a multi-ethnic country. We have shown that even among parents of children with thalassaemia major, there is a low acceptance rate for prenatal diagnosis and termination of affected pregnancies. Acceptance rates for abortion vary markedly depending on the ethnic and religious backgrounds where perceived religious non-permissibility is an important factor leading to a high rate of non-acceptance among the Malay Muslims. In a country with such a diverse cultural background and religious beliefs like Malaysia, it is essential for us to understand the issues in depth so that we can implement effective strategies towards successful prevention. This study suggests that scholarly deliberations among the Malaysian Muslim religious authorities that result in a supportive stance in prenatal diagnosis and selective abortion may contribute to a more successful prevention programme, but this issue should be approached with a lot of tact and sensitivity. Besides that, education should be improved among parents with affected children, and this should include adequate information on the long-term morbidity and mortality associated with thalassaemia. They can thus acquire a balanced viewpoint, and an informed decision may then be made with regard to their reproductive options. Our survey seeks to address the views of thalassaemia carriers, although their views might not necessarily translate accurately into actual behaviours. We therefore suggest that future studies should employ a more objective approach by observing and reporting the actual choices made by participants when offered genetic counselling and prenatal diagnosis. Last but not the least, the health authorities will need to explore various other strategies such as public education, an effective national screening programme as well as the accessibility of counselling, genetic and antenatal services so as to ensure an effective prevention programme which will be sustainable in the long run. ---

Thalassaemia is a public health problem in multiethnic Malaysia which mainly affects the Malays, Kadazan-Dusuns and Chinese. This study, the first in Malaysia, aims to evaluate the acceptability of prenatal diagnosis and abortion among Malaysian parents who have a child or children with thalassaemia major and the socio-demographic factors affecting their decision-making. A pre-structured questionnaire was distributed to parents of children with thalassaemia major. Response rate for completed surveys was 99.1 %. Out of 116 respondents, the majority (83/71.6 %) were agreeable for prenatal diagnosis, but only 33 (28.4 %) agreed to both prenatal diagnosis followed by termination of affected foetuses. Of parents who declined abortion, 77.6 % cited religious restriction as the main reason, and their religious background was a significant factor (p=0.001), with 73.4 % of Muslim participants against termination compared to 25 % of Christians and 13.3 % of Buddhists. Gender, age, highest education level and number of children affected with thalassaemia were non-significant predictors in decision-making regarding abortion. The acceptance rate for termination of foetuses with thalassaemia major in Malaysia is low especially among the Muslims due to religious non-permissibility. Therefore, scholarly deliberations among the Malaysian Muslim religious authorities that result in a supportive stance in this issue may contribute to a more successful prevention programme.

Introduction Aboriginal and Torres Strait Islander knowledge and wisdom has long recognised the role of social and cultural factors on health and well-being . Aboriginal and Torres Strait Islander holistic health philosophy describes social and emotional well-being as the interconnection of social emotional, spiritual, cultural factors on health and wellness of not just individuals but communities . Social and emotional well-being as conceptualised by Gee et al. recognises the ongoing influence of historical, political and social factors on health and social outcomes . These social factors are internationally described and recognised as the social determinants of health and are estimated to contribute to 34% of the overall burden of disease experienced by Aboriginal and Torres Strait Islander people . Both internationally and cross-culturally peer-reviewed literature has established associations, explored pathways and biological mechanisms providing a critical knowledge base on the role of social factors on health . Despite these understandings, there is limited evidence on effective intervention strategies that address how these social factors influence health outcomes within the population . Recent government consultations highlight the importance of self-determined and timely action on the social determinants of health for Aboriginal and Torres Strait Islander communities incorporating system responses that are coordinated, culturally relevant and strengths-based . Health systems face challenges in responding to the complex nature of the social determinants of health with collaborations required across health and social services; nonetheless, the clinical frontline workforce have been recognised as a potential catalyst for change in any systems response . Clinical workforce approaches that include screening clients for social and emotional well-being facilitate the early identification and management of needs, planned and coordinated responses and the monitoring of progress and outcomes . In a current context, Aboriginal Community Controlled Health Services and primary health care services are 'doing whatever it takes' to meet the social and emotional well-being needs of Aboriginal and Torres Strait Islander people which includes addressing the social determinants of health in service delivery . Consultations with ACCHOs have highlighted key principles which inform holistic approaches to the social determinants of health including self-determination, accessible and culturally safe care and strong partnerships that support clients to navigate social services . A recent document analysis of 67 ACCHO annual reports found that all services were working to improve clients' intermediary social determinants of health, specifically material circumstances, biological, behavioural and psychosocial factors . Whilst structured and funded Aboriginal and Torres Strait Islander health assessments for preventative care are widely implemented, these assessments are limited by a biomedical focus that inadequately addresses social and cultural factors . Across organisations there are varied responses depending on the capacity of the primary health care service . Furthermore, service delivery protocols for addressing the social determinants of health and more broadly data systems for monitoring their actions are not well established . Strengths-based, person centred and empowerment approaches are often used synonymously to describe the delivery of health care for Aboriginal and Torres Strait Islander people. These approaches promote individuals control over their own lives and focus on abilities and resources to enable self-determination . Aboriginal and Torres Strait Islander people who have increased control and mastery over their lived experiences are empowered in their engagement with social and health services . Health care services commonly describe intentions to deliver strengths-based approaches, yet the practical and genuine implementation with Aboriginal and Torres Strait Islander people is still emerging in practice . Holistic case management models are well suited for strengths-based practice which focuses on empowering people to take charge of their own lives and to support the identification of existing strengths and resources . Case management approaches whilst diverse across disciplines and in different contexts usually include the following core functions; assessment, planning, linking, monitoring, advocacy and outreach services . Case management approaches in primary health care with Aboriginal and Torres Strait Islander people report improvements in selfrated health status, reduction in depression and improved measures of diabetes control . These findings suggest that patient-led case management has the potential to enhance holistic approaches to social and emotional well-being . The effects of colonisation and the continuing social and political oppression and dispossession of Aboriginal and Torres Strait Islander communities have contributed to significant socioeconomic and health inequities . Persistent and disproportionate inequalities experienced by Aboriginal and Torres Strait Islander people highlight the need to better understand and respond to social and emotional well-being needs which includes the social determinants of health. There is a pressing need for coordinated best practice responses to social and emotional well-being screening and management, dedicated resources, training and ongoing monitoring . Existing evidence has not yet described approaches that collectively inform health care responses for Aboriginal and Torres Strait Islander social and emotional well-being. To address this gap, a pilot program has been designed within a research setting and includes the following key elements: i) identifying unmet needs, ii) strengths-based case management, iii) document and monitoring, iv) culturally relevant supervision and v) evaluation. The aim of this manuscript is to describe and critically explore the program's key elements from an Aboriginal and Torres Strait Islander perspective as part of strengthening practice-based evidence on social and emotional well-being. --- Discussion --- Program context The Cultural Pathways program is implemented by Wardliparingga Aboriginal Health Equity research team in the South Australian Health and Medical Research Institute, Adelaide, South Australia. Wardliparingga undertakes research that is of relevance to South Australian Aboriginal and Torres Strait Islander communities through partnerships, collaboration, respect, reciprocity and for the benefit of community . The Cultural Pathways Program is designed and implemented by Aboriginal and Torres Strait Islander people as a response to community identified needs. The program is implemented within an Indigenous methodological framework and from inception to implementation the program has been underpinned by Aboriginal and Torres Strait Islander ways of knowing, being and doing . Priority areas for research were established through extensive consultation and engagement with the community . All programs of work implemented by Wardliparingga have Aboriginal and Torres Strait Islander leadership and governance, through these structures the community consistently highlighted that more holistic responses, which included the social determinants of health, were required. The research team is predominantly Aboriginal and Torres Strait Islander researchers who bring wisdom and experience to the development of the program approach and implementation ensuring consistent alignment with Aboriginal and Torres Strait Islander ways of knowing, being and doing. The program described in this manuscript was approved by the Aboriginal Health Research Ethics Committee of South Australia . The program approach includes comprehensive screening utilising a specifically developed holistic screening tool to identify unmet social and emotional well-being needs. Following screening, facilitators implement strengths-based case management through goal setting, prioritisation and brokering connections to services. program structures embed documentation and monitoring of the program's social and emotional well-being responses, actions taken to address needs and outcomes for participants. These elements are underpinned by culturally relevant supervision, reflective practice and evaluation. The program approach critically explores the benefits, cultural relevance and responsiveness of common practices in case management. Through a combined understanding of these approaches, the program seeks to inform the evidence base for strengthened and coordinated responses to Aboriginal and Torres Strait Islander social and emotional well-being. Program delivery is undertaken by male and female facilitators with workforce roles informed by a navigator approach, to assist individuals' engagement with the health care system and to overcome any barriers to care . Referrals are received from a large-scale population-based biomedical cohort study of Aboriginal and Torres Strait Islander South Australians. As part of the study, all participants receive a comprehensive health assessment that includes questions regarding their social and emotional well-being. Further to this, community engagement and consultations highlighted that post-study follow-up responses for participants would require addressing social and emotional well-being needs such as psychosocial health, financial literacy, food security and material circumstances. Participants are offered a referral to the Cultural Pathways Program, if unmet social and cultural needs are identified during the assessment. The implementation setting replicates real-world service delivery models where presentation may initially be for a physical health need. Upon receipt of referrals from the study team, the Cultural Pathways Program facilitators connect with participants and implement the flexible participant led case management process . Program elements informing a social and emotional well-being response have been detailed within the following sections, providing the theoretical underpinnings, Cultural Pathways Program approach, embedded Aboriginal and Torres Strait Islander ways of working and opportunities for strengthening practice. --- Identifying unmet needs Screening and assessment is a common first point of engagement in health settings and appropriate screening delivered as part of routine practice can enhance the timely and effective identification of needs and accordingly inform responses or prompt a more comprehensive assessment . Indigenous specific health assessments are associated with improved preventive care for a range of health needs; however, a greater focus is needed on social and cultural factors . Cultural Pathways Program facilitators implement a modified Social Needs Screening Tool to identify unmet social needs of participants. Developed through an Aboriginal and Torres Strait Islander researcher led process, with community input to ensure cultural relevance and responsiveness, the adapted holistic tool covers well-being domains including mental health and cultural and community connection and social domains including financial and food security, transport, employment, housing and social isolation. The process of cultural development ensures the questions are relevant, asked the right way, with cultural meaning and are best able to identify the unique needs of Aboriginal and Torres Strait Islander participants . Screening processes for the social determinants must be accompanied by plans for action , and as part of the program's case management approach the screening process assists the Facilitator to understand participant needs and enables the identification and prioritisation of participant goals. By implementing a structured and consistent approach, identifying and documenting unmet needs enable the measurement of actions, activities and the monitoring of participant outcomes. --- Strengths-Based case management The program's case management approach includes goal setting, prioritisation and brokering connections to services. Facilitators work in partnership with participants and tailor responses to individual circumstances and needs. A strengthsbased approach to case management ensures facilitators focus on clients' abilities, talents and resources to enable client's self-determination skills, develop resilience and the ability to respond or navigate similar situations in the future . Goal setting is a common step in the case of management process with theoretical concepts highlighting the importance of collaboration for effective goal setting . An individuals' sense of control and autonomy influence their willingness to set goals and efforts for achieving them . The Australian Integrated Mental Health Initiative is an existing framework that uses strengths based story telling . The Cultural Pathways Program implements a goal and priority setting framework utilising the AIMhi Pictorial Care Plan to explore physical, emotional, spiritual, cultural, family, social and work contexts to identify worries, strengths and resources. Consistent with Aboriginal and Torres Strait Islander ways of working, facilitators work in partnership with participants to identify and prioritise issues of most importance that will support improved well-being. As part of the strengths-based, empowerment and person-centred approaches, participants define their own priorities contributing to enhanced autonomy, control and self-efficacy. As part of the 'brokering' approach to case management, facilitators connect participants with services to meet their needs. Making a referral to other services, organisations or agencies are widely implemented in health and social services. Social and emotional well-being and social determinants of health needs span across sectors with often multiple services and agencies involved, this requires coordination to minimise the burden on service users and to enable referrals and connections . Brokering connections relies on relationships, understandings of what is available across the breadth of health and social needs and understandings of culturally relevant services . To support this approach, facilitators undertake service mapping exercises to identify the available services and will pro-actively seek the most appropriate service to connect a participant to and reduce barriers to access these services . Facilitators actively support participants to access services by contacting services on behalf of participants, supporting participants when they contact services themselves and follow up contact with participants to monitor progress. If necessary, facilitators address any challenges or barriers to support the best possible outcome. The active and coordinated approach to brokering connections enhances service access for participants and enables the program to also monitor brokerage outcomes. --- Documentation and monitoring Program monitoring involves measuring and reporting on progress and creates opportunities for continuous quality improvement . Currently, health services rarely systematically collect data about or measure activity on the social determinants of health and require a mechanism to monitor and evaluate the impact of social and emotional well-being services they provide to address health outcomes . Comprehensive understandings of the most appropriate measures for Aboriginal and Torres Strait Islander social and emotional well-being and the social determinants of health are still emerging. Existing national measures of well-being include psychological distress, positive well-being, anger, life stressors, discrimination, cultural identification and removal from natural family . Measures for the social determinants of health as described by the World Health Organization Conceptual Framework and outlined in the Aboriginal and Torres Strait Islander Health Performance Framework include domains such as connection to country, education, employment, health system, housing, income and transport. The Cultural Pathways Program combines social and emotional well-being and social determinants of health measures as part of the programs' monitoring framework. The program utilises REDCap , a secure web platform for managing online databases . The platform collects participant information, demographics and activity data which include when and how people are contacted and the services provided by social/health domain. The program measures factors such as unmet needs, identified goals, whether they have been achieved and the service connections made. The program utilises routine data for ongoing monitoring, quality improvement and as part of funding requirements and obligations. The data collected by the program was informed by Aboriginal and Torres Strait Islander understandings of health and well-being and the wisdom and expertise of the research team and community. The process included the collective development of culturally relevant measures in relation to social and emotional well-being, specifically practical ways to measure progress towards addressing complex social and cultural factors. This process enabled the program to capture information that is useful and relevant for Aboriginal and Torres Strait Islander people. A structured and consistent approach to identifying needs and a specifically designed monitoring framework enables the program to measure progress or outcomes which can be used to understand the needs of service users, to plan responses and to advocate for resources . --- Culturally relevant supervision Reflective practice and clinical supervision are recognised by many professions for their role in supporting enhanced clinical practice as well as the health and well-being of the workforce . This is particularly important for Aboriginal and Torres --- 4 Tina Brodie et al. Strait Islander health workers and practitioners who have complex experiences including burnout and vicarious trauma . The Aboriginal and Torres Strait Islander health workforce and non-Indigenous workers in Aboriginal and Torres Strait Islander health contexts require access to high-quality cultural and clinical supervision which supports cultural safety, improved practice and well-being . Available frameworks for culturally appropriate supervision with Aboriginal and Torres Strait Islander people include considerations for working with community, looking after self, understanding of roles and professional practice (Koivu et al., 2012;Nelson et al., 2015;Scerra, 2012 The Cultural Pathways Program utilises these existing frameworks as well as the knowledge and experience of program staff to implement a culturally relevant reflective practice and supervision model. An experienced Aboriginal clinician supports facilitators through a range of structures including weekly clinical yarning, one to one yarning and debriefing opportunities as required. Facilitators share perspectives, feelings, challenges, barriers and enablers in relation to both clinical practice as well as system, policy and organisational factors which impact the participant, Facilitator, or the program. Fundamentally, the supervision and reflective practice model are culturally grounded in relationships and yarning to support the cultural safety for Aboriginal and Torres Strait Islander participants whilst also enabling the retention and well-being of the Aboriginal and Torres Strait Islander workforce. --- Developmental evaluation Evaluating health programs and initiatives supports implementation across different contexts utilising insights into how and why they work and whether they have been effective . There is an increasing recognition of the important role of evidence-based programs featuring high quality and culturally relevant evaluation . The Cultural Pathways Program is underpinned by an Indigenous methodological evaluation framework which utilises Developmental Evaluation, an approach to evaluation that supports innovation and adaptation in complex environments , and is consistent with Indigenous methodology and participatory approaches requiring partnerships, trust and shared decision making . The key to developmental evaluation is that the evaluator works with the team in real-time, asking evaluation questions, examining and tracking implications of adaptations and providing timely feedback as the program is implemented and modified or adapted as needed. The evaluator as an Aboriginal woman is immersed in as an insider drawing heavily on reflective practice and utilising the cultural knowledge and expertise of the evaluator as part of the evaluation method. The aim of the evaluation is to understand the process including what was delivered, how it was implemented and the experiences of program participants. The evaluation through reflective and formative methods supports further understanding on the interactions between facilitators, program participants and the broader health and social service contexts. The evaluation framework includes community engagement, governance and approaches which have been purposely selected for their consistency with Indigenous methodologies. This framework ensures that the participation and voice of the community are therefore embedded throughout implementation to support tangible benefits to the community . --- Conclusions There is a knowledge to action gap on how to assess and address the social determinants of health within clinical practice to inform the development of coordinated, culturally relevant and strength-based responses to meet the holistic social and emotional well-being needs of Aboriginal and Torres Strait Islander people and communities. Primary health care services, often as the entry point for accessing health services, are well positioned to implement coordinated health equity responses which include addressing the social determinants of health . The absence of a readily applied model creates challenges for the provision of coordinated, resourced and systemic responses to the social determinants of health . Routine screening for unmet needs, implementing strengths-based practice, connecting people to what they need, monitoring service provision and providing clinical and cultural support for the Aboriginal and Torres Strait Islander workforce align to existing practice and are transferable across contexts. Continuous quality improvement and monitoring enables primary health care services to embed new practices into services, systems and routines . The ability to implement holistic approaches to Aboriginal and Torres Strait Islander health through the intersection of health and social services requires adequate resources, training and support to clinical workforce , including consideration of roles, responsibilities, scope of practice and readiness to implement strengths-based approaches. These changes cannot be implemented without addressing the ongoing impacts of racism and oppression of Aboriginal and Torres Strait Islander people, allowing for culturally safe systems which are able to meet holistic social and emotional well-being needs . The Cultural Pathways Program builds on existing approaches to contribute to practice-based evidence of culturally relevant case management approaches which can be utilised as part of routine care to strengthen the systematic identification and response in primary health care delivery. The combined understandings of the elements outlined in this manuscript provide a framework to inform service planning and tailored implementation which can strengthen social and emotional well-being responses for Aboriginal and Torres Strait Islander people. policy, health systems and inequity comprises a unique comprehensive skillset relevant to existing and emerging complexities of Aboriginal health and well-being. Specifically, Odette seeks to understand how institutional policies and practices drive health and social inequities experienced by Indigenous populations. Her novel approach is the use of community-level information to show and explore the reasons for variations in disadvantage both within the Aboriginal community and between the Aboriginal and non-Aboriginal community. Integral to her research is the inclusion of Aboriginal communities in defining their health and well-being and how Indigenous data can be governed in the future to derive greater benefit for the population. Luke Cantley has family connections to the Gunditjmara nation of Victoria and is a Research Associate located within Flinders University. Through his research, Luke is determined to solidify Aboriginal culture as a protective factor within the child protection system, whilst exploring the nuances between child safety and cultural safety. Luke holds extensive knowledge on the role unmet social and cultural needs have on positive health outcomes within the Aboriginal and Torres Strait Islander community and holds a strong passion for advocating for increased health care utilisation for health care consumers. Luke has gained extensive experience working as an Aboriginal Health Worker within a strengths-based approach across diverse sectors including Prison health, Primary Health Care, Public Housing and Mental Health Services. Developing expertise in 1) Culturally appropriate and ethical ways of engaging within the community, 2) Health and Well-being assessment methods fostering participatory action research, and 3) Social inequities generated by reduced access to services or resources. Peita Cooper has a Bachelor of Social Work and currently works within the justice sector. Peita commenced as a Graduate in Wardliparingga Aboriginal Health Equity Theme, at the South Australian Health and Medical Research Institute . As a Program Facilitator, Peita contributed to the delivery of strengths-based case management and developing culturally responsive practice with Aboriginal and Torres Strait Islander peoples' and communities. Her previous experience includes working in the disability sector. Seth Westhead has family connections to the Awabakal and Wiradjuri nations of NSW and is a Research Associate with Wardliparingga Aboriginal Health Equity Theme, South Australian Health & Medical Research Institute. Through his research work, Seth strives to better understand how social and cultural determinants drive health and social inequities within society, particularly as it relates to the Indigenous population. He seeks to better equip communities and young people with tools and evidence for public health advocacy and enable communities to translate health research into meaningful action. Specifically, Seth has expertise in the: 1) conceptual development of Aboriginal specific social determinants and well-being frameworks and tools, 2) implementation of projects involving community engagement and community-led governance structures 2), and 3) undertaking of qualitative research methodologies and community and stakeholder participation interpretation of findings. Alex Brown is an Aboriginal medical doctor and researcher, he is the Theme Leader of Wardliparingga Aboriginal Health Equity and Professor of Medicine at the University of Adelaide. He grew up on the south coast of New South Wales with family connections to Nowra, Wreck Bay and Wallaga Lake on the far south coast of NSW. Over the last 20 years, Alex has established an extensive and unique research program focused on chronic disease in vulnerable communities, with a particular focus on outlining and overcoming health disparities. He leads projects encompassing epidemiology, psychosocial determinants of chronic disease, mixed methods health services research in Aboriginal primary care and hospital settings, and randomised controlled trials of pharmacological and non-pharmacological chronic disease interventions. Alex has been heavily involved in engaging government and lead agencies in setting the agenda in Aboriginal cardiovascular disease management and control and chronic disease policy more broadly. He sits on a range of national committees, and co-chairs the Indigenous Research Health Fund through the MRFF. Natasha Howard is the Wardliparingga Platform Lead: Implementation Science. The platform incorporates a systems view and privileges Indigenous knowledge to deliver mixed-method inter-disciplinary perspectives which aim to generate policy and practice-based evidence on the social determinants of health. Her experience spans both the health and social sciences, applying population approaches to investigate how the social and built environment enables and promotes cardiometabolic health and well-being, notably for priority populations. She has been active in advocacy and mentoring of the local population health community in both research and practice. Conflict of interest. None. Ethical approval. This program was approved by the Aboriginal Health Research Ethics Committee of South Australia .

Aboriginal and Torres Strait Islander holistic health represents the interconnection of social, emotional, spiritual and cultural factors on health and well-being. Social factors (education, employment, housing, transport, food and financial security) are internationally described and recognised as the social determinants of health. The social determinants of health are estimated to contribute to 34% of the overall burden of disease experienced by Aboriginal and Torres Strait Islander people. Primary health care services currently 'do what it takes' to address social and emotional well-being needs, including the social determinants of health, and require culturally relevant tools and processes for implementing coordinated and holistic responses. Drawing upon a research-setting pilot program, this manuscript outlines key elements encapsulating a strengths-based approach aimed at addressing Aboriginal and Torres Strait Islander holistic social and emotional well-being. The Cultural Pathways Program is a response to community identified needs, designed and led by Aboriginal and Torres Strait Islander people and informed by holistic views of health. The program aims to identify holistic needs of Aboriginal and Torres Strait Islander people as the starting point to act on the social determinants of health. Facilitators implement strengthsbased practice to identify social and cultural needs (e.g. cultural and community connection, food and financial security, housing, mental health, transport), engage in a goal setting process and broker connections with social and health services. An integrated culturally appropriate clinical supervision model enhances delivery of the program through reflective practice and shared decision making. These embedded approaches enable continuous review and improvement from a program and participant perspective. A developmental evaluation underpins program implementation and the proposed culturally relevant elements could be further tailored for delivery within primary health care services as part of routine care to strengthen systematic identification and response to social and emotional well-being needs.

'moving furniture' in the context of tango dancing only became clear to me five months later. According to a common way of evaluating female performance in tango dancing, women who deploy the ability to respond immediately and effortlessly to their dancing partner's movements with the complementary ones, are considered to be 'light women' . In this context 'light' means kinetically imperceptible and predictable, so as to allow for male choreographic improvisation. Trying to understand the self-image of Latin American women, many Anglophone studies have turned to the influence of Catholic feminine images, especially the Virgin Mary -following the analysis of Emily Stevens in Female and Male in Latin America . Albeit later studies have criticized this emphasis on Marianismo for providing a passive image of women the prevalence of gender images of Catholic origin continues to dominate research on the continent . Studies continue to discuss women's interpretation of these images , the moral lessons derived from them and how they use them to maintain valued self-concepts . Based on 12 years of participant observation, in tango milonguero lessons -first as a student of the 'woman's role', then of the 'man's role' and finally as a teacher -and at night-time milongas in downtown Buenos Aires, I take here a different stance, analysing the image of the 'light women' as an allegorical rendition of heterosexual relationships originating in tango dance halls. This analytical frame is inspired by cultural critiques viewing feminine figures elsewhere as the product of the allegorization and feminization of the relationships of colonizers with the colonized and conquerors with uncultivated lands . At a difference with this literature, it considers the image under study as emerging from the immediate context where it is exchanged. Thus, this work aims to make a twofold contribution to sociological knowledge on feminine images. First it seeks to enlarge knowledge on the diversity of feminine imagery in Latin America by depicting a feminine figure that shows no connection to Catholic imagery. Second, by showing the resonance of the image of the 'light woman' and the characteristics of heterosexual relations in tango social dance events, this article suggests the possibility that gender images may be understood as the feminization and allegorization of relations involving men and women in the very contexts where such images circulate. --- Milongas and tango milonguero lessons In Buenos Aires the word milonga designates either a musical genre, or a public social event where people dance tangos, milongas and valses criollos. I here employ the term in the latter sense, as a dance event. With few exceptions, while Saturday night milongas are traditionally attended by couples who go to dance with each other and their common friends, those taking place from Sundays to Fridays are attended by people who go alone, or with tango classmates and friends, to dance with partners they meet there. Among the latter there is a small but growing number of milongas where homodancing prevails. Throughout this article I refer to milongas taking place on Sunday to Friday nights in downtown Buenos Aires where heterodance prevails. Regular attendants, called milongueros, distinguish 'orthodox' or traditional milongas from 'relaxed' ones -milongas relajadas, sometimes called 'practices' . 'Orthodox milongas' is a native term that indicates dance events that are regulated by strict 'codes' . Accordingly, men and women sit opposite each other on either side of the dance floor, there is a strict dress code of smart clothes and leather shoes and no jeans or trainers are allowed. Men invite women to dance by fixing their gaze on those who are watching them, and then moving their heads upwards and sideways -a move called cabeceo. If the woman nods acceptingly, the man will go up to her table and then escort her to the dance floor. After the dance, he will accompany her back to the table. The couple's dance lasts four tangos, four milongas or four waltzes, depending on each specific tanda -as these sets of music are called. In these milongas, the dancing tandas only include these rhythms, played in succession and divided by cortinas, that is, excerpts of other musical genres that nobody dances to. At the beginning of each tango, the couple spends a few seconds in small talk, and then starts dancing, anti-clockwise, without overtaking or touching other couples, and only stopping to make some steps for no more than two or three counts -if the space allows -or else they keep moving forward. The couple's feet do not lift high from the floor, specially to avoid movements that may hurt nearby dancers. Once the tanda is finished, the man walks the woman back to her chair and returns to his, in order to invite another partner when the following set starts. The distribution of the places where one can sit is dictated by the time and frequency a person has been attending the milonga or others within the same circuit. When people are sitting at the tables, they do not talk or only do so in a low voice, in order not to disturb the dancers. The music played is mostly the recordings made by the 'typical orchestras' in the 1940s. These orchestras are known by the names of their directors . Each tanda usually pays tribute to one of these artists -thus providing rhythmic homogeneity. Generally speaking, the 'orthodox' milongas are more populated by older people than the 'relaxed' ones, although there may be a few younger dancers, especially women. In the former, there is a criteria based on 'dance standards' when choosing a partner. The evaluation of these standards is related to the years spent in the milonga and the reputation of the dancers . There is a tendency to choose partners who have the same dance level, although young women can quickly climb up -or skip -the hierarchy because of their beauty . Most of the 'codes' respected in the 'orthodox' milongas do not exist in the prácticas and 'relaxed' milongas. Here, people attend in mixed groups and share the same table, talk while others are dancing and pay less attention to what happens on the dance floor, except when a new tanda begins and invitations are made to dance. Besides tangos, waltzes and milongas, it is common to have a tanda of other rhythms during the night. The dance floor is not as clearly separated from the walking areas and there are verbal invitations to dance, as well as cabeceos. On the dance floor, figuras on the spot are frequent, with legs unfolding upwards, backwards or sideways. Contact with other couples is, if not welcomed, at least tolerated and excused with a smile, as long as it is followed by an apology. The circulation is still anti-clockwise, but it becomes slower as each couple stops to make more intricate movements, taking up more space and time. Often, couples of women or men dance together and, in some, though it is still rare, women 'lead' and men 'follow'. Many of these milongas include tandas of live music and recordings that would be unsuitable in an 'orthodox' milonga, either because of their nature -far from the 'typical orchestras' from the 1940s and their compositions for piano, bass, violins and bandoneones -or because they include more recently composed tangos. In general, in these prácticas or 'relaxed' milongas, the entrance fee is lower than in the 'orthodox' ones and, occasionally, the fee is a la gorra, that is, at the discretion of the attendees who contribute with what they can. Despite the differences regarding codes, both 'orthodox' and 'relaxed' milongas forbid any erotic gesture beyond the limits established by the dance itself. With rare exceptions, dancers leave the room on their own, with a person of their own gender or, in the 'relaxed' milongas, in a group. There is no apparent sign of attraction or desire beyond the dance's embrace, and the couples are renewed in each tanda. It is still rare to see the same couple dancing more than a tanda in the same milonga, as this is considered a sign of mutual attraction. The organizers of 'orthodox' milongas make sure the codes are respected and warn newcomers who may challenge them. They may even expel them if their warnings are not followed. The exclusion of any sign of heterosexual attraction beyond the dance's embrace also characterizes the relaxed milongas, albeit without a direct control from the organizers. During my fieldwork I attended both 'orthodox' and 'relaxed' milongas, first with my teachers and classmates and then with my tango students and fellow teachers. With them I also shared and organized dinners, birthday and Christmas parties; paid and received visits; attended shows; looked for places to offer tango lessons and participated in negotiations with their owners and administrators; travelled on long weekends and trained new teaching assistants and teachers. Because of the secrecy surrounding affairs originating in milongas, the close friendship relations that I established with many of them along the past 12 years were central to gathering data for this article. Although at the beginning of my fieldwork I attended a number of tango classes in diverse styles, I finally opted for remaining in tango milonguero ones -which I took in four different 'schools'. Tango milonguero is a style created by the codification and commodification of the dance of the milongueros as practised in downtown Buenos Aires. It was developed by two female teachers and one male teacher at the beginning of the 1990s: Susana Miller, Ana María Schapira and Cacho Dante. Old milongueros from the city centre had always considered their way of dancing tango impossible to teach collectively, as they had acquired their knowledge in secret practices with more experienced dancers and by constant participation in the milongas. After collaborating with Dante, however, Miller started creating with Ana María Schapira a method to teach the dance. In this way, they appropriated the only recognized knowledge in the tango dance -the man's role -and codified it . As part of the process, they selected elements coming from various movement techniques and they codified the combinations of steps made by many different milongueros. Thus, they created the 'milonguero style' and developed a teaching method. From these milongueros they inherited the concept of 'the light woman' as the ideal dancing partner, and then they spread it among their students. Once this method came into being, the tango milonguero classes in the city -as well as in various European and North American countries -multiplied due to their teachings and that of their collaborators, students and assistants. There were also some old milongueros who, with less success, decided to start their own classes and obtain the financial benefits that these produced. The tango milonguero classes were added to the other pre-existent styles: the tango salón , escénico and Nuevo classes that professional dancers and milongueros from the neighbourhood of Villa Urquiza and its surrounding areas had been delivering in Buenos Aires before the 1990s . Therefore, these classes became part of a growing process of commodification and collectivization in the teaching of the tango dance that had some precedents in previous decades, but never enjoyed the success and growth that this decade would bring. Because of the sustained, direct and frontal contact between the upper part of the torso of both members of the couple, the milonguero style, in general, is still considered by present dancers as the most obviously erotic experience within tango dancing. It is often stated that milongueros created this style in the sweet shops of the city centre in order to 'hold tight' or 'grope' the women they met there -who were far from the family control exercised in the neighbourhood clubs. Following the tradition started by Miller and Schapira, tango milonguero teachers use a vocabulary that makes reference to changes of weight, balances, stretches, flexions, muscles and also torso, hips, legs, knees and feet positions that remain unknown to most dancers who were dancing it socially before the 1990s. In the lessons, the use of technical and anatomical vocabulary somehow blurs the erotic connotations of the close embrace and of the 'chest to chest' body contact that characterize both the milonguero dance of the city centre and the 'milonguero style'. The language, taken from classes in other dance genres like ballet or contemporary dance, makes tango lessons similar in appearance to them. --- Light women dancing tango Older milongueros attending Buenos Aires downtown milongas often criticize and dismiss as potential dance partners women who, intentionally or not, perform moves of their own, or at their own rhythm, as it hinders the development of the choreography the male dancers are creatively improvising. Professors who inherited and teach their art during tango lessons also transmit these expectations regarding the dance of women. So both at traditional milongas and during social tango lessons it is said that women who move without their companion's indication are 'like truckdrivers' . This figure, stereotypically associated with lack of good manners and delicacy, conveys the idea that by leading instead of letting their partners lead, they become masculine. Women who delay their response to the male's moves are said to be 'heavy women', 'pieces of furniture' or 'refrigerators' that hinder the creative freedom of their male companions. Each tango step is initiated with an almost imperceptible movement of the man's torso known as la marca . This movement announces for the woman what his next step will be. Then, ideally she will perform the expected complementary movement at the same time that he performs his step. The more precisely and on time a woman responds to her partner's movements, the less he has to worry about what she is doing. If he knows the music by heart, as is usually the case, he can create a choreography of various movements in advance for the couple to perform, being sure that she will answer as expected or, as tango dancers usually say, that she will 'follow him'. He will then conclude that she is a 'light woman'. Instead, when he has to stop the performance of his ideal choreography and adapt it to his partner's rhythm because she does not respond as fast or precisely as he expected, he will assert that she is a 'heavy woman', or a 'piece of furniture'. The 'heaviness' or 'lightness' of female dancers is therefore independent of their body weight. Tango 'lightness' does not relate, either, to the ethereal quality of movement expected from ballet performers since the 19th century. Thus, in the context of tango lessons and milongas women's 'lightness' is a necessary complement to the male's creative freedom. Creativity, improvisation and the ability to surprise their dancing partners are central attributes of the esteemed male dance. Tango dancing in downtown milongas allows for advancing and retreating, stopping and going around, left and right displacements, sudden stops and rhythm changes, doubling beats and skipping beats, changing from following the rhythm of the bandoneón to the melody of the violin, initiating advances that will end up in backward movements, long walks and long pauses. The ways milongueros move around the dance floor resonate with the ways Buenos Aires dwellers drive and once played football . It is full of amagues, minimal movements indicating they will go in one direction to suddenly change to the opposite. So each milonguero dances adapting his own movements to a moving scene that changes all the time in unpredictable ways, while each milonguera dances adapting to a partner who moves all the time in unpredictable ways. In the film Tango Bayle Nuestro, director Jorge Zanada includes a scene in which a group of milongueros watch the performance of a group of tango stage dancers who are being filmed for an English television programme. At the end, almost unanimously, the milongueros remark with disgust that the dances were previously choreographed. Tango, according to their own way of describing their art, should always be improvised, spontaneous, never planned or repeated. Every tango danced by a milonguero is, in their own eyes, a transient creation, freely invented at every instant, product of the moment's inspiration. For them, like De Certeau's tactics, tango dancing neither follows nor leaves footprints. Creatively solving the many problems they may encounter on the dance floor -a couple that suddenly stops in front of them, another one that quickly approaches from the side, a third one pushing from behind -while they interpret with their bodies the variation played by a violin, the sudden dash of a piano, or the slowing movement of a bandoneón at the end of a tango, milongueros see themselves as free creators who improvise their dance following the moment's inspiration, and are proud of this ability. To experience this freedom, what they call 'a light woman' as their dance partner is considered a necessary requirement. A light partner is, for these social tango dancers, one that 'relaxes and allows herself to be carried' through the floor, never responds to their cues with an impromptu movement, never stops unexpectedly, never generates, by herself, a movement that will surprise them. With a 'light woman', the milonguero can feel that he 'freely creates the dance in her body', as a young and famed professor likes to say. A light woman is 'a Stradivarius', as legendary dancer Juan Carlos Copes calls his equally renowned partner María Nieves, or 'a Ferrari' allowing a swift ride through the dance floor, in the words of a student of the school where I teach. As a canvas that absorbs paint in a constant fashion, or a violin that produces exactly the desired sound, 'the light woman' always moves as expected in response to the improvised, imaginative, inspired move of her partner, thus contributing to his creative freedom. --- Feminine images in Latin America Following the analysis of Emily Stevens in Female and Male in Latin America a considerable number of Anglophone studies addressed the issue of the influence of Catholic feminine images, especially the Virgin Mary, on the self-image of Latin American women. While subsequent studies criticized the passivity that the author attributed to them in her study of Marianismo as a Latin American ideology of women's spiritual superiority, the prevalence of images of Catholic origin continues to dominate this body of literature . Thus studies continue to discuss how women interpret these images , what moral lessons they derive from them and how they use them to maintain valued self-concepts or legitimize their choices . Some studies on masculinities in Latin America have also analysed how men label women with categories derived from images such as saints and virgins . Argentine social scientists studying the field of music have departed from the assumption that feminine imagery necessarily derives from the Catholic Church, but not always from the one that assumes this imagery as inspired mostly or exclusively by the behaviour of women . Thus Martín suggests that images of women in the lyrics of a local genre called cumbia villera may indicate transformations in the female role in intimate relations. Following her lead Vila and Semán argue that the frequent labelling of women as whores and of performing explicitly sexual practices in this genre is a reaction to what the authors call 'the sexual activation of women'. Images depicting women seeking sex are a result of the infantile attitude with which men, dominant but besieged, contemplate the lively sexual activity of some women and fantasize that most of them exhibit this behaviour. Accordingly, the masculine voice heard in cumbia villera lyrics no longer describes women as passive objects; instead, it describes them as powerful. Thus, for Vila and Semán, through 'name calling lyrics' , men try to keep control of women's sexuality when they have already lost it. Feminine images are, for these authors, a masculine reaction to changes in women's practices. These changes are, in turn, indicative of a subversion of heterosexual power relations. Archetti has made a similar argument in relation to the frequent mention of the figure of the milonguita in tango lyrics during the first decades of the 20th century. Around 1920, the daughters of recently arrived European immigrants ventured to public entertainment places that constituted a new presence in downtown Buenos Aires. Dancing tangos in public places, sharing embraces that were neither approved nor controlled by their families and establishing heterosexual relations with men on the basis of personal preferences, these young women inspired tango composers, themselves descendants of immigrants, to create an impressive amount of lyrics portraying female figures they named milonguitas. For Archetti, these tangos were a masculine reaction to women's escaping the close control exercised by their parents at a time when the vision of a non-accompanied woman in a public place constituted a rare occurrence -and to watch one embracing a man was downright inconceivable. Influenced as they were by romantic literature, tango writers did not call them sluts, but their lyrics attributed to those milonguitas characteristics that bear a strong resemblance to those associated with whores. These include their demeanour -they were always depicted as laughing and drinking champagne; an interest in acquiring luxury symbols as a result of their relationships with men; and a marked disposition to abandon true love in the pursuit of a good catch. Thus both Archetti's and Vila and Semán's analyses argue that lyrics either labelling women as sluts or attributing them whore-like traits are male reactions to novelties in female behaviour: women dancing in new, more sensual ways and establishing heterosexual relations in ways that were until then considered improper . Other authors have analysed stereotypical feminine figures in tango lyrics as controlling images, which have pedagogical effects on women . Armus interpreted a widespread tango image, women dying from tuberculosis as the feminization of illness. For the author this feminization served moralizing purposes. Along the same lines, Campodónico and Gil Lozano have analysed the contrasting figures of the milonguita and the mother both in tango lyrics and films as constructions also having moralizing effects over women. Some studies on gender images in Latin America have departed from the assumption that these are reactions to changes in women's behaviour and primarily affect them. Authors have made efforts to show that recurrent female and male images reflect not only gender but also class and ethnic inequalities . Savigliano takes a similar stance, asserting that in tango gender is a metaphor condensing racial, class and colonial inequities. But in other sections of the same book, she also considers that tango lyrics reflect the real behaviour of women, describing their abilities to subvert gender relations both while dancing and loving a man. The author states that milonguitas are apparently docile women but they finally leave men who oppress them . Based on 12 years of participant observation in tango lessons and milongas in downtown Buenos Aires, I take a different stance, and analyse the image of the 'light women' as an allegorical rendition of heterosexual relationships originating in milongas. Etymologically allegory means 'other speech' , it conveys an open declamatory act that contains another layer of meaning. In western painting and sculpture the female form has been very frequently used to allegorically represent a variety of concepts such as freedom, justice, revolution, youth and death among many others . Rarely used by sociologists, the concept of allegory has sometimes been employed by cultural critics to analyse colonial relationships. Jenny Sharpe has used it to analyse narratives of rape written by colonizers in India. She argues that the images of the white innocent woman and the Indian rapist are allegories of the colonizer-colonized relationship elaborated by the former when the continuity of the empire was threatened by rebellion. In turn, Rebecca Blevins Faery has noticed that images of naked women and virgins were produced by colonizers and conquerors as allegories of America to depict it as a vacant and uncultivated land. With a slight twist on this literature, I consider the image of 'the light woman' as emerging from the immediate context where it is exchanged, showing the resonance of the image of the 'light woman' in dance and the characteristics of heterosexual relations in tango social dance events. --- Heterosexual relations in Buenos Aires downtown milongas --- The dance floor as a place of transient and generalized eroticism Tango is almost universally defined as an erotic, sensual or passionate dance by those who watch it and by those who advertise it for international consumption . However, the way social dancers and teachers refer to it in Buenos Aires does not match this description. Old milongueros publicly state that they dance for the pure joy that it brings them, and thus hide from novices any possible erotic attraction -a link that, however, they will acknowledge in a whisper when they are among friends. The majority of those who started dancing from the 1980s onwards clearly distinguish between the pleasure that the dance produces from that associated with sexuality or with any other emotional link. They frequently state that there are partners they like dancing with because of the way they dance; that there are others who dance badly, but feel sexually attracted to; and that there are partners with whom they enjoy dancing, because they are funny or 'have a good vibe' or 'are friends'. This feeling of friendship is particularly encouraged in tango classes. Finally, whatever pleasure one may derive from a particular partner, most milongueros consider this should only be articulated within the couple's inner sphere and never made evident to those around . Thus, even if social tango dancing may seem erotic to the onlooker, this eroticism appears to be evenly distributed between all the dancers with no special liaisons arising between any of them but for a brief moment. Tango eroticism also seems extremely mobile, for couples reshuffle as soon as a musical tanda ends. When tango dancers experience pleasure or erotic attraction during the dance, the signs used to indicate this are usually subtle and only understood by the partner: a touch delayed for a few seconds, a hand that slides upwards to reach the flesh in the neck above the shirt collar or dress cleavage, some unavoidable contact that is a bit longer than the music concedes, or the imperceptible repetition of certain movements that allow more and sustained contact. An example of the expected limits of such signs happened when several advanced students, teachers and assistants of a milonguero tango school were watching the organizer of one of the so-called 'relaxed' milongas at the same place where they taught and learnt once a week. Because of the unusual explicitness of the erotic contact he displayed when dancing with his partner, the women nicknamed him 'The Groper': Lara, Carina, Camila, Andrés and I were looking at 'The Groper' dancing with one of his friends. He invited her to open her legs wide and he put one of his legs between hers, embracing her around her waist and making both bodies go upwards and downwards. Then, he put his hand with his fingers downwards on her bottom and he slid it upwards slowly but surely. We all watched from our chairs with our bodies and heads together, moving from side to side in order not to miss any of the scenes that the column that separated us from the dance floor hid from us. Lara put her hands in her stomach and made a painful face . 'It's like watching porn', Andrés says. 'Yes', answers Carina, 'I don't have a problem with porn movies, but this is like watching one at home, in the living room, with your family.' Martín, another teacher, ironically asks: 'Do you know the name of that guy's dance style? -"Uruguayan Tango", it's the one Walter and Waldo taught in Cha-Cha-Cha', making reference to a TV comedy in which a pair of men parodied tango classes through grotesque performances. We all laughed until the moment Luciano said: 'No, it's called Erect Tango'. This unanimous censure when confronted by visible and openly sensual movement as displayed by the organizer shows the preference for invisible and subtle, tactile and only perceptible to the occasional partner, demonstration of attraction and erotic pleasure during the tango dance. It also explains why, for many milongueros, explicit demonstrations of passion associated with stage tango look ridiculous and become the object of parody and jokes within the context of social tango. As a result of this subtlety, when watching the dance floor in a milonga no special attraction appears to link the members of any given couple. This impression is still more vivid since all couples are reshuffled every set of four pieces of music. When a musical tanda finishes, everybody goes back to their seats, searching for a new transient dance partner as soon as the music resumes. Thus, no special liaisons appear to arise between any of them. Tango eroticism also seems extremely mobile, since those who seemed attracted to each other moments before are likely to be showing identical interest in new dance partners as soon as they begin to dance with them. However, every milonguero or milonguera recognizes that, occasionally, they have experienced an exceptional 'connection' during the dance that developed into erotic encounters beyond the confines of the milonga. When sexual encounters recur, milongueros refer to the relationship as 'a story' -a native term I will use from now on. There are, as we will see, hidden liaisons involving sex between those attending the milongas and that remain unnoticed by others or, when noticed, unspoken. In the common milonguero language, this concealment is reinforced by statements such as 'you go to the milonga to dance' and 'I choose my partners because of their dance', though they may add afterwards in a whisper 'and, if I also like them, all the better'. --- Light heterosexual relations Everybody knows that there is flirting in a milonga, but how you hit on someone is still a mystery to me. Since the rules of conduct in milongas differ markedly from those applying in the world outside, los códigos de la milonga -the milonga codes -keep the limit between those in the know and those who do not know, between old-timers and newcomers, keeping the latter away from the dance floor and making their initiation into a very gradual process. Although everybody in the milonga knows that there are códigos, and milongueros like to convey some of them to newcomers, they never mention rules organizing sexual encounters among them. The prospective milonguero has to discover this implicit aspect of the code by observation, spending time in milongas, catching words and attitudes and, especially, by listening to the warnings and secret anecdotes of close friends. As a general rule, in the milongas of the city centre that take place from Sunday to Friday, dates are arranged without any public display of affection or attraction beyond the tango's embrace. For unknowing watchers and newcomers, milongas appear as a space clearly divided between a dance floor where the dance evolves in all its erotic flair and its surroundings, where eroticism is completely absent. Thus, most of the 'stories' born in the milongas remain publicly hidden, except to the most intimate friends of those involved who, in turn, will never acknowledge knowing, even while interacting with the participant in the story who is not the one that revealed it. Only a fraction of those attending a milonga regularly get to know some of the 'stories' through whispers or gossip that, on the other hand, is always suspected of being made up. Since further evidence is very difficult to gather, the evaluation of the gossip's veracity will depend on the intimacy between the listener, the teller and those involved. A dramatic illustration of this invisibility regarding relationships took place at a milonga where I used to go weekly and involved a woman who used to sit next to my table. We had always greeted with a hug and we used to exchange a few words during the night. On one of these occasions, she started crying. When I asked her what had happened, she told me that one of the regulars in the milonga, whom I also knew, had died. She then told me that she had had a 'story' with him for over two years, but had never told anyone because he did not want others to know. Another friend then confirmed this revelation, while adding that the man in question had forbidden her to tell because he was 'a womanizer' and did not want other women to know about their affair. Throughout the night, the woman revealed her relationship with her deceased partner to each person who greeted her with a hug after seeing her crying, and she remained sitting at her usual table and without concealing her grief. Without such drama, there were several occasions on which I got to know, by chance, about 'stories' that involved friends and acquaintances and even students and classmates I used to meet regularly. Either I would see them together in the street after they had parted their separate ways; or, after a fight, one of the partners would reveal the 'story'; or a close friend could tell in revenge, after an argument with a member of the couple. I have also helped cover up 'stories' of close friends by leaving the milonga with them so that they could meet their partner later outside . The occasional couple who has met during the context of a dance will hardly ever leave the room together or at the same time. On the contrary, it will be during the initial musical phrases that start each tango that the date will be set. Then, they will return to their respective tables and they will meet again in another place, at another time, far from the other participants' gaze. These meetings can either be set for the same night, at a specified time, establishing who leaves the milonga first and in which bar or street corner nearby they will meet, or it may be set for a different night. If it is not possible to agree on a date and time, telephone numbers will be exchanged in the corridors, in discreet and coordinated visits to the toilets, or at the door, in the entrance hall, on the balconies or when going out to smoke or take fresh air, always depending on the hidden and empty spaces that the milonga may offer. In milongas attended by younger people, where social gatherings outside the dance floor are more common, there is another unwritten code that enables the possibility to coordinate secret dates: friends will never approach a man and a woman when they are speaking privately, unless both of them belong to a group of intimate friends, classmates generally, and they are certain that there is no possible 'story' between them -though, as I have realized, sometimes that certainty may not match the reality of the situation. I remember the shock of a female tango student, who worked as producer of a rock band and had just started attending milongas, when her teacher got angry with her because she had approached him while he was talking to another woman in the entrance hall. 'I just can't understand this', she said in disbelief, 'In the rock environment these things don't happen.' On the other hand, 'stories' do not have visible consequences in the way their main protagonists will interact in future milongas, and thus remain invisible. They may dance together in the future or not; if she ignores the male's cabeceo or if he disregards her enticing stare. If they do dance again, they may arrange a new date or not. If one of them suggests meeting on the same night, the other can always make up an excuse: 'I have to work early', 'I'm knackered' or 'I have to take my mum to the hospital tomorrow'. This excuse may allow a secret meeting with another person from the same milonga on that very night. Thus, secrecy allows milongueros and milongueras to be involved in several 'stories' at the same time and to choose the person they want to 'leave' with on each night. It also allows them to try and start other affairs with new partners. As one of them once illustrated: 'If I let other people know that I have a "story" with a woman in a milonga, I am stupid, because I am missing out on pulling all the other ones.' In general, the few women who publicly proclaim their 'stories', either verbally or in gestures in the milonga confines, are either described as 'clumsy' or 'crazy'. Moreover, those involved will spread the idea that 'they imagine things'. On the other hand, those men who show their attraction towards a woman through gestures are called 'slimy'. This is more common when they are newcomers. Generally speaking, most people who attend milongas learn quickly the code of secrecy and respect it. However, in private conversations, women complain more often than men about keeping their 'stories' secret, especially when they continue over time. Though both men and women sometimes exercise the right to this free sexual circulation guaranteed by secrecy, it is -as far as I know -much more common for men than for women to be involved in several long 'stories' developing at the same time. This practice is not absent among women, but especially those above 30 tend to finish a long 'story' before starting a new one. Since women more frequently complain about events regarding 'stories' one could argue that milongueros are more suited to this way of sexually relating than milongueras. But it can also be the case that these complaints reflect more closely the gender distribution of what can and must be said about the number and variety of one's own sexual experiences than that of actual experiences. To be sure during my fieldwork I heard men loudly bragging about the richness and variety of their sexual encounters in front of their friends while women just whispered about the same subject in the ears of a very close friend. I must point out that though the usual endings of 'stories' originated in the milongas are to give birth to new ones, a few of them become official and develop into public relationships. However, only some of these relationships imply giving up the emergence of new secret 'stories', especially if one or both members continue dancing in the milongas. The secrecy code allows couples to initiate new relationships and set dates without being noticed, even if their partners are present in the same room. Occasionally, a new 'serious' relationship may mean the abandonment of those milongas where they used to interact and so, the new couple will start going dancing together in other places or they may stop dancing altogether. The secrecy that milonga practices impose on the erotic relationships that take place in its realm allows for free sexual circulation and the development of multiple 'stories' that can happen at the same time and that may or may not be resumed during any given night, depending on chance and the will of the moment. --- The unbearable heaviness of women who fall in love The most often-stated reason by my milonguero friends to finish a liaison that started in a milonga is that 'she fell in love and started getting heavy'. As I heard this reason time and again as an excuse for the termination of several of their involvements -or of their friends' -I finally asked one of them: 'but don't you ever fall in love?' He answered: 'I might fall in love, but I never tell them, because if I do they start getting heavy.' Men going to milongas, therefore, seem to establish a close connection between falling in love and female 'heaviness'. The diagnosis of falling in love made by the men is not always acknowledged by the women involved. On more than one occasion I have learned of women who were very much surprised to learn that they were being dumped because 'they were falling in love'. I have heard male milongueros diagnose infatuation and foresee increasing 'heaviness' of the women they were seeing in the following circumstances: • When they found out that a woman had told other women in the milonga that they had 'a story' with them. • When a woman asked that they go together to a milonga. • When a woman demanded that the man not leave the milonga with another woman if she was present. • When a woman engaged in public demonstrations of affect in the context of a milonga. • When a woman, after a period of not seeing them, said she missed them. • When a woman called asking to see them or to chat. The first three of these behaviorus entail a breach of the implicit code of secrecy that regulates relationships in the milongas, the last two involve an attempt, on the part of the woman, to have a say in the frequency of encounters and in the development of an affective link. The limits of specific milongueros are variable, and some may take certain behaviours listed above and not others as proof of unwarranted female love. Many regular milongueros have told me that when they see a woman repeatedly they consciously deploy tactics to prevent their falling in love. These vary from individual to individual. Some avoid any word that may suggest a romantic relationship: 'miss', 'love', 'want', even 'like' are forbidden words that they force themselves not to utter during sex. Some even confided that they do not tell a woman that she is pretty, to prevent her from falling in love. Other tactics include abstaining of meeting in a public place -even if it is outside the milongas -so that the woman does not think they want to officially acknowledge their 'story'; or not going out on Saturday nights, which would also give the relationship unwanted recognition and perhaps provide fuel for love. Interestingly, all these tactics construct female sentiments and love as a consequence of the males' actions. I have heard them express their surprise when women fell in love 'without them doing anything' to warrant it. For these men, frequent sexual encounters over long periods of time with women they met in the milongas do not seem to lead to the development of sentimental feelings. Romantic words, public demonstrations of affection, seeing each other on Saturdays or being introduced to their friends, however, are considered possible causes for women to fall in love and thus turn into 'heavy' ones. I must emphasize that I have rarely heard women complain because men they were seeing had fallen in love. Quite to the contrary, a frequent complaint is that they could only have one-night stands with milongueros. The male version is that they only do this if women become 'heavy' immediately afterwards -for example, if they call them during the week to chat or ask how they are doing. If not, they are willing to repeat the experience each time there is no other more desirable choice at hand. This may happen for an unlimited amount of time, as long as the woman shows no signs of 'falling in love'. Foreign women, known to social dancers as las extranjeras, are an interesting exception to the taboo of romance in the milongas. As one of my milonguero friends put itmuch to the approval of others present -'a foreigner can become your girlfriend because she is leaving anyway'. Milongueros often explain that they can treat foreign women differently, express their love to them or show themselves together publicly because their eventual and inevitable departure ensures the fleetingness of the relationship. Foreign women are seldom considered 'heavy' and are allowed to fall in love, as long as they do not plan to put down roots in the city: their impending departure makes them extra-light women. An identical exception holds for tourist women in relation to their ability to dance. Differently from Argentine women, blond light-eyed or very dark-skinned foreign women are frequently invited to dance as soon as they arrive to the milonga, independently of their dancing skills . However, if they plan to stay in the city they lose their 'foreigner' identity and they are judged as harshly as their porteñas counterparts . Their 'falling in love' is avoided by the same tactics already discussed. The exceptional lightness attributed to foreign women show the intimate relation linking the attribution of that trait to the fleetingness of 'stories'. Although, as our analysis of 'heavy women' has shown this fleetingness should not be considered identical to briefness A 'story' continues to be light even when composed of many episodes as long as recurrent occurrences do not imply any obligation to be repeated at any given moment in the future. --- Conclusions The secrecy that milonga practices impose on the erotic relationships that originate within the milonga allows the free sexual circulation and the development of multiple 'stories' that can happen at the same time and that may or may not be resumed during any given night, depending on chance and the will of the moment. They are, as 'light women' in dance, relationships that allow for the creative freedom of the participants and do not impose any obligations on them. They are characterized by secrecy, detachment, multiplicity and freedom. Conversely, public recognition, affective attachment and the obligation to meet at any given moment are enemies of this free circulation. They turn heterosexual relations 'heavy'. But in the milonga context lightness and heaviness are not depicted as attributes of 'stories' and 'love relationships': they become feminine images. They acquire an anonymous female form, as did justice or liberty when statues were made to represent them. Thus the 'light woman' can be understood as an allegory of those ways of heterosexually relating that milongueros call 'stories' while the one that becomes 'heavy' allegorizes monogamous affective attachments. In this process of gendered allegorization the role of men in heterosexual relationships is erased. One could wonder if a similar process is not involved in the making of the figure of the slut in cumbia villera or of milonguitas in early tango lyrics. Couldn't they be understood as allegories of heterosexual relations established in the social contexts where these genres are or were publicly performed? In order to answer the question much more information of how relationships develop beyond the limits of dance halls would be needed. No doubt when attributed to specific persons, images of women can become a social identity . Such attributions have observable consequences for women and they develop diverse tactics to deal with them. They can sometimes choose to contextually identify with them, embodying them through kinetic behaviour, clothing and demeanour , reject them or even use them to their advantage . But here I wished to emphasize that in the context of the milongas the image of 'the light woman' gives a feminine form to a prevailing way of heterosexually relating that engages both men and women. être envisagées comme la féminisation et l'allégorie des relations impliquant les hommes et les femmes dans les contextes mêmes où ces images circulent. --- Mots-clés Cours de danse, images des femmes, milonga, sexes, tango --- Resumen Tratando de entender la auto imagen de las mujeres latinoamericanas, muchos estudios anglófonos han recurrido a la influencia de las imágenes femeninas católicas, especialmente a la Virgen María. Aunque estudios recientes han criticado este énfasis en el Marianismo por proporcionar una imagen pasiva de las mujeres, la preeminencia de las imágenes de género de origen católico continúa dominando las investigaciones sobre el continente. Basado en doce años de observación participante en clases de tango milonguero -primero como alumna del 'papel de la mujer', luego del 'papel del hombre' y finalmente como profesora -y en milongas nocturnas en el centro de Buenos Aires, asumo una postura diferente sobre esta cuestión, analizando la imagen de la 'mujer liviana' como una versión alegórica de las relaciones heterosexuales originadas en los salones de baile de tango. Este marco analítico está inspirado por las críticas culturales que observan las figuras femeninas en otros lugares como la alegorización y feminización de las relaciones de colonizadores y colonizados, y de conquistadores con las tierras sin cultivar. Sin embargo, en contrapunto con esta literatura, este artículo considera la imagen estudiada como emergiendo del contexto inmediato donde es intercambiado. Al mostrar la resonancia de la imagen de la mujer liviana, y las características de las relaciones heterosexuales en los bailes y eventos de tango social, este artículo sugiere la posibilidad que las imágenes de género serían entendidas como la feminización y alegorización de las relaciones que involucran hombres y mujeres en los propios contextos donde dichas imágenes circulan. --- Palabras clave Clases de danza, género, imágenes de género, milonga, tango --- Author biography Maria Julia Carozzi got her PhD in Anthropology from the University of California Los Angeles. She is a tenured Professor of Ritual and Performance at the Institute of Higher Social Studies, National University of San Martin and tenured researcher at the National Council for Scientific and Technological Research, Argentina . Her latest book is entitled Las palabras y los pasos: Etnografías de la danza en la ciudad . --- Résumé Dans leurs tentatives de comprendre l'image de soi des femmes latino-américaines, beaucoup d'études anglophones se sont tournés vers l'influence des images féminines catholiques, en particulier de la Vierge Marie. Malgré des études plus récentes qui ont critiqué l'accent mis sur le marianisme dans la formation de cette image passive de la femme, la place importante accordée aux images féminines d'origine catholique continue à dominer la recherche sur le continent. Basée sur douze années d'observation participante menée lors de leçons de tango milonguero -afin d'étudier en premier lieu le 'rôle de la femme', puis le 'rôle de l'homme' et enfin celui du professeur -et lors de séances nocturnes de milonga dans le centre de Buenos Aires, cet article propose une approche différente de la question, celle de l'analyse de l'image des 'femmes légères' comme une interprétation allégorique des relations hétérosexuelles dans les salles de danse de tango. Cette grille d'analyse est inspirée des critiques culturelles qui considèrent d'autres figures féminines comme le produit de l'allégorie et la féminisation des relations entre les colonisateurs et les colonisés et entre les conquérants et les terres incultes. Cependant, à la différence de cette littérature, cet article analyse cette image directement dans le contexte immédiat où elle est échangée. En montrant les relations entre l'image de la 'femme légère' et les relations hétérosexuelles dans les endroits où l'on danse le tango, cet article suggère que ces images féminines puissent

Trying to understand the self-image of Latin American women, many Anglophone studies have turned to the influence of Catholic feminine images, especially the Virgin Mary. Albeit later studies have criticized this emphasis on Marianismo for providing a passive image of women, the prevalence of gender images of Catholic origin continues to dominate research on the continent. Based on 12 years of participant observation in tango milonguero lessons and at night-time milongas in downtown Buenos Aires the author take a different stance on the issue, analysing the image of the 'light woman' (la mujer liviana) as an allegorical rendition of heterosexual relationships originating in tango dance halls. This analytical frame is inspired by cultural critiques viewing feminine figures elsewhere as the product of the allegorization and feminization of the relationships of colonizers with the colonized and conquerors with uncultivated lands. However, at odds with this literature, it considers the image under study as emerging from the immediate context where it is exchanged. By showing the resonance of the image of the 'light woman' and the characteristics of heterosexual relations in tango social dance events, this article suggests the possibility that gender images may be understood as the feminization and allegorization of relations involving men and women in the very contexts where such images circulate.

Background An extensive amount of research has consistently found associations between childhood socioeconomic status and mental health functioning [1][2][3], with marital conflict and parenting practices seeming to mediate these associations. SES is a construct that consists of multiple dimensions of social position [4,5]. Previous related empirical and theoretical research has focused on economic and educational aspects as SES indicators. Family income has been associated with children's developmental outcomes, as have parental educational levels [6][7][8][9][10][11][12]. However, despite the many studies conducted in this area, few have simultaneously investigated the influence of family income and maternal and paternal education levels as predictors in the relationships between SES, family processes , and child mental health functioning. Additionally, despite extensive studies concerning the relationships between SES, family processes, and child mental health functioning, most have only minimally considered the effects of the positive dimensions of marital conflict and parenting practices , rather than the negative dimensions thereof , as mediators in the link between SES and child mental health functioning [7,[13][14][15][16]. Moreover, a limitation of previous empirical work concerning these associations is that these studies focused on negative developmental outcomes [17,18]. Further studies examining positive dimensions of child mental health functioning, especially the issue of social competence, are needed. Social competence, which is defined as an individual's ability to act in a socially appropriate manner [19,20], has received comparatively less attention as a mediator in the link between SES, family processes, and child behavioral adjustment, despite preliminary evidence suggesting it may be an important indicator. When considering the complex relationships between these variables, it is important to consider independent associations, while controlling for other variables. However, previous studies have primarily examined individual relationships between different types of SES, marital conflict, and parenting practices, as well as child social competence and behavioral adjustment, without considering these associations in a comprehensive model. Therefore, this study examined mediators of the associations between SES and children's functioning in greater detail. Specifically, destructive and constructive marital conflict, negative and positive parenting practices, and child social skills were investigated as mediators in the associations between SES indicators, including family income and parental education levels, and children's internalizing and externalizing behaviors in a unified model. Regarding social skills, we especially focused on the mediating role of social competence in the relationships between family factors and child behavioral problems, from preschool to the first grade. --- Socioeconomic status and child adjustment Research in the past decade has shown that SES is an important contextual factor that strongly predicts child outcomes [1][2][3]. Extensive research has shown that SES affects the well-being and development of children, including their internalizing and externalizing symptoms, as well as their cognitive and language development [1,3,[21][22][23][24][25][26][27]. It has been well documented that economic problems, such as low income and financial instability, adversely influence inter-parental and parent/child interactions, which in turn are related to a range of harmful outcomes for child development [28]. Studies have shown that economic problems are associated with destructive parental interactions that predict increased domestic problems and lower levels of marital quality. Furthermore, it has also been shown that economic problems place children at an increased risk of exposure to family conflict [7,[29][30][31][32]. Economic problems are also predictors of negative parenting, including lack of warmth and involvement, parental harshness, and authoritarian parenting methods [28,[33][34][35][36]. The family stress model , which was proposed by Conger et al., explains the relationships among SES, marital conflict, and parenting style, while also providing solid evidence for the negative effects of family economic problems on both parents and children [15,37]. The FSM proposes that economic hardship predicts economic pressure, which in turn exacerbates emotional distress for both parents [37]. In turn, parental emotional distress has a direct, negative impact on the parents' relationships with each other, as indicated by conflict. This conflict then spills over into parent/child relationships, in the form of negative parenting, resulting in harsh, uninvolved, and/ or inconsistent child-rearing practices; these parenting styles are associated with an increase in negative outcomes for children [29,[37][38][39]. Educational status and economic aspects are typical quantitative SES indicators [4,5]. Many previous studies have focused on the educational aspects of SES in the relationship between SES and child development, with parental educational levels being associated with child developmental outcomes [1, 2, 10-12, 25, 26]. However, despite the many studies completed in this area, few have simultaneously investigated the influence of multiple components of SES, including family income, and maternal and paternal education levels, as predictors in the relationships among SES, family processes, and child mental health functioning. In several studies that include both educational and economic aspects of SES indicators, educational status has often either previously been used as a control variable, or it has been combined with income in the construction of an overall index of SES indicators [6,7]. Furthermore, a limitation of previous empirical work on the FSM is that studies have also focused exclusively on the economic aspect of SES in the relationship between SES and family processes, dedicating little research attention to the educational aspects of SES [28]. It is well known that education is an important predictor of family income across the life course [40]. Therefore, it may be reasonable to expect the influence of educational status on parental interactions and parent/ child interactions to be indirect and mediated by economic well-being. Education is an important component of SES that helps identify a social class or position, and has been linked to individual competence [4]. Higher education is likely to enhance various individual skills for competent functioning, such as problem-solving skills, cognitive skills, and capacity to cope with change. People with higher levels of education tend to be able to solve problems that are more complex and perform jobs with more autonomy and creativity [41][42][43][44]. Moreover, educational achievement provides persons with more employment opportunities, enhances their ability to make significant contributions to their fields, and demonstrates significant positive associations with occupational prestige and income [40,[45][46][47]. Furthermore, according to human capital theory, the education level of an individual's spouse also helps accumulate human capital and has an important impact on economic outcomes [48,49]. For example, a spouse with a higher education might provide constructive advice and information that can affect career and decision making in the family, such as consumption, fertility, and where to live [50][51][52]. Additionally, spouses are likely to affect each other through values, attitudes, and other abilities associated with education. Many studies have revealed common findings that the education level of an individual's spouse is positively correlated with the individual's earnings. Especially, numerous studies have suggested that a wife's education affects her husband's earnings [51][52][53][54][55][56], and vice versa. Additionally, other studies have shown that an individual's earnings are positively correlated with their spouse's education level [53,57]. This correlation might be due to marital matching, as individuals that are more productive are more likely to marry better-educated individuals. However, despite the fact that parental education levels strongly interact with income, education levels and economic conditions could have different effects on family processes and child mental health functioning, possibly acting through different pathways. Regarding the relationship between educational level and marital relationship, higher education is likely to help parents to strengthen their communication and analytical skills, allowing for more effective problem solving between parents [44,50,58]. Moreover, higher education is also likely to enhance self-control and coping mechanisms of parents, possibly increasing the positive association between education and psychological well-being [58]. Consequently, parental education levels might positively affect marital relationship through parental psychological well-being [44,[59][60][61]. A large amount of evidence for the beneficial nature of education on marriage exist, as studies have demonstrated a negative relationship between parental educational levels and marital conflict [62], a positive association between educational attainment and greater marital satisfaction [30,63], and higher levels of educational attainment are associated with greater marital stability [64,65]. In addition, previous research has suggested that parental education is the strongest and most important predictor of parenting behavior [66]. Regarding the relationship between educational level and parent/ child interactions, higher education is likely to promote the ability to process information, and enable parents to acquire more knowledge and skills about childrearing and child development, allowing parents with higher education to use more effective strategies for childrearing [66][67][68]. Moreover, as mentioned above, a higher level of education is likely to boost parental psychological wellbeing, which, in turn, could positively influence parenting style [69][70][71]. Many studies found that higher maternal education levels are associated with more supportive parenting [72,73], which is also associated with positive cognitive, behavioral, emotional, and physical child outcomes [74][75][76][77]. While few studies have investigated the influence of paternal education levels on fathers' involvement in childrearing, some studies have found paternal education levels to be somewhat associated with parent/ child interactions. For example, several studies revealed that fathers with higher educational attainment tend to be more involved, show more positive engagement, and be more accessible to their children than fathers with a lower education level [78][79][80]. However, other studies have found little association between paternal educational attainment and fathers' involvement, after controlling for factors such as family income and maternal education level [6][7][8][9]. As there are conflicting results in the literature regarding the influence of paternal education level on parental involvement, it is possible that parental education levels may influence parenting attitudes directly, or they may do so indirectly through family economic factors or other SES indicators. Given this information, we are unable to form strong expectations regarding the possible pathways of how both maternal and paternal education levels may influence childhood mental health problems. When considering the complex relationships in the above-mentioned variables, it is important to consider independent associations, while controlling for other SES variables. However, few previous studies have primarily examined individual relationships between SES, including family income and parental educational levels, interparental interactions, parent/child interactions, and/ or child mental health functioning, taking into account associations in a comprehensive model. Therefore, investigations into SES, including family income and parental educational levels, are needed to clarify how each SES indicator flows through the family processes to influence child development. Studying individual markers of SES, including family income and maternal and paternal education, enables us to study the unique and combined contributions of family income and parental education towards family functioning and child adjustment. --- Family processes and child adjustment As mentioned earlier, the FSM has shown that economic hardship predicts greater economic pressure, in turn exacerbating emotional distress among parents, which then negatively affects their relationship with each other, as indicated by parental relationship conflict [29,39]. This marital conflict spills over into parent/child relationships, which are characterized by more hostile, harsh, emotionally neglectful parenting, and less warmth. These types of relationships are associated with more negative outcomes in childhood and adulthood [7,15,16]. The "spillover hypothesis" has been proposed to explain this relationship between marital conflict and child outcomes. According to this hypothesis, the negativity and positivity experienced in the inter-parental relationship transfer to the parent/child relationship, affecting child outcomes [17,18,[81][82][83]. The hypothesis further posits that destructive marital conflict, such as verbal and physical aggression, requires excessive energy that makes parents less emotionally available and less sensitive to the needs of their children. The negative interactions "spill over" into the parent/child relationship, resulting in an increase in negative parenting practices, such as poor monitoring, inconsistency, and harsh discipline. In contrast, constructive marital conflict, such as satisfaction, support, and positive interaction, spills over into the parent/child relationship, which is characterized by increased availability to meet children's needs, and results in more positive parenting practices, such as involvement and praise. Moreover, several studies examining the effects of conflict on children's emotional and behavioral outcomes, have also demonstrated ways of categorizing conflict into destructive and constructive marital conflict [84][85][86][87][88]. These studies suggest that destructive marital conflict make children more vulnerable to developing adjustment problems including aggression, conduct disorders, anxiety, and depressive symptomatology. Conversely, these studies also suggest that constructive marital conflict, including progress towards the resolution of the conflicts and explanations about how conflicts were resolved, is likely to be beneficial to children, helping them learn effective problem-solving and communication skills. Therefore, the findings illustrate the need to examine marital conflict as a multidimensional construct to understand how conflict affects children. However, despite the extensive research completed in this area, studies have minimally considered the impact of positive dimensions of marital conflict and in turn, parenting practices , rather than negative dimensions , as mediators in the link between SES and child mental health functioning. Previous studies have consistently found that destructive marital conflict fosters negative spillover, resulting in more negative parent/child interactions [18]. Furthermore, a limitation of previous empirical work is that studies have focused exclusively on negative outcomes [17,18]. Further studies examining a positive association between family factors and child mental health functioning, including positive outcomes, have been called for. Therefore, investigations into positive spillover practices are needed to clarify how family functioning affects child development in a comprehensive model. --- Social competence and child adjustment School maladjustment is one of the most prevalent and significant health problems threatening children. Previous studies have suggested that one of the factors related to child maladjustment is a child's inability to adjust socially, as a result of a lack of social competence [89]. Social competence has been broadly defined as effectiveness in social interactions [20]. Social skills are discrete abilities that contribute to social competence [19]. Specifically, these skills have been defined as socially acceptable learned behaviors that enable children to interact effectively and avoid unacceptable responses from others [90]. In short, social competence refers to an individual's overall ability to act in a socially appropriate manner [19], whereas social skills refer to specific and distinct behaviors representing social competence [91]. Social skills are some of the most important accomplishments in childhood. Aspects of social skills, such as cooperation, self-control, and assertion, which were clustered by Gresham and Elliott [90], affect social adaptation in later life. Social skills help children initiate positive peer interactions, which help them learn positive behaviors through peer modeling and provide them with resources, such as support and acceptance [92][93][94][95]. Conversely, children who fail to develop social skills in early developmental phases often display social problems. Children who persistently exhibit deficits in social skills experience both short-and long-term negative consequences, which may often be precursors to more severe social problems later in life [96,97]. Children who lack social skills may experience emotional difficulties, and tend to have trouble interacting with their peers, teachers, and families [97][98][99][100]. Furthermore, social skill deficits frequently demonstrate a negative association with behavioral adjustment [99][100][101][102]. Behavioral adjustment is generally associated with two broad symptom dimensions: internalizing and externalizing behaviors. Internalizing behaviors include worry, anxiety, depression, and somatic complaints; while externalizing behaviors include hyperactivity, inattention, aggression toward peers, and management problems [103][104][105][106][107][108][109][110]. Internalizing and externalizing behaviors consistently influence each other over time, with prior studies showing that internalizing behaviors predict later externalizing behaviors, and vice versa [111][112][113][114][115][116]. Further, there is evidence of co-morbidity with internalizing and externalizing behaviors later in the life course. Social competence predicts internalizing and externalizing behaviors across longer periods in childhood, adolescence, and adulthood. Additionally, lower social competence forecasts higher levels of both internalizing and externalizing problems [99-102, 117, 118]. Children who lack social skills have difficulties in expressing themselves and understanding others, such as sending appropriate social messages and responding to their peers, teachers, and families. They have fewer positive interactions and have more trouble interacting with others. Consequently, these individuals are more prone to be disliked and deemed socially incompetent by others [119]. Therefore, children with social skill deficits are at an elevated risk for social isolation, including anxious solitude and peer rejection. Social isolation is associated with behavioral adjustment. For instance, increased childhood social isolation longitudinally predicts depressive symptoms [120][121][122]. Therefore, early peer difficulties with social skill deficits are predictive of later maladjustment. The cross-sectional and longitudinal associations between social competence deficits and internalizing symptoms have been well documented from preschool to adolescence [123][124][125]. Similarly, several studies suggest childhood peer rejection longitudinally predicts externalizing behaviors, including aggression, conduct disorders involving peers, and other under-controlled behaviors during the school-age years and into adolescence [101,102,126]. However, several social skill abilities among children that are associated with externalizing behaviors, such as abilities in emotion regulation, verbally expressing emotions, and self-regulation of behavior, generally increase with age [127,128]. Therefore, as social skills improve with age, the rates of externalizing problems tend to decrease in comparison to internalizing problems [127][128][129]. Eventually, the failure to develop social skills and successful childhood interpersonal relationships could promote mental health difficulties and both internalizing and externalizing problems over time. Early childhood is a pivotal period for social development. The transition period from early childhood to elementary school first grade is a pivotal period for social development that leads to school readiness. Previous research has indicated that the preschool years are a sensitive period for the acquisition of social skills and related abilities [130][131][132][133][134][135]. Preschool-aged children learn and frequently display various prosocial behaviors [136]. Therefore, this period is an important developmental stage during when children are expected to acquire social skills to prepare them for broader social activity. Social skill deficits in early childhood gradually become permanent over time, are related to poor academic performance, and are predictive of social adjustment problems and serious psychopathology in adolescence. Understanding the factors that influence these developmental processes in early childhood may enable the prevention of later socio-emotional difficulties. There is an extensive body of literature demonstrating that the development of social competence among children is significantly affected by environmental factors in childhood [137][138][139]. For example, family functioning has been shown to predict children's social competence. Positive parenting, such as emotional expressiveness, responsiveness, and support, has been shown to enhance empathy and social functioning in children [140][141][142][143], while negative parenting behavior, such as harsh discipline, emotional neglect, or rejecting behavior, is often associated with lower sociability/ social competence and increased problem behaviors in children [16,25,143]. Many previous studies have also shown that destructive marital conflicts negatively affect social competence [144]. This type of marital conflict may put children at risk of developing adjustment problems, including internalizing and externalizing disorders, due to their inability to control their emotions. Moreover, they may learn through these interactions to solve problems through aggressive behavior [18,[145][146][147]. Since research has primarily focused on destructive marital conflict, few studies have investigated constructive marital conflict, which may foster social competence. Constructive marital conflict may also aid in the development of problem-solving, coping, and conflict resolution abilities by teaching children how to effectively communicate with others to solve issues [148][149][150]. Previous studies consistently suggest that destructive conflict increases the risk of adjustment disorders, whereas constructive conflict may positively influence adjustment. Despite the differential effects of destructive and constructive conflict on child development, there is no distinction between these two types of conflict and their implications for social development within the literature. Moreover, even though marital conflict and parenting practices affect social competence [144,151], few studies have addressed the various ways that this may occur within a comprehensive model. As mentioned previously, a limitation of empirical work on the FSM is that studies have focused exclusively on negative outcomes, such as internalizing and externalizing problems [7,15]. This myopic focus leads to a strong need for the examination of positive associations, such as positive developmental outcomes among children . The current study highlights the ways that family processes within the FSM promote desirable child outcomes, specifically focusing on the development of social competence. Various studies have demonstrated the significant effects of family processes on social competence, primarily examining the individual relationships between different types of SES, marital conflict, parenting practices, and child mental health functioning, without considering associations in a comprehensive model. When considering the complex relationships among these variables, it is also important to consider independent associations, while controlling for other variables. For a more detailed exploration of the early protective factors potentially influencing diverse developmental maladjustment, the purpose of this preliminary study was to examine, in greater detail, social competence as a mediator of the relationships between SES, family processes, and children's adjustment. --- Present study Although several studies have demonstrated a significant impact of SES and family processes on general adjustment among children, few have considered the relationship between child behavioral problems and SES, including family economic and parental educational levels, negative and positive aspects of marital conflict and parenting practices, and child social competence, in conjunction with one another. Most prior studies including the FSM have focused little attention on the educational domain of SES or the positive aspects of family functioning and child outcomes. When considering the complex relationships between these variables, it is important to consider independent associations, while controlling for other variables in a comprehensive model. Most studies have examined these complex relationships in a more piecemeal fashion, rarely integrating them into a unified conceptual model. Within the risk and resilience research framework, relational risk or protective factors are thought to make either additive or contingent contributions to adjustment. Based on the observations above, the aim of this study was to clarify the roles of SES , marital conflict , parenting practices , and child social competence and behavioral problems , by analyzing these relationships in a comprehensive model. In the present study, we used longitudinal assessments of children's externalizing and internalizing behaviors to evaluate the hypothesis that SES, marital conflict, and parenting practices predict children's social competence, which is then related to later child adjustment. The mediational model in Fig. 1 was tested to estimate the direct effects of Time 1 SES, marital conflict, and parenting practices on Time 2 behavioral problems, and to examine the indirect effects of T1 variables, through their effects on T1 social competence, on T2 behavioral problems. As a result, our study provides theoretical contributions to the FSM by incorporating additional critical factors . Investigating the role of social competence as a mediating process in the link between relational risks such as SES and later child adjustment will enable important theoretical contributions to the understanding of processes involved in the development of adaptation among children with higher relational risks, and will provide implications for prevention and intervention efforts. We hypothesized the following pathways: SES indicators are, as predictors, differentially associated with family processes and child mental functioning through distinct pathways; both negative and positive aspects of family processes will mediate the relationship between SES and child mental health functioning; and social competence in preschool, which is influenced by multidimensional family factors, will reduce the risk of behavioral problems in the first grade. Hosokawa and Katsura Child Adolesc Psychiatry Ment Health 11:62 --- Methods --- --- Ethics statement The children's parents and teachers were informed of the study's purpose and procedures, and they were made aware that they were not obligated to participate. The teachers provided their written informed consent, and the parents submitted the same on behalf of their children prior to participating in this research. Ethical approval for this study was obtained from Kyoto University's Ethics Committee in Kyoto, Japan . --- Measures All the questions used for the self-developed questionnaire were questions translated into Japanese. --- Predictors Socioeconomic status At T1, SES was defined as information about family income levels, as provided by the parents, and parental education. Parents were asked to report their total yearly family income, their education in years, and their completed education levels by choosing one of the following response options: compulsory education , vocational upper-secondary school/general upper-secondary school , less than 4 years at college/university , and over 4 years at college/university . Each of the SES scores were converted to z scores. --- Mediators Marital conflict At T1, the Quality of Co-parental Communication Scale , a 10-item self-report questionnaire, was used to assess each parent's feelings or behaviors within the context of the co-parenting relationship [120]. This measure is composed of the following two subscales: Co-parental Conflict and Co-parental Support . Items are rated on a 5-point Likert scale ranging from 1 to 5 . The Conflict and Support subscales assess parents' perceptions of the co-parenting relationship. The Conflict subscale measures the negative aspect of the co-parenting relationship, with higher conflict scores indicating more co-parental communication conflict [152]. In the current study, we con- sidered Co-parental Conflict as destructive conflict. Conversely, the Support subscale measures positive aspects of the co-parenting relationship, with higher support scores indicating more supportive co-parental communication [152]. Specifically, the Support subscale measures "general support" including helpfulness, resourcefulness, and cooperation [152], as opposed to the constructive aspects of conflict. However, in the current study, we considered Co-parental Support as constructive marital conflict. The scales have adequate internal consistency and construct validity [152][153][154]. The internal consistency was .88 and . 74 for Conflict and Support scales, respectively [152]. The current study found internal consistencies of .77 and .86 for the Conflict and Support scales, respectively. Each QCCS total score was converted to a z score. Parenting practice At T1, the Alabama Parenting Questionnaire , a 42-item self-report questionnaire, was used to assess various aspects of parenting behavior [155,156]. The measure is composed of the following five subscales: Poor Monitoring/Supervision, Inconsistent Discipline, Corporal Punishment, Positive Parenting, and Involvement. Items are rated on a 5-point Likert scale ranging from 1 to 5 . Participants self-reported their own parenting behavior. The developers have reported that the measure has adequate internal consistency and construct validity [156]. The internal consistency of the subscales ranges from .46 to .80 [156]. In this study, the subscales' internal consistency ranged from .71 to . 76. In this study, we standardized the separate positive and negative parenting composite scores [157]. Scores on the Poor Monitoring/Supervision, Inconsistent Discipline, and Corporal Punishment subscales of the APQ were combined to form a negative parenting composite score, whereas scores on the Positive Parenting and Involvement subscales were combined to form a Positive Parenting composite score. The Negative Parenting composite score was calculated by converting the Poor Monitoring/ Supervision, Inconsistent Discipline, and Corporal Punishment subscale scores to z scores and then averaging them, with higher scores indicating more negative parenting. Similarly, the Positive Parenting composite score was calculated using the same method for the Positive Parenting and Involvement subscale scores, with higher scores indicating more positive parenting. Child social competence At T1, the Social Skills Questionnaire was used as an index of observer ratings of child social competence. In the current study, the children's teachers evaluated their social skills using this scale. The SSQ is a 24-item measure of children's social competence in relation to "cooperation", "self-control", and "assertion" [158][159][160], as factors affecting social adaptation in later life [90]. These clusters of social behaviors can briefly be characterized as follows: Cooperationbehaviors such as helping others, sharing with a peer, and complying with rules such as sharing and obeying; Self-control-behaviors that emerge in conflict situations, such as responding appropriately to teasing or corrective feedback from an adult; and Assertion-behaviors such as asking others for help/information and responding to others' actions . The SSQ has the following three subscales: Cooperation , Self-control , and Assertion . These factors are based upon, and positively correlated with, the Social Skills Rating System [90], which is one of the most widely used social skills scales and was used in the National Institute of Child Health and Human Development study [161,162]. The SSQ's items are rated on a 3-point scale ranging from 0 to 2 , yielding total scores for cooperation, self-control, and assertiveness. The SSQ has adequate internal consistency and construct validity; the subscales' internal consistency has previously ranged from .91 to .93 [158], with a range from .84 to 94 in the current study. Furthermore, the present study combined total scores for cooperation, self-control, and assertiveness to form a social skills score, with higher scores indicating better social skills. The social skills score was calculated by converting scores on the Cooperation, Self-control, and Assertion subscales to z scores, and then averaging them. --- Criterion variables Child adjustment The Strengths and Difficulties Questionnaire is a 25-item measure of parents' perceptions of their children's prosocial and difficult behaviors, and it is designed to assess general internalizing and externalizing emotional and behavioral problems [163]. In this study, children's mothers evaluated their behavioral adjustment using this scale at both T1 and T2. The measure is composed of the following five subscales: Emotional Symptoms, Conduct Problems, Hyperactivity-Inattention, Peer Problems, and Prosocial Behavior. Items were rated on a 3-point Likert scale ranging from 0 to 2 . The scales' internal consistency and construct validity were reported as adequate [164][165][166]. In this study, the Emotional Symptoms and Peer Problems subscales of the SDQ were combined to form an Internalizing Problems scale , while the Conduct Problems and Hyperactivity-Inattention subscales were combined to form an Externalizing Problem scale , as suggested by Goodman et al. [167], with higher scores indicating more behavioral problems. Each SDQ total score was converted to a z score. --- Procedure To conduct our study, we asked the kindergartens and nursery schools with 50 or more students, in Nagoya city, to participate. As a result, principals of 130 facilities gave us permission to conduct our survey and meet with participating parents. To recruit families at T1, self-reported questionnaires were distributed at the participating facilities to all parents of 5 year olds . Participants received an information sheet and questionnaires on childrearing, in relation to family factors , and child behavioral adjustment . Participants provided written informed consent and agreed to participate. The parents completed the questionnaires at a single time point and returned these to participating facilities in sealed envelopes to prevent teachers from seeing the questionnaires. Then, the teachers evaluated the children's social skills using the SSQ. All sealed envelopes containing questionnaires and SSQ evaluations were returned to the researcher from the respective principals. At T2, 12 months later, participants were contacted again when the children were in the first grade. At T1, the researcher obtained the address of participants, and, at T2, the researcher mailed the participants questionnaires on childrearing in relation to family factors and child behavioral adjustment. Participants who completed the questionnaires returned them to the researcher by mail. Access to the data was restricted to the researchers of the current longitudinal study. --- Data analyses First, prior to developing a model of the relationships among SES, parental relationship, parenting practices, and child social competence and adjustment, correlation analyses were utilized to determine the associations among SES , marital relationship , parenting practices , child social competence , and child adjustment . Second, path analyses were conducted to estimate direct and indirect paths between SES, parental relationship, parenting practices, and child social competence and adjustment. Structural equation modeling analyses were conducted using full information maximum-likelihood estimation in the presence of missing data. The hypothesized model is presented in Fig. 1. In the models, SES was specified as a predictor of the marital relationship , parenting practices , child social competence , and behavioral adjustment . We estimated how family factors and child social competence in preschool influenced the children's behavioral adjustment in the first grade. The model also included T1 behavioral adjustment as control variables; through controlling for initial levels of maladjustment, the model would appropriately address changes in behavioral adjustment. Based on previous findings in the literature, we expected the effect of T1 SES indicators on T2 behavioral adjustment to be mediated by the T1 parental relationship, parenting practices, and social competence. Moreover, we expected an inverse effect between T1 social competence and T2 adjustment. To assess fit, we examined the Comparative Fit Index [168], the Incremental Fit Index [169], and the Root Mean Square Error of Approximation [170]. Good model fit is reflected in CFI and IFI values above .90 [168,169]. Regarding the RMSEA, good fit was represented by a value smaller than .05 and reasonable fit was represented by values ranging from .05 to .08 [171]. All the statistical analyses were conducted using SPSS version 23.0 and Amos version 23.0. --- Results --- Preliminary analyses SES indicators are shown in Table 1. Other descriptive statistics for all variables measured by the scales are presented in Table 2. A correlation matrix of the SES indicators, marital conflict, parenting practices, and child social competence and behavioral adjustment is shown in Table 3. Analyses in study composites showed that all correlations of the study composites were statistically significant. The indicators of SES, marital conflict, parenting practice, and child social competence and behavioral adjustment were interrelated, supporting our hypotheses and previous empirical findings. Each SES variable was negatively related to destructive marital conflict, negative parenting, and the children's externalizing and internalizing behavioral problems. Conversely, it was positively related to constructive marital conflict, positive parenting, and children's social skills. In turn, social skills inversely correlated with children's externalizing and internalizing behavioral problems. --- Mediational models for SES, marital conflict, parenting practices, child social skills, and child adjustment Longitudinal models examined the impact of SES, marital conflict, and parenting practices on child social competence and behavioral adjustment . Figure 3 depicts the final path models, and the path diagram specifies both direct and indirect paths linking T1 SES indicators to T2 child behavioral adjustment . The standardized coefficients are shown in Fig. 3. Model fit was tested with multiple indices; the model provided a good fit to the data [χ 2 = 31.89, p = .023; CFI = .99; IFI = .99; RMSEA = .02]. In the model, several statistically significant direct and indirect paths were found between the predictors and criterion variables. Family income was found to be a significant predictor of lower levels of destructive marital conflict , lower levels of negative parenting practices , higher levels of constructive marital conflict , higher levels of positive parenting practices , higher levels of child social skills , and lower levels of T2 internalizing problems and T2 externalizing problems . The indirect paths from family economy to child mental health functioning through marital conflict and parenting practices were also significant. Maternal education level was found to be a significant predictor of lower levels of negative parenting practices , higher levels of constructive marital conflict , higher levels of positive parenting practices , and lower levels of T2 internalizing problems and T2 externalizing problems . The indirect paths from maternal education level to child mental health functioning through marital conflict and parenting practices were also significant. Paternal education level was found to be a significant predictor of lower levels of destructive marital conflict , lower levels of negative parenting practices , higher levels of constructive marital conflict , and higher levels of child social skills . The indirect paths from paternal education level to child mental health functioning through marital conflict and parenting practices were also significant. Notably, in terms of the negative dimension of family processes , T1 destructive conflict was directly, negatively related to social skills , and indirectly, negatively related to T1 social skills through T1 negative parenting practices. T1 negative parenting practices were directly, negatively related to social skills . Regarding the positive dimension of family processes, T1 constructive conflict was directly, positively related to social skills , and indirectly, positively related to T1 social skills through T1 positive parenting practices. T1 positive parenting practices were directly, positively related to social skills . In turn, T1 social skills were found to be a direct and significant predictor of lower levels of T2 internalizing problems and T2 externalizing problems , while controlling for behavior problems at T1. Therefore, consistent with the hypotheses, each SES indicator was significantly and independently associated with child mental health functioning through positive and negative dimensions of marital conflict and parenting practices. Notably, T1 social skills in preschool, which were affected by T1 family factors, predicted lower levels of T2 behavioral problems in the first grade. --- Discussion Our longitudinal study explored, in a comprehensive model, marital conflict , parenting practices , and social competence as mediators of the association between SES in preschool and child behavioral adjustment in the first grade. Our extension of previous research investigating the relationships between SES and child behavioral adjustment comprised the following three points. We included both family income, and maternal and paternal education levels as SES indicators, and as predictors of family processes and mental health functioning of children , in a unified model. We expected each SES indicator, as predictors, to be differentially associated with family processes and child mental functioning through distinct pathways. We included not only negative mediators , but also positive mediators , as mediating mechanisms in the link between SES and child mental health functioning. We expected both negative pathways and positive pathways in the family process model. We included not only negative child developmental outcomes , but also desirable child developmental outcomes in the relationship between family factors and child mental health functioning. Moreover, we focused on social competence as a mediator of the relationship between family factors and child behavioral problems. We expected social competence in preschool, which was affected by different types of family factors, to be inversely related to the symptoms of behavioral problems in the first grade. Our main findings were the following. Family income and parental education levels were differentially associated with child mental health functioning through distinct pathways. This result provides evidence that lower SES is both directly and indirectly associated with more destructive marital conflict, more use of negative parenting practices, less constructive marital conflict, less use of positive parenting practices, poorer social competence, and more symptoms of behavioral problems. This suggests that, by contrast, higher SES is both directly and indirectly associated with less destructive marital conflict, less use of negative parenting practices, more constructive marital conflict, more use of positive parenting practices, higher social competence, and fewer symptoms of behavioral problems. We identified both negative and positive pathways between SES and child mental health functioning. Positive mediators included constructive marital conflict and positive --- Path of family economic situation, family processes, and child mental health functioning In this study, family income was directly linked to marital conflict, parenting practices, and in turn, child mental health functioning . This result is consistent with previous research findings identifying a direct path of family income to destructive marital conflict and negative parenting practices, and in turn, child outcomes [7, 28-30, 35, 36, 63]. Furthermore, this result supports the FSM's prediction that family income affects children's socioemotional development through its influence on parents' psychological well-being and, therefore, the inter-parental relationship and parent/child interactions [15]. The result also supports the notion of negative spillover effects and is consistent with family systems theory [17,18]. Conversely, we found a positive pathway within which a higher family economic status was associated with more constructive marital conflict, and in turn, more use of positive parenting practices, resulting in higher mental health functioning. This result supports the notion of the positive spillover effect, with the positive inter-parental relationship spilling over into the parent/child relationship, resulting in more positive parenting practices. Similar to negative spillover effects and consistent with family systems theory [18], positive emotions from interparental relationships may transfer to parent/child relationships [82,83]. This result, that there is a positive spillover effect in the family process model, is an extension of previous studies. Additionally, we found that family income was directly related to child mental health functioning , while controlling for other variables. There are likely to be other factors that were not accounted for in our model. For example, the Family Investment Model , which is concerned with the advantages reaped by the developing child because of family wealth [28,172,173], may explain this association. The FIM proposes that families with more economic resources can make significant investments in the development of their children, whereas those with lower incomes must invest in more immediate family needs [1,7,174]. Income enables families to invest in building their children's human capital. These investments in children involve several dimensions of goods and services, including parents' direct and indirect stimulation of learning , the family's standard of living , and living in a more advantaged neighborhood environment that fosters a child's development [7,175,176]. According to this perspective, children in disadvantaged families tend to fare worse because they have limited access to resources that help them develop. Mayer demonstrated that children in disadvantaged families lived under worse conditions, owned fewer stimulating materials, and were less likely to engage in stimulating activities [176]. After controlling for other family background characteristics, these resources were associated with children's developmental outcomes [176]. Therefore, the apparent direct effect of family economic status found in the current study could possibly be mediated by factors that were not accounted for in our model. Future studies should investigate this possibility by including more family factors related to child mental health functioning in their models. --- Path of parental educational level, family processes, and child mental health functioning As mentioned earlier, despite the many studies completed in this area, few studies have simultaneously investigated the influence of family income and maternal and paternal education levels as predictors in the relationships between SES, family processes, and child mental health functioning [6,7,28]. Although most of the previous FSM studies have focused primarily on economic conditions, we suspect that they tend to capture a limited Hosokawa and Katsura Child Adolesc Psychiatry Ment Health 11:62 scope of the influence of educational achievement. In this study, both maternal and paternal educational levels were independently linked with parental functioning and parent/child interactions, and in turn, with child mental health functioning in a unified model, while controlling for economic conditions. In addition, this result also supports the notion of both positive and negative spillover effects [17,18], as educational levels were positively related to higher levels of constructive marital conflict, and in turn, higher levels of positive parenting, resulting in better developmental outcomes. Therefore, the results regarding the effects of multiple components of SES, including family income and maternal and paternal education levels on child mental health functioning through distinct pathways, are an extension of those found in previous studies. In terms of the relationship between educational level and marital conflict, the results of the current study are consistent with those of previous research showing educational attainment to be inversely related to destructive marital conflict [62], and parental educational attainment to be positively related to greater marital satisfaction and marital stability [30,[63][64][65]. More precisely, paternal education was linked to both destructive and constructive conflict; however, maternal education was linked to only constructive conflict. This might be due to difference of effect of maternal and paternal education on decision-making in the home. As mentioned earlier, previous studies have suggested that higher education helps parents strengthen their communication and problem-solving skills, and promotes effective problem solving between parents [50,58]. In addition, higher education tends to make fathers positively participate in decision-making in the home, whereas, fathers with lower education negatively participate [177][178][179]. Therefore, in this study, paternal education might more strongly affect both destructive and constructive than maternal education. Furthermore, in terms of the relationship between educational level and parental involvement, we found that maternal education was associated with positive parenting practices, but not paternal education; however, both maternal and paternal education were linked to negative parenting practices. This result might indicate that the effects of parental education on involvement is larger for maternal education than for paternal. This might be due to mothers tending to be the main provider of care within the households of Japan. Many studies suggest that mothers assume the primary parenting role, in that mothers were found to be more intrusive toward father/child interactions [180][181][182]. In addition, this result is consistent with previous research findings. A large number of studies suggest higher maternal educational attainment to be positively related to positive parenting attitudes, such as talking to children warmly or supportively [72,73], whereas lower educational levels have been found to be predictors of negative parenting, such as harshness and physical disciplinary tactics [33,34,[183][184][185]. However, although many studies suggest maternal educational attainment is related to parenting attitudes, few studies have comparatively investigated the influence of paternal education levels on parental involvement. These results imply the possibility that both maternal and paternal educational levels are independently related to parenting attitudes. One of the important mechanisms in the effect of parental education levels on family processes and children's development is likely to be parental knowledge about childrearing and child development. Lower levels of parental education are associated with negative parenting attitudes, such as physical and authoritarian disciplinary tactics [33,34,[183][184][185]. It has been suggested that this is due to a lack of knowledge concerning the counterproductive outcomes of severe disciplinary responses and appropriate alternatives to harsh discipline [33,183]. Higher levels of parental education have also been positively associated with sensitivity, positive regard, and cognitive stimulation of children [186]. Further, it has been suggested that higher educational levels are associated with increased knowledge about childrearing and child development, and more supportive parenting [72,73]. Therefore, both maternal and paternal education levels may influence parenting attitudes, even when controlling for family income, whereas educational attainment affects parenting attitudes through the adverse effects of poor family economic situations on parents' mental wellbeing. Therefore, we assume that findings related to economic predictions based on the FSM are likely to reflect educational differences in SES as well. Educational levels are likely to play an important role in the relationships among SES, family processes, and child mental health functioning. In addition, we found a direct association between parental education levels and child mental health functioning , while controlling for other variables. There are likely to be other factors that were unaccounted for in our model. The FIM may also explain this mediating pathway to provide evidence for the plausibility of parental education level as an important aspect of the investment process [1,7]. The model proposes that, similar to family income, parental education level has an influence on parental investments, and that these investments, in turn, will have a positive relationship with child development. Parents with higher education levels acquire more knowledge Hosokawa and Katsura Child Adolesc Psychiatry Ment Health 11:62 about child development, have a greater understanding of strategies to encourage social competence, and may be more effective in teaching children [72,73,187]. Families with higher educational levels and more knowledge about childrearing and child development may be more willing to make significant investments in their children's development. Despite the reasonableness of this hypothesized mediating process, there have been limited investigations into the impact of parental education level, in terms of the FIM. However, some evidence is consistent with the aforementioned ideas. For example, a previous study found education level to be positively correlated with parental investments involving a more enriched and positive child-rearing environment, characterized by the availability of play and learning materials, and the organization and diversity of the physical environment [188]. Investment in this regard is not only material , but also emotional [189]. For example, more highly educated parents create a richer and more complex language environment for their children [190]. They also spend more time communicating with their children [173,191]. A previous study found parental education to be positively related to children's language skills, including vocabulary and reading skills [192]. The richness of the language environment in inter-parental and parent/child interactions may mediate the association between parents' education levels and a child's productive vocabularies, and enhance the children's social competence. Therefore, there are likely to be other factors in family processes that were unaccounted for in our model. This result is likely to support the FIM, including its suggestion of parental educational attainment as an SES indicator. More precisely, regarding the path between parental education and social competence, we found that paternal education was directly linked to social competence, but maternal education was not. There are likely to be other factors of paternal characteristic roles that were unaccounted for in our model, in addition to factors of the FIM. For instance, paternal involvement tends to be more physical and challenging than maternal [193,194]. Physical and challenging play is an important component of human socialization [195,196]. Father/child physical play is likely to help children learn to regulate their own behavior, and practice coping with failure or frustration and interpreting others' emotions. This is because father/ child physical play has been linked to children's emotionregulation and peer competence [196][197][198][199]. The positive association between father/child physical play and child social competence is a common empirical finding [195,[200][201][202][203]. In addition, several studies have suggested that fathers with higher educational levels tend to be more involved, have more positive engagement, and are more accessible to their children [78][79][80]. Therefore, fathers with higher educational levels might promote child social competence through not only factors of FIM, but also characteristic parental involvement, such as physical and challenging play. Moreover, regarding the path between parental education and behavioral problems, we found that maternal education was both directly and indirectly linked to T2 internalizing behavior and externalizing behavior; however, the link for paternal education was only indirect. There are also likely to be other factors of maternal characteristic roles that were unaccounted for in our model. For instance, mothers with higher education tend to have higher quality of mother/child interactions, such as sensitivity and responsiveness [188,204]. Past researchers have found that maternal sensitivity and responsiveness significantly shape children's cognitive development. Furthermore, cognitive competence deficits have also been reported as a vulnerability factor in causing behavioral problems [205][206][207][208]. Therefore, maternal educational achievement might affect behavioral problems through the effect of specific mother/child interactions. Future studies should investigate the possibilities of the direct effect of parental education levels, as found in this study, being mediated by factors not accounted for in our model. This could be done by including more factors in future models. --- The role of social competence in the relationships among SES, family processes, and adjustment We focused on both negative child developmental outcomes and desirable child developmental outcomes in the relationship between family factors and child mental health functioning. We also highlighted the ways that family processes within the FSM promote positive developmental outcomes. In the current study, social competence mediated the association between family factors and children's behavioral adjustment in a comprehensive model. SES was positively related to social competence and inversely related to internalizing and externalizing symptomatology, through positive and negative dimensions of parents' marital relationships and parenting styles. This result is an extension of those of previous studies, in which multidimensional family factors were related to both negative and positive outcomes in a comprehensive model. This result is consistent with several previous research findings identifying the direct individual path within which marital conflict and parenting practices are associated with child mental health functioning. In terms of parenting practices and child mental health functioning, in this study, negative parenting practice was directly linked with poorer mental health functioning . By contrast, positive parenting was directly linked to higher mental health functioning . Previous studies have suggested that negative parenting behaviors, such as harsh discipline, being emotionally neglectful, or demonstrating rejecting behaviors, are often associated with lower sociabilitycompetence and increased problem behaviors in children [16,25,143], while positive parenting behaviors, such as emotional expressiveness, responsiveness, and support, have been shown to predict better empathy and social functioning in children [140][141][142][143]. Additionally, in terms of marital conflict and child mental health functioning, in this study, marital conflict was not only indirectly related to child outcomes through parenting practices, but also directly related to child outcomes. Parents' destructive marital conflict was directly linked with poorer mental health functioning . By contrast, parents' constructive conflict was directly linked to better mental health functioning , and in turn, fewer symptoms of behavioral problems. These results are consistent with previous studies indicating that exposure to marital conflict is associated with different responses in children, depending on the type of inter-parental relationship [146,209]. Many previous studies have shown that destructive marital conflict negatively affects social competence [144]. In addition, the relationships between inter-parental destructive conflict and negative psychological adjustment among children are well established [146,149,[209][210][211]. That is, destructive marital conflict has been shown to adversely influence children's social competence [212][213][214][215], internalizing symptoms [211,216], and externalizing problems [210,211]. However, limited research has investigated the impact of constructive marital conflict on child mental health functioning. Therefore, the current result is an extension of those in previous studies, which demonstrated constructive marital conflict's direct association with child social development. One of the important direct mechanisms of the effect of inter-parental relationship on children's development is likely to be modeling. According to social learning theory, children's social development can be influenced by modeling the behaviors and attitudes of significant persons in their lives, such as parents [217]. Child social development may be both positively and negatively related to parents' social development, due to the effects of modeling [218][219][220]. Consistent with the modeling mechanism proposed by the spillover hypothesis, children may directly model conflict behavior exhibited by their parents. In the case of destructive marital conflict, children whose parents resolve their problems through aggressive behavior are more likely to learn that aggression is an acceptable way of dealing with disagreements, and thus, may act aggressively when interacting with their peers [149,221,222]. Therefore, destructive marital conflict is likely to directly limit children's social development. By contrast, in the case of constructive marital conflict, children whose parents resolve problems through supportive cooperation are more likely to learn from the negotiations between their mothers and fathers during the decision-making process, allowing them a blueprint to communicate more effectively and efficiently when interacting with their peers [150]. Therefore, constructive marital conflict is likely to directly enhance social development. In addition, in this study, social skills in preschool, which were affected by family factors, inversely predicted later internalizing and externalizing symptomatology in the first grade, after controlling for preschool behavioral symptomatology. This result is consistent with previous research. A number of studies have shown negative correlations between social competence and behavioral problems. Early social competence among children is an important predictor of later social adjustment and psychopathology [223][224][225][226]. For example, social competence promotes child development in a number of domains, including social adjustment and interpersonal relationships [223,227,228]. Conversely, social competence deficits have been linked to social maladjustment and several problem behaviors, including aggression and delinquency [105,223,[229][230][231][232][233][234]. Previous studies have primarily examined individual relationships between different types of SES, marital conflict, parenting practices, social competence, and child outcomes, without considering these associations in a comprehensive model. However, when considering the complex relationships between these variables, social competence was adversely related to later behavioral problems, as a mediating mechanism in the link between SES and child adjustment. Preschool social competence played a potential protective role in preventing later behavioral problems in the first grade. This result is an extension of previous studies, in which social competence was found to influence later adjustment, as shown in the complex relationships among these variables. The prevailing model of prevention holds that reducing risk factors associated with adverse outcomes, and increasing protective factors that moderate the effects of exposure to risk, will reduce the possibility of later maladjustment [235]. The effectiveness of this approach towards prevention rests on the extent to which identified risk and protective factors are actually causal. Therefore, the current study findings, which focus on multidimensional family factors' simultaneous promotion of social competence among preschoolers, may provide an effective strategy for promoting later social adjustment among children. --- Limitations and future directions Our findings should be interpreted in light of several limitations. First, although this study's design was longitudinal, the design was partially cross-sectional, identifying the relationship between family factors and social competence at T1. The cross-sectional design poses several restrictions that make it difficult to assume causality among the factors. Statistical evidence from studies using a cross-sectional design may not be as informative as longitudinal data [236,237]. Prior studies have found that children's mental health functioning influences interparental relationship and parenting styles, as well as the influence of inter-parental relationship and parenting styles on children's mental health functioning [238][239][240][241]. Children's mental health functioning and family factors are likely to influence each other. Furthermore, followup period of the current study was only 1 year. Although the transition period from early childhood to elementary school is an important period of mental development for children, 1 year may not be enough follow-up time to estimate the effects that have taken place, leading to the possibility of underestimating the impact of SES. Future studies should primarily focus on longitudinal research to examine the effects of family factors on later social competence. Specifically, it is necessary to have longitudinal research with surveys distributed at least three different time points and more long term to clarify the extent to which family factors flow through social competence to affect later behavioral problems. Second, the majority of the data in this study was obtained from only mothers; therefore, there is a risk of reporting bias. This vulnerability to reporting bias can pose a serious potential problem to interpretation of the findings [242][243][244][245]. Single respondents views' toward family factors and child mental health functioning may be skewed either more positively or negatively, thus resulting in misleading findings. The arguments for the examination of the complex relationships between components of SES, family processes, and child mental functioning would seem to be not fully realized with data provided only from mothers. Paternal and maternal education levels or other background information may also influence their views of family factors and children's adaptive functioning; several studies have showed there are discrepancies between the views of fathers and mothers [246,247]. Therefore, this study's data may obscure the extent to which paternal education is associated with the inter-parental relationship, parenting styles, and children's adaptive functioning, since information from the point of view of fathers was absent. Furthermore, other factors may also influence the views of the informants. For example, regarding the interparental relationship, prior studies have shown that views of conflict vary across men and women; women tend to report more conflict episodes than men do, whether for the better or worse [248]. In addition, regarding parenting styles, the data provided by only maternal reports did not reveal information concerning fathers' involvement. Generally, fathers and mothers each have their own parenting styles. Many studies have shown that fathers and mothers play similar or complementary roles in terms of parenting behavior, simultaneously suggesting that their qualities of parenting behavior differ, in particular concerning the amount of physical play; fathering may prove to be more challenging [249][250][251]. Views of children's adaptive functioning behavior may vary across fathers, mothers, and children's teachers. Many study findings indicate that there are several discrepancies among informants, including fathers, mothers, and children's teachers. These discrepancies are particularly prevalent between children's parents and teachers, in terms of their assessment of the children's psychological well-being [242][243][244][245]252]. The discrepancies may reflect children's symptoms, or the opportunities to observe them. Generally, it is not easy for parents to assess early maladaptive behaviors. In particular, parents have difficulty identifying behavior that is indicative of internalizing problems in young children. For instance, it is difficult for parents to distinguish behavior that is reflective of underlying psychopathology from behavior that is reflective of immaturity in self-regulatory competence. Conversely, teachers have the advantage of having the opportunity to observe the behavior of many other children simultaneously. Furthermore, behavioral problems are likely to be more apparent at school than at home. Therefore, obtaining teacher reports may be particularly important for young children to aid in the assessment and forecasting of their school maladjustment and mental health problems [253]. Furthermore, several studies have suggested that the combination of teacher and parent reports with independent assessments is more sensitive than either assessment alone [254]. Therefore, in future studies, reports from several dissimilar informants, including those from fathers and teachers, in addition to mothers, will be needed to more precisely evaluate how family factors affect child mental health functioning. Third, in the current study, we did not consider the interplay between maternal and paternal education, or the interplay between positive and negative aspects of inter-parental functioning. We studied the independent contributions of both maternal and paternal education, and those of the positive and negative aspects of interparental functioning; the framework used in this study does not lead to an examination of the actual interplay among any of these factors. Regarding parental education, we included the independent contributions of both maternal and paternal education level, as we expected each SES indicator, as a predictor, to be differentially associated with family processes and child mental functioning through distinct pathways. However, the argument is incomplete and not generally consistent with theoretical perspectives, including family systems and developmental systems theories [1][2][3][4][5]. Theoretical perspectives suggest there is a more dynamic interplay than the simple additive contribution of maternal and paternal education. Not modeling the interaction between maternal and paternal education achievement may mislead the influence of each maternal and paternal education achievement. Regarding inter-parental functioning, we also included the independent contributions of both positive and negative aspects of inter-parental functioning, as there are reasons we expected each positive and negative aspect of inter-parental functioning to be differentially associated with other variables through distinct pathways. Most studies empirically investigating the FSM have focused exclusively on the negative aspect of inter-parental functioning [15,37]. Previous research suggests the interplay between the positive and negative aspects of inter-parental functioning is more complex than simply looking at the independent contributions of each [15,37]. Previous research also suggests that it is not easy to distinguish the positive and negative aspects of inter-parental functioning, and that children respond to the whole instead of just the parts [29,[37][38][39]. The model including the independent contributions of both the positive and negative aspects of inter-parental functioning may not precisely assess the influences of each. Therefore, the inclusion of maternal and paternal education, and the positive and negative aspects of inter-parental functioning are both strengths and weaknesses of this study. Fourth, we could not exactly assess the positive aspects of inter-parental functioning as a constructive marital conflict. As mentioned earlier, we used the Quality of Co-parental Communication measure to assess the positive and negative aspects of inter-parental functioning. The QCCS captures two aspects of the interparental relationship: Co-parental Conflict ; and Co-parental Support [152]. The Support subscales of this scale measured only "general support"; it has not precisely measured the constructive aspects of conflict. However, in the current study, we treated Co-parental Support, as measured by the Support subscales, as constructive conflict. Thus, the "constructive conflict" we used may not precisely assess the influence of the positive aspects of inter-parental functioning on the other variables. Future studies should investigate this possibility further by using other scales to more precisely assess the constructive aspects of conflict. Fifth, there are likely to be other factors that were not accounted for in our model. As mentioned earlier, we found a direct association between SES and child mental health functioning, while controlling for other variables. There are likely to be other family environmental factors . Furthermore, although we found the effects of certain hypothesized family environmental factors on child mental health functioning, we did not consider genetic factors in our model; it is important to realize children's behavioral problems may be influenced by genetic risks, as well as their family's environmental factors. A large body of evidence supports the conclusion that children's behavioral problems are moderately heritable [255][256][257][258]. Several studies have suggested the extent to which children's mental health functioning is affected by family environmental factors depends on genetic and early temperamental characteristics; environments help determine how genes express themselves [259][260][261]. Children with different genetic attributes will respond differentially to the same environmental circumstances. Therefore, it is difficult to distinguish genetic effects from the effects of family environmental factors on child mental health functioning because genetic factors were not examined in this model. Consequently, there are likely to be other family environmental and genetic factors that need to be included in this model. Future studies should investigate this possibility further by including more family environmental factors related to child mental health functioning. Specifically, these studies could include a genetically informative design , as these types of studies would be useful in accounting for the interplay between individuals and environmental circumstances. Furthermore, although we described earlier that the FIM contends that family SES is associated with neighborhood conditions as one aspect of parental investment, our studies did not assess areal characteristics . Family's socioeconomic resources are likely to largely determine the kind of neighborhood in which they reside [262]. Wealthier parents are expected to reside in areas that have a positive community environment, which provides resources for the developing child, such as parks, good schools, community involvement among residents, and access to conventional friends. Conversely, poor parents are constrained in their choice of neighborhoods. Children reared in neighborhoods without these resources experience a number of negative consequences. Lower income may lead to residing in extremely poor neighborhoods, which are characterized by few resources for child development, such as playgrounds, childcare, health care facilities, and after-school programs. Children who live in areas of disadvantaged neighborhoods tend to have poor physical and mental health [263,264]. Furthermore, several studies suggest that the affluence of neighborhoods is associated with child outcomes over and above family poverty [265]. Thus, future studies will need to include an assessment of neighborhood quality. Finally, these findings may not be generalizable to all families, because there is a risk of attrition bias, and the sample was drawn from a limited geographical area in an urban metropolis of Japan. As mentioned earlier, the retention rate from T1 to T2 was 51.6%, and the T2 returning participants tended to be relatively higher in SES than the non-returning participants. This indicates there is a risk of attrition bias. Therefore, there is the possibility that our analyses could not exactly evaluate the mechanism of children with lower SES, and our analyses may underestimate the influence of SES. Furthermore, some characteristics of Japanese society, such as low levels of economic disparity and high education levels among the general population, may have contributed towards the current results. The reproducibility of the current results should be confirmed using data from other regions in a variety of settings. In summary, future research on these topics would benefit from longitudinal designs and samples with higher retention rates , and greater demographic and clinical diversity. --- Conclusions Despite the above-mentioned limitations, our findings help advance our understanding of the relationships between different types of SES, marital relationships, parenting styles, and child social competence and behavioral problems. This study highlights the need to simultaneously explore the interrelations between multiple family factors to further our understanding of child mental health functioning. Emphasis is placed on the importance of examining both family income and educational levels of parents as SES indicators, to elucidate the relationships between family factors and child adjustment. Additionally, consistent with a developmental psychopathology perspective, this study emphasizes the need to explore both positive and negative aspects of family processes , with a particular focus on the positive dimensions of family functioning. This study also emphasizes social competence as a potential protective factor that prevents later behavioral problems. The current study advances the understanding of SES, marital conflict, and parenting, utilizing a family systems explanation for child development. This study adds to previous literature concerning the relationship between SES and child mental health outcomes by demonstrating that both family income and parental education levels simultaneously and independently influence child mental health outcomes through marital conflict and parenting practices. In addition, the current study adds to previous literature concerning the relationship between SES and child mental health functioning, by demonstrating the positive pathway where constructive marital conflict was shown to be related to higher levels of affirmative parenting and, in turn, more positive outcomes. The current study supports not only the notion of negative spillover effects, but also of positive spillover effects. In addition, social skills, which were affected by multidimensional family factors , adversely influenced later internalizing and externalizing behaviors. Therefore, our study suggests the possibility that theoretical models, including the FSM, should be included with parental educational levels and positive aspects of family functioning and child outcomes when examining the effects of SES. These findings offer preliminary evidence for the need to explore SES by including family income and parental educational levels, and both negative and positive aspects of family functioning. They advance our understanding of SES, marital conflict, and parenting practices, using a family systems explanation for child development. Therefore, our results suggest that we should be sensitive to social inequalities in children's mental health problems and developmental outcomes, and strive to reduce social inequalities. In the long-term, it may be necessary to focus not only on economic support, but also on education, as providing equal access to suitable educational opportunities can positively affect the next generation, and is likely to have a more permanent impact on the child-rearing environment than a temporary increase in income. If more parents can become better educated through an improved social system, it might lead to better developmental outcomes for children. In addition, simultaneously focusing on the marital relationship and parenting style in negative and positive domains may be an effective strategy for developing social adjustment among children. The current study suggests that marital relationships and parenting skills in negative and positive domains may be appropriate for interventions promoting social competence among children to prevent later social maladjustment among parents and children who are socioeconomically disadvantaged. Our findings have important clinical and policy implications. --- --- Abbreviations QCCS: The Quality of Co-parental Communication Scale; APQ: The Alabama Parenting Questionnaire; SSQ: The Social Skills Questionnaire; SDQ: The Strengths and Difficulties Questionnaire. --- --- Competing interests The authors declare that they have no competing interests. --- --- ---

Objective: Using a short-term longitudinal design, this study examined the concurrent and longitudinal relationships among familial socioeconomic status (SES; i.e., family income and maternal and paternal education levels), marital conflict (i.e., constructive and destructive marital conflict), parenting practices (i.e., positive and negative parenting practices), child social competence (i.e., social skills), and child behavioral adjustment (i.e., internalizing and externalizing problems) in a comprehensive model.The sample included a total of 1604 preschoolers aged 5 years at Time 1 and first graders aged 6 years at Time 2 (51.5% male). Parents completed a self-reported questionnaire regarding their SES, marital conflict, parenting practices, and their children's behavioral adjustment. Teachers also evaluated the children's social competence.The path analysis results revealed that Time 1 family income and maternal and paternal education levels were respectively related to Time 1 social skills and Time 2 internalizing and externalizing problems, both directly and indirectly, through their influence on destructive and constructive marital conflict, as well as negative and positive parenting practices. Notably, after controlling for Time 1 behavioral problems as mediating mechanisms in the link between family factors (i.e., SES, marital conflict, and parenting practices) and behavioral adjustment, Time 1 social skills significantly and inversely influenced both the internalization and externalization of problems at Time 2.The merit of examining SES, marital conflict, and parenting practices as multidimensional constructs is discussed in relation to an understanding of processes and pathways within families that affect child mental health functioning. The results suggest social competence, which is influenced by the multidimensional constructs of family factors, may prove protective in reducing the risk of child maladjustment, especially for children who are socioeconomically disadvantaged.

Introduction Since 1990, interculturality has been an important concept worldwide. It has progressively gained prominence in the public, social, and political spheres as the focus of interest of educational improvements at both national institutional and inter/transnational levels [1,2]. If we focus on the educational field, schools have become an important part of intercultural action [3]. Interculturality is increasingly used as a strategy to promote reflection and a culture of dialogue in educational institutions. There is a clear orientation towards changing certain hegemonies of legitimization of the knowledge of what is taught and how it is taught in schools. The aim is to transform those organizational and structuring routines of the school institution that reproduce exclusion. This circumstance, which can be linked to the historical struggles of social movements and their demands for social justice, "can also be seen, at the same time, from another perspective: the one that links it to the global designs of power, capital and market" [1] . It is therefore necessary to understand the current complex situation surrounding the concept of interculturality in formal education. In this sense, given that the term "interculturality" seems to have been used in multicultural, neoliberal discourses, policies, and strategies in an assimilationist and bureaucratic tone, it is useful to understand that "functional interculturality" unquestioningly complies with the established hegemonic rules. On the other hand, "critical interculturality" as a political, social, epistemic, and ethical project suppresses the causes of social asymmetry and cultural discrimination that hinder the possibilities of authentic cultural inclusion. In the field of education, the influence of the neoliberal discourse has tinged the battles for social justice that emerged in the pedagogies of the oppressed [4] and, later, of hope [5,6]. Those that [7,8] were later raised as banners concerning the transformation of the established system refs. [1,2] highlighted the usefulness of most educational reforms and pointed out that these reforms, coinciding with neoliberation policies at different historical moments and in contexts close to globalization which recognize the multiethnic and multilingual character of countries, introduced specific policies for "the different" as part of this multicultural logic of transnational capitalism. --- The Critical Approach to Interculturality in Different Contexts Based on this socio-critical historical perspective, it is useful to point out the divergences, different nuances, and particularities in the way in which each context has developed policies and instruments of public action to respond to the intercultural phenomenon [9]. Conceptions of interculturality differ substantially depending on the countries on which the analysis is focused on. As it was pointed out in [10], the notion of interculturality and intercultural education is not the same in countries that have been created thanks to the migration of diverse cultures, such as Canada, as it is in other European countries and the United States, which have migrations from other countries with a lower socioeconomic level or status. The author also discussed other countries, such as Mexico, where multiculturalism is part of their history and where "cultural differences between the dominant group and the indigenous cultures have led to everything from extermination attempts to concerted national assimilation efforts" [10] . In analyzing intercultural education in terms of the "paradoxical closeness" between the European and Latin American contexts, ref. [11] argued that in both territories coexist "different models and approaches to tackle cultural diversity, which, far from contributing to the transformation of schools and society, tend to deepen social inequalities" [11] . This has happened in many countries because the effectiveness of intercultural proposals depends not only on how much the actions generated are consistent with the needs of each educational context but also on other structural measures that transcend the school environment. If these conditions are not complied with, the "so-called intercultural initiatives could be used as a camouflage for inequalities, as an alibi to avoid the adoption of initiatives that are genuinely respectful of cultural diversity, or as a mere showcase for the most "touristy" and superficial vision of cultural manifestations" [12] . In this respect, it is believed that the European Union's policy initiatives have focused less on the positive effect of improved schooling, or on the political and social participation of immigrant students, but rather on cohesion. The logic governing European policy documents is based on achieving better integration of immigrants into society and, more specifically, into the labor market [13]. In this sense, the authors pointed out that, at these governmental levels, the empowerment of the social and political participation of this group through a more inclusive education is only addressed as a "side effect" of the policies developed. This is consistent with [14] analysis of the clear inconsistencies between the statements of European rhetoric, which stress democracy and human rights, and what is emphasized by national educational policies, basically focused on the acquisition of language and national values to "fit in" migrants. These policies, which supposedly promote social justice, may fail to achieve inclusive citizenship if, while acknowledging demographic changes, they neglect everyday injustices and the European historical imprint of racism and xenophobia [15,16]. Until recently, it would seem that what has prevailed is to find channels, different in each country, to integrate the "different" from the educational system and making them adapt to the "majority". However, there are other indications that their potential can contribute very much to the development of intercultural competence of all citizens due to their capacity for resilience, acceptance, and openness to novelty, especially in second-and third-generation migrant citizens [17]. As [18] stated, the potential of the individual to demonstrate intercultural competence increases as the experience of cultural diversity becomes more sophisticated, moving from ethno-centrism to ethno-relativism. It is in this sense that situations of cultural dialogue can be fostered, based on the empathy and flexibility that are associated, in many cases, with the development of intercultural competence. As [19] defines intercultural competence as "the ability to interact effectively and appropriately in intercultural situations, based on one's intercultural knowledge, skills and attitudes" , what leads to intercultural competence is intercultural learning. In this regard, Cortazzi and Jin stated that intercultural learning refers to how we understand other cultures and our own culture through interaction, generating a cultural synergy that "yields a holistic benefit that is larger than separated elements. It suggests the importance of peer dialogue as well as teacher-student discussion about ways of learning to develop local, contextualized ways of learning." [20] . --- Evolution of the Concept of Intercultural Education All this conceptualization emerged in the field of analysis that has been forged around the central subject of this study, which is intercultural education. The expression "intercultural education" first appeared within the European context in 1983 when the European legislative framework emphasized the intercultural dimension of education in a resolution for the schooling of immigrant children [21]. Since then, as developed by [22], there has been a clear evolution in the conception and approach given to the expression. Since 2015, research on intercultural education begins to move towards a more dynamic, ethical, and transcultural perspective, which is brought to educational practice through interaction and cultural dialogue [23,24]. The static view of intercultural education as a skill in dealing with other cultures, ethnicities, or languages has been disregarded. It has become increasingly evident that cultural diversity itself, promoted by national education systems, is not appropriate for intercultural learning and dialogue to take place [25,26], always in a living process of formation. In this field, where discourses and training practice, sometimes unrelated, have multiplied worldwide, a rallying point is extracted: intercultural education, supported by inclusion and social justice, should address the entire school community, fostering dialogue and democratic coexistence, going beyond the mere recognition and acceptance of differences [27,28]. The authors of [22] highlight an important contradiction: the national education system and the teaching and learning approach it fosters is precisely what raises obstacles to intercultural dialogue. Thus, it is easy to understand that the school recognizes this intercultural approach in declarative terms but maintains an assimilationist logic in its praxis [29]. The conflict between the official discourse on intercultural education and its actual application in school practice should not be overlooked. In specific areas, where social coexistence with diverse cultures has significantly intensified , publications have also increased significantly in the last two decades [30][31][32]. One of the fundamental reasons concerns the interest in caring for the intercultural coexistence that has developed given the high rate of immigration in Spain. According to [12], there is a clear definition of the concept of intercultural education that emphasizes the perspective of interaction and the benefit of cultural pluralism typical of democratic societies as a pedagogical resource. Intercultural pedagogy is understood as a "reflection on education, [ . . . ] based on the recognition of cultural diversity value" [12] . What is important about this proposal is that it is based on a model of analysis and intervention that includes all the dimensions of the educational process. At the time when intercultural education was taking shape in Spain, a clear vision was provided showing the existing gaps in the research carried out up to that time in relation to both structural and curricular measures. The academic centers had a limited margin to put faith in an educational project based on a medium/long-term consolidated center where interculturality was hardly existent. At that time, it became clear that "organizational and curricular changes do not usually affect the general curriculum, but rather have a compensatory character" [12] . From the results of these studies, important recommendations were drawn regarding the grouping and selection of students. It was aimed at the need for flexible groups adapted to teaching-learning processes. In relation to formative and evaluative activities, cooperative learning was emphasized to favor interaction. Regarding another critical dimension, the participation of families and students in the center, it was suggested that it is not so much a matter of continuing to facilitate family-school-community bridges, but rather of "creating educational spaces where all those involved have a role to play in the definition and development of the educational model" [12] . Until 2003, intercultural education was an illusion rather than a reality; it was necessary to state and support actions towards the creation of models of educational projects in schools that would build a solid and consistent proposal for intercultural education. Later, in their systematic review, ref. [33] point out two fundamental lines that have marked a trend between 2009 and 2019: the one related to the perceptions and beliefs of teachers in relation to ethnocultural diversity, and on the other hand, the concern to offer a consistent and sustained response to this diversity in schools and communities [34,35], bearing in mind the crucial role of the family in the important dynamics of change both inside and outside the classroom. --- Towards an Inclusive Approach to Intercultural Education in Schools Since 2015, there has been a certain abundance of studies emphasizing an inclusive and dynamic conception of intercultural education based on quality interactions generated in schools. However, it is necessary to review what has happened in the last five years, analyzing whether this approach has permeated organizational and curricular processes and pedagogical practice in schools and whether the "terrain marked by political discourse" on the inclusion of cultural diversity [25] has developed. From the perspective that has been reflected, beyond neoliberal slogans, a review is proposed on the way educational institutions and school communities may be succeeding in inclusive practice and process practice towards creating an authentic intercultural education. This review considers that an inclusive intercultural approach must be supported by an involved and collaborative institutional organization, the progress of which is achieved through the construction of a sense of belonging to the educational community [36][37][38][39][40]. The need for collective responsibility in a climate of exchange between educational agents for the improvement of competent learning for all students [41] derives from this principle. The importance of transforming practice by reviewing the ways of understanding relationships and interactions between the staff, with the students, and with the educational community must be recognized [42]. From this perspective, it is important to assume a meso-, micro-, and macro-vision of intercultural education, analyzing whether the political purposes and the type of institutional conditions that educational administrations impose on schools are consistent with the daily margins of action of teachers in their centers. As [43] points out "the school, in the last decades, has been accumulating an excess of expectations and missions, as a consequence of having delegated everything that the family or society could not do or was concerned about". This takes the risk of making the school responsible for what it cannot achieve in isolation and on its own. In this sense, it is necessary to strengthen the conception of educational contexts as spaces for dialogue beyond the physical barriers of schools [44], where critical reflection on dilemmas and paradoxes can be enriched so that inclusion can take root in praxis [45]. It must be assumed that in order to achieve a true culture of diversity, the traditional view of cultures and the compensatory curricular approach for minorities must be overcome [25,44,46]. At the same time, intercultural education must be considered beyond specific ethnic markers, from which a process of homogenization of cultural groups is sought to be assembled starting from a cultural logic [4]. The message that [47] conveyed a few years ago in a monograph focused on intercultural education culminates the theoretical framing of this systematic review: Today it is more necessary than ever to promote intercultural education with a pedagogical approach of inclusive nature and focused on the community, and it is not only a matter of addressing the organizational, didactic, and curricular responses of the intercultural school, but the nature, meaning, imprint and character of educational actions that generate peaceful and respectful coexistence, equity and participation, education and democratic values in academic institutions. --- Objectives Based on this approach, the aim of this systematic review is to examine the role of the school community in the implementation of authentic inclusive intercultural education. The specific objectives are as follows: To conduct an updated review of the reorganization processes that the school community have determined in order to offer an inclusive educational response to cultural diversity. To analyze the current response of schools as participatory and collaborative communities in intercultural education. To evaluate the current practical reality of schools as inclusive and intercultural education centers. --- Materials and Methods To achieve the proposed objectives, a systematic literature review was carried out applying the protocol of the updated PRISMA statement [48] for the identification and selection of documents. This methodology aims to identify, analyze, and assess those studies that highlight the institutional dimension of schools as participatory and collaborative educational communities. The intercultural education approach is analyzed in relation to the following dimensions: 1. The construction processes of the participatory school climate. --- 2. The organizational and pedagogical structure of inclusive schools. --- 3. The impact of practice and experiences created in these inclusive intercultural schools to generate intercultural learning. To study these facets, concepts related to intercultural education, cultural diversity, coexistence, and school climate, as well as the school context, were chosen as fundamental elements that make up a true educational community based on the model of collaboration and participation. Scopus and Web of Science databases were used for English language articles and Dialnet for Spanish ones. The following descriptors from the UNESCO thesaurus were used to search for documents: school, "intercultural education", participation, "cultural diversity", practice, "primary education", "secondary education", and "elementary education". These descriptors were combined with the Boolean operators AND, OR, and NOT. Several combinations were tried until the final search formula was: "intercultural education" AND . The search was carried out using the topics in both English and Spanish. After each search, the documents were selected based on the inclusion and exclusion criteria established beforehand . f. Analysis of documents, essays, or theoretical reflections. g. Studies focused on intercultural education from the inclusive cultural diversity approach in schools. g. Studies focused on other types of specific educational support needs unrelated to intercultural school contexts and with a compensatory and non-inclusive approach. For the analysis and classification of the documents based on the established dimensions , an analysis was applied to evaluate the degree of agreement and disagreement among the 3 judges who participated in the classification of the documents based on the inclusion and exclusion criteria that were assessed. Considering that some agreements could be the result of chance and that their random effect could give a greater reliability effect than the real one [49], the Perreault and Leigh coincidence coefficient was used to amend this result. Its values range from 0 to 1 , and the coincidence value was I = 0.83 which made it possible to consider the final selection of documents as adequate. I r = Fo N - 1 K • K K -1 1/2 if Fo N ≥ 1 K I r = 0 if Fo N < 1 K Fo = number of opinions on which the judges agree N = total number of opinions K = total number of categories --- Procedure The exhaustive search began by using the keywords, then the inclusion criteria were applied, eliminating those documents that did not meet them. Using the Mendeley bibliographic manager, duplicates were located and subsequently eliminated from the initial selection. After this, titles and abstracts were read and the studies that met the inclusion and exclusion criteria were selected. A complete reading was performed out of the chosen texts, and those that did not focus on the objectives of the study after an in-depth analysis of their content were excluded. The objectives, sample, methodology, main results, and conclusions were analyzed for each article from the sample of selected documents. The initial selection of documents was as follows: A total of 3755 results were obtained. After applying the inclusion and exclusion criteria, 3732 documents were eliminated. The final sample of studies consisted of 21 documents. The following flow chart shows the selection process and the thorough examination of the articles . --- Results --- Characteristics of the Included Studies All articles published on the subject up to the time of the closing of this review were reviewed. The 21 articles selected for the review were studies published mainly in English between 2018 and 2023, which ensured results on experiences developed in educational centers around the new models of promotion and work of interculturality in schools. The studies were conducted mainly in Spain and Chile, although they were also carried out in countries such as Mexico, Brazil, Colombia, Russia, Italy, Austria, Denmark, Slovenia, Poland, and the United Kingdom). It is noteworthy that 11 out of the 21 studies were conducted exclusively in Spain. Four studies used a purely quantitative methodology and the other fifteen used qualitative methodology . Two studies were mixed . Some evaluation tools were used in the studies, including questionnaires , observation , interviews , focus groups , and documentary analysis . In 47.1% the sample was made up of teachers, management teams, or social professionals, in 32.4% the sample was made up of students, in 14.7% of family members, and in 2.9% there were programs designed for the intervention of children with cultural diversity. All the information can be seen in Table 2, which also lists the authorship, year of publication, purpose, sample, country, methodology with the evaluation tool, and main results. --- Results --- Characteristics of the Included Studies All articles published on the subject up to the time of the closing of this review were reviewed. The 21 articles selected for the review were studies published mainly in English between 2018 and 2023, which ensured results on experiences developed in educational centers around the new models of promotion and work of interculturality in schools. The studies were conducted mainly in Spain and Chile, although they were also carried out in countries such as Mexico, Brazil, Colombia, Russia, Italy, Austria, Denmark, Slovenia, Poland, and the United Kingdom). It is noteworthy that 11 out of the 21 studies were conducted exclusively in Spain. Four studies used a purely quantitative methodology and the other fifteen used qualitative methodology . Two studies were mixed . Some evaluation tools were used in the studies, including questionnaires , observation , interviews , focus groups , and documentary analysis . In 47.1% the sample was made up of teachers, management teams, or social professionals, in 32.4% the sample was made up of students, in 14.7% of family members, and in 2.9% there were programs designed for the intervention of children with cultural diversity. All the information can be seen in Table 2, which also lists the authorship, year of publication, purpose, sample, country, methodology with the evaluation tool, and main results. To analyze the challenges of school leaders when including migrant students in a Chilean school. To identify school leadership practice that favor the inclusion of migrant students. To understand school leaders' perceptions about the inclusion of migrant students and their strategies to promote it. To identify barriers and challenges in the process of inclusion of migrant students and how school leaders can address them. Qualitative methodology A total of three members of the management team ; three teachers with migrant students and three inspectors, three teachers of the school integration program, and six parents of foreign origin. Semi-structured interviews and a focused discussion group. Three strategies were identified for school leaders to foster the inclusion of migrant students in a Chilean school and establish an inclusive school climate. They provide emotional and social support and adapt pedagogical practice to meet the needs of migrant students. --- Chile León et al. [51] To know the management tasks that promote inclusion. Quantitative methodology A total of 243 families and 154 teaching teams from 17 public and subsidized primary and secondary schools. Questionnaire: leading inclusive education . The management teams of schools in socioeconomically disadvantaged areas carried out fewer inclusive actions than those located in more favored areas. --- Spain Jiménez-Vargas [52] Analyze how schools manage the tensions between an inclusive model and an advanced neoliberal model, aiming to identify effective strategies to promote inclusion and educational equity. To determine teachers' beliefs about cultural diversity and how they influence planning and teaching practice. --- Qualitative methodology A total of 45 elementary school teachers. Interview There is a discrepancy between the theories and normative documents of intercultural education and the beliefs and practice of teachers. There are biases in the teachers' statements which led to a moralizing discourse and the repetition of practice without reflecting on their appropriateness to the specific context. --- Italy Garreta-Bochaca et al. [55] To analyze the degree of recognition of cultural diversity in the school's documentation, the actions carried out in relation to cultural diversity , and the practical development of intercultural educational practice. --- Quantitative methodology A total of 1730 representatives of the management teams of primary education schools. --- Questionnaire The results indicate that there is still a long way to go in intercultural education. Schools with a lower presence of students of foreign origin tend to perceive this issue as less of a priority and act less in this direction compared with other schools. The misconception persists that it should only be addressed in culturally diverse schools. --- Spain Antolinez Domínguez and Jorge Barbuzano [56] To analyze the models of cultural diversity management through two case studies carried out in Spain and Mexico during the rise of intercultural education. --- Qualitative methodology Two schools, one in Andalusia and the other in Oaxaca . Key players in the centers: 55 interviews in Spain and 47 in Mexico. Case studies: school and community ethnographies It highlights the importance of identifying the aspects that can function as analytical facets for intercultural education from the paradigm of diversity. The paradigms of inequality and difference are those that underpin educational policies based on culturalism, which has a double logic. First, diversity is reduced to cultural criteria, which can confuse diversity with inequality. Second, cultural homogenization occurs within groups, leading to a distinction between "them" and "us". --- Spain and Mexico Dežan and Sedmak [57] To analyze the influence of the school environment on the well-being of adolescent migrants, focusing on the pedagogical practice and interpersonal relationships established between them, their peers, and teachers. How these factors may affect their academic, social, and emotional adaptation and development in the school environment. Analyze the process of implementation of activities and reflect on the notion of interculturality in education from a broad perspective related to global citizenship and education for development. --- Qualitative methodology Two primary education centers, with few migrants in the Basque Country. A total of 16 teachers. Focused discussion group and teacher's logbooks . Three perceptions of interculturality were distinguished: generalist; focused on cultural aspects; and social justice approach. The intercultural logic is based on assumptions such as equality and justice, which cannot be imposed but must be adopted at a personal level. --- Spain Molina Díaz and Sales [63] To know the process of building the school's educational community and the factors involved in this process. --- Qualitative methodology Interview of the principal, two teachers, two parents, two students, and one member of the city council. Focused discussion group: four parents, four teachers, and three students. Interviews, focused discussion group, and participant observation). The results remark the importance of highlighting the collective identity of the community at the beginning of the construction process and how affiliation to a group is not enough for people to identify with it. The complexity of the collective identity-building process in the community is highlighted. The importance of searching for identity symbols and motivating community members on an ongoing basis is pointed out. Dichotomous discourse that influences the evolution of collective identity. --- Spain Chile Torrelles Montanuy [65] To analyze the predominant approaches and actions in the practical development of intercultural education in primary education in Spanish schools considering the ownership and the percentage of students of foreign origin in each of them. --- Mixed methodology A total of 1730 primary school students; 19 interviews with teachers from five schools. Questionnaire and interview. The results show that the presence of foreign students is the most influential variable in the implementation of intercultural education in schools. Although there is a discursive recognition of the value of intercultural education, no actions with an intercultural perspective are identified in its practical application in most schools. Democratic participation in the community is essential and is related to co-responsibility and shared sovereignty. The relationship between democratic participation, collaborative culture, and distributed leadership in the inclusive educational community is highlighted, and the importance of learning to participate is emphasized. --- Spain --- Gromova et al. [66] To identify the educational practice used by elementary school teachers to provide linguistic and academic support to immigrant children and to promote a welcoming classroom environment that favors their psychological adjustment. --- Qualitative methodology A total of 20 elementary school teachers who have experience with immigrant students. --- Interview Teachers do not use special methods for teaching Russian as a foreign language and recognize the need for training in teaching methods for immigrant students. There is a need for teacher training programs that include teaching multilingual and culturally diverse students in the classroom. To describe the development of the learning and service project as a curricular practice linked to the territory. To analyze the strategies of curriculum negotiation and student participation in this educational practice. --- Qualitative methodology Students in sixth grade of primary education, two teachers, the families involved in the ApS, the school's management team, and the education councilor of the city council. Participant observation, interviews, focused discussion groups, audiovisual records, documentary analysis, and field diary. The ApS curricular practice carried out among all sectors of the educational community, through cooperative dynamics and participatory social diagnoses such as social mapping, cooperative learning, and class assemblies, has allowed students to reach their maximum learning potential and achieve true democratic participation. The educational experience has enabled all participants to become critically and actively aware of their reality and has led to a transformative and emancipatory education. The leadership exercised by the teachers is within the parameters of distributed leadership, central to the experiences of democratic schools and social justice models. --- Spain Guzmán Vargas et al. To identify a pedagogical strategy that would allow the interaction of the Wounaan and Mestizo cultures present in the classrooms of a school in Bogota. --- Qualitative methodology A total of 27 students of the Compartir Recuerdo IED school . Action research . The project "Sharing our worlds : Construction of a mural through collaborative work, sharing experiences, languages, cultural diversity and ways of life with empathy, solidarity and respect for diversity" is presented. Intercultural processes have been achieved in students, parents, and teachers, which highlights the importance of pedagogical strategies that promote the formation of individuals who transform formative dynamics based on cultural diversity. It is necessary to reconstruct the meaning of inclusive education in order to take into account cultural exchanges in the classroom and allow the general population to achieve human dignity and quality of life. To describe and understand how intervention in cooperative learning and the use of cell phones in kindergarten and primary school classrooms favors school coexistence in a context of socioeconomic vulnerability and cultural diversity. --- Qualitative methodology A total of 52 students from two kindergarten classrooms and their respective tutors, and 55 students from two primary school classrooms and their respective tutors. Participant case study, semi-structured interviews, and focused discussion group. The results point to a reduction in disruptive behaviors, academic disaffection, and exclusion, which is mainly attributed to the combination of some cooperative elements that are favored by the cell phone . The educational use of cell phones with a cooperative methodology improves school coexistence. --- Brazil Mela-Contreras et al. To encourage dialogue and critical reflection among teachers and students on the respect and appreciation of cultural and ethnic diversity through cinematography. --- Qualitative methodology Five public schools, 11 teachers, and 145 seventh grade students in visual arts and history, geography, and social sciences. Documentary and narrative analysis of film language. The Cineduk program generated in the students the acquisition of adequate competencies to interpret and signify the diversity of the world . It advocates cinematographic praxis to enrich teaching practice and learning through the curricular and cultural content of students. The development of cinematography as a didactic strategy promotes non-discriminatory and dialogic teaching. Based on the dimensions established for the analysis of the documents, Figure 2 shows the distribution of studies by dimensions/fields. As can be seen, most of the studies focused on organizational aspects and the pedagogical approach, while the aspects centered on the processes of building a participatory/collaborative school climate were the least frequent studies. Based on the dimensions established for the analysis of the documents, Figure 2 shows the distribution of studies by dimensions/fields. As can be seen, most of the studies focused on organizational aspects and the pedagogical approach, while the aspects centered on the processes of building a participatory/collaborative school climate were the least frequent studies. --- Discussion To facilitate the reading and discussion of the results, we display the analysis arranged according to the dimensions that made it possible to identify the main lines of work developed by the schools in the framework of intercultural education. Regarding the processes of organization and management of the educational response and pedagogical approaches, it should be noted that several studies highlight the importance of school leadership and the adaptation of pedagogical practice in promoting the inclusion of migrant students and the implementation of an alternative intercultural education model focused on the needs of migrant groups [50,64]. It has been observed that management teams in socioeconomically disadvantaged areas have more difficulty carrying out inclusive actions [51], suggesting the need for support and improvement plans that are more ecological and sensitive to the nature of the community to which it intends to respond. Studies such as [52] highlight the role of teachers in educational inclusion and the importance of training models focused on fundamental aspects for the improvement of attitudes and respect for cultural diversity. The coexistence of different models for the promotion of respect for cultural diversity generated resistance practice that strained the management of a model truly adequate to the needs of the context. In this sense, the adaptation of pedagogical practice is essential to meet the needs of migrant students [50]. The presence of foreign students is a key factor in the implementation of intercultural education in schools [65], but it is also essential to overcome the misnamed "folklorization" that has led to the identification of different cultures by reference to their traditions, gastronomy, dances, etc. [65]. The construction of a collective identity in the educational community is a dynamic and complex process in which all members of the community must participate [63]. The existence of a dichotomous discourse in intercultural education can lead to cultural homogenization within groups and a distinction between "them" and "us" [56]. As [54] points out, there is a discrepancy between the theories and normative documents of intercultural education and the beliefs and practice of teachers, which can generate contradictions and make it difficult to effectively implement intercultural education. Taken together, these findings suggest that it is crucial for schools to adopt an inclusive pedagogical approach tailored to the needs of migrant students and to promote cultural diversity supported by committed school leadership and adequate teacher training in intercultural education [50,52,64]. It is essential that the educational administration provides support and specific improvement plans to schools in socioeconomically disadvantaged areas to facilitate the implementation of inclusive actions [51]. Intercultural education should be 57% 24% 19% 00% 10% 20% 30% 40% 50% 60% Organizational structure and pedagogical approach The impact of the practices and experiences created in the schools The processes of building a participatory/collaborative school climate --- Discussion To facilitate the reading and discussion of the results, we display the analysis arranged according to the dimensions that made it possible to identify the main lines of work developed by the schools in the framework of intercultural education. Regarding the processes of organization and management of the educational response and pedagogical approaches, it should be noted that several studies highlight the importance of school leadership and the adaptation of pedagogical practice in promoting the inclusion of migrant students and the implementation of an alternative intercultural education model focused on the needs of migrant groups [50,64]. It has been observed that management teams in socioeconomically disadvantaged areas have more difficulty carrying out inclusive actions [51], suggesting the need for support and improvement plans that are more ecological and sensitive to the nature of the community to which it intends to respond. Studies such as [52] highlight the role of teachers in educational inclusion and the importance of training models focused on fundamental aspects for the improvement of attitudes and respect for cultural diversity. The coexistence of different models for the promotion of respect for cultural diversity generated resistance practice that strained the management of a model truly adequate to the needs of the context. In this sense, the adaptation of pedagogical practice is essential to meet the needs of migrant students [50]. The presence of foreign students is a key factor in the implementation of intercultural education in schools [65], but it is also essential to overcome the misnamed "folklorization" that has led to the identification of different cultures by reference to their traditions, gastronomy, dances, etc. [65]. The construction of a collective identity in the educational community is a dynamic and complex process in which all members of the community must participate [63]. The existence of a dichotomous discourse in intercultural education can lead to cultural homogenization within groups and a distinction between "them" and "us" [56]. As [54] points out, there is a discrepancy between the theories and normative documents of intercultural education and the beliefs and practice of teachers, which can generate contradictions and make it difficult to effectively implement intercultural education. Taken together, these findings suggest that it is crucial for schools to adopt an inclusive pedagogical approach tailored to the needs of migrant students and to promote cultural diversity supported by committed school leadership and adequate teacher training in intercultural education [50,52,64]. It is essential that the educational administration provides support and specific improvement plans to schools in socioeconomically disadvantaged areas to facilitate the implementation of inclusive actions [51]. Intercultural education should be extended to the entire student body and educational community, fostering the construction of a collective identity that values and celebrates diversity [63,65]. It is necessary to promote teacher training in inter-cultural competencies and to encourage critical reflection on own beliefs and practice in relation to cultural diversity [56]. Concerning the situation of schools in relation to creating participatory and collaborative communities, with a climate favorable to intercultural education, the selected studies highlight the importance of building a participatory and collaborative school climate to promote the well-being and inclusion of migrant students [57]. Intercultural education and successful interpersonal relationships between teachers and students are fundamental to achieving this goal [58]. Likewise, it is necessary to address the specific difficulties faced by migrant students, such as migratory experience, gender role conflicts, and belonging to segregated groups [59]. Overcoming these challenges requires a caring school culture and teacher training in intercultural and emotional competencies, as well as in the collaborative relationship with migrant families [59]. Promoting intercultural sensitivity among native pupils and paying attention to the academic self-concept and life satisfaction of immigrant students are key elements to improve school coexistence and the educational inclusion of this group [60]. The strategies identified by the different studies analyzed include fundamentally: teacher training in interculturality, the development of intervention programs to improve communication and emotional self-concept, and the participation of families and social entities to promote respect and trust among students [60]. In this sense, the affective dimension emerges as a fundamental aspect of educational research, emphasizing the importance of understanding and intervening in education at the local level by recognizing intercultural educational regions and their particularities [58]. Establishing horizontal relationships in the process can expand the possibilities of understanding the cultural meanings and senses that guide people's actions [58]. Where the practical and experiential reality developed from schools on intercultural education is concerned, the selected studies show the impact of the practice and experiences created in schools to generate intercultural learning. For example, ref. [66] examines the teaching methods used in Russia to teach Russian to immigrant children and highlights the need for training in teaching methods for immigrant students. The findings suggest that teacher training programs are needed that include teaching multilingual and culturally diverse students in today's classrooms [66]. According to [67], educational practice allows students to reach their full learning potential and achieve true democratic participation. These educational experiences promote a transformative and emancipatory education, and the leadership exercised by teachers is recognized as essential to the experiences of democratic schools and social justice models [67]. Where [68] is concerned, there is an emphasis on the importance of pedagogical strategies that promote the development of transforming individuals of formative education dynamics based on cultural diversity. Pedagogical proposals should be based on relational, functional, and critical interculturality to generate spaces of exchange considering the knowledge of each student [68]. As examples of practical strategies, ref. [69] indicates that the educational use of mobile phones with a cooperative methodology is a good example of a useful strategy to improve school cohabitation and reduce disruptive behaviors, academic disaffection, and exclusion [27]. On the other hand, refs. [70,71] are committed to the use of cinematography as a didactic strategy to promote non-discriminatory and dialogic teaching by using film language in audiovisual creations on cultural diversity. --- Limitations and Future Work One of the main limitations found in this study focuses on the nature, objectives, and scope of the interventions around the working model in intercultural education. The evidence points to the need to consider new organizational models for managing and approaching pedagogical styles focused on the construction of an intercultural school climate. Isolated, sporadic interventions are disconnected from the organizational and ped-agogical development of the centers without a global and integrating vision of intercultural education within the curricular design of them. These are highly frequent and ineffective, as they only cover areas of the curriculum and are not integrated into the culture of intercultural inclusion. The reality of the intercultural fact of the centers is hardly visible, few experiences are shared, and the real scope of the actions carried out in the schools is unknown. The specific and particular initiatives that are being proposed could be redirected toward the construction of innovations at the center or institutional level and could be triggered for more global actions that mobilize the entire educational community. In this way, the deployment of inclusive intercultural education could result in repercussions of greater scope, reach, breadth, and significance. It is recommended that future studies consider the experiences of centers, the pedagogical principles of which contemplate the educational response to intercultural reality. Field studies could improve the analysis, quality, and scope of the experiences that have been collected in the studies presented in this manuscript. These studies would provide detailed information on how this practice is being applied in schools and how it is perceived by students, teachers, and families. Therefore, the knowledge of this practice could be deepened and provide more precise and effective recommendations for educational leaders who wish to implement strategies to foster a culture of inclusion and diversity in their institutions. --- Conclusions This study dealt with three key dimensions in the search for inclusive and intercultural education in diverse school environments. Through an updated systematic literature review, management reorganization processes and pedagogical approaches, participatory and collaborative school climate construction, as well as the impact of practice and experiences created in schools to generate intercultural learning were assessed. The study provides a comprehensive view of the advances and challenges in inclusive and intercultural education in diverse school sceneries. Although significant progress has been made in implementing inclusive practice and building participatory and collaborative school climates, there are still areas that require greater attention and development. One of these areas is the ongoing training of teachers in intercultural competencies and socio-emotional skills. The inclusion of cultural diversity into the curriculum and the adoption of inclusive pedagogical approaches not only enhances the educational experience of students but also promotes fairness and social justice in the educational system. In addition, it is essential to strengthen the relationship between schools and families to support the process of integration and adaptation of migrant students and their families to the school environment. It is important to continue researching and evaluating the impact of educational practice and experiences on the generation of intercultural learning. Identifying and promoting effective pedagogical strategies and innovations in teaching and learning can contribute to a more inclusive and fair education. In addition, it is essential to promote the participation and collaboration of all actors involved in education, including students, parents, teachers, administrators, and educational leaders, as well as political authorities and civil society organizations. In summary, this study provides an updated and comprehensive overview of the situation of inclusive and intercultural education in diverse school environments, linking the results obtained in the three dimensions with the objectives. The findings can help future research and practice in this field, as well as inform educational policies and strategies to address cultural diversity in schools. Through an interdisciplinary and collaborative approach, it is possible to move towards a more inclusive, equity-based, and socially fair educational system that values and celebrates cultural diversity as a stimulating resource for all students. --- Data Availability Statement: Information and queries on the data used can be obtained from this article. ---

In today's society, high-quality educational contexts must include intercultural education and educational inclusion as main elements of school culture. Equity, social justice, and equal opportunities for everybody require the construction of flexible processes, relationships, and organizational structures open to diversity. This paper presents an updated review of studies focusing on models of educational responses adapted to cultural diversity. It analyzes the response of schools as collaborative communities in intercultural education and their reality as inclusive and intercultural communities. An exhaustive search of documents was carried out, consulting the Web of Science (WoS), Scopus, and Dialnet databases. After analyzing and applying inclusion and exclusion criteria, 21 documents were identified that showed the structural, cultural, and relational transformation of educational centers and the improvement of their organizational and public response and adaptation to current needs. The challenge of building intercultural educational contexts is a concern for schools and the educational agents who coexist in them.

INTRODUCTION Non-communicable diseases are estimated to be responsible for two-thirds of global mortality. 1 Cardiovascular diseases are the most prevalent NCD, both globally and in South Africa. 2 3 NCDs remain vastly undersupported; currently less than 0.1% of South Africa's national health budget is spent on NCDs. 4 Access to care is a vital element of universal health coverage and a key challenge for South Africa; economic, geographical and racial inequalities together create vast disparities. [5][6][7] The South African government is now in its second phase of implementing National Health Insurance. 8 This second phase has come with an emphasis on strengthened and community-based primary care: a national programme, Re-Engineering Primary Health --- Strengths and limitations of this study ► This is the first known investigation combining both mortality trends and barriers to access to care for people who have died of cardiovascular diseases. ► In combining quantitative and qualitative analysis to develop a three delays model specific to cardiovascular disease in the Agincourt population, this study provides a basis for development and application to other settings to deal with care access and quality issues in the final stages of chronic illness. ► Using two components of verbal autopsy data provides complimentary information and a basis for refining circumstances of mortality indicators, but direct comparison is limited by methodological differences in data collection. ► The non-specificity of verbal autopsy cause of death categories, particularly 'other and unspecified cardiac diseases,' reduces the ability of this study to analyse trends in specific cardiovascular disease conditions. ► Other factors influencing equity in access to care, such as migratory status and socioeconomic status, cannot be included due to lack of available data within verbal autopsies. --- Open access Care, was initiated in 2010 with the aim of strengthening health promotion and prevention and early detection of disease. 9 As a response to South Africa's dual burden of HIV/AIDS and NCDs, the National Department of Health introduced an integrated chronic disease management model in 2011. 1 Although other studies have quantitively examined gaps in care-access using a continuum or 'cascade' of care approach, 10 11 knowledge of where barriers in access to care for CVD occur and the identity of such barriers in rural South Africa, as well as who is less likely to seek care, is a key gap in the literature. This is a crucial area for research, as delays in accessing care are an important predictor of patient morbidity and mortality for CVD outcomes. 12 Due to a lack of civil registration data and death certification systems in many low-income and middle-income countries , existing data on trends in CVD deaths may vary in reliability. 13 There is a civil registration system in South Africa, with marked improvements in recent years, but evaluations have highlighted persisting quality issues that may which influence its reliability and validity . 14 Verbal autopsy is a method of ascertaining cause of death and the circumstances surrounding that death in situations-such as deaths in the community-where other methods are not available or reliable. VAs are done by trained fieldworkers who ask a standard set of questions to carers of the deceased. Either physician judgement or computer algorithm is used to assign cause of death based on these data, 15 which has been shown to be reliable and reproducible. 16 Since 2012, versions of the WHO VA tool have included questions on circumstances surrounding deaths. 17 There is also a free text section where fieldworkers record summaries about the death and its circumstances, which may include information not captured on the binary responses from the rest of the VA tool-including access to care. VA presents a unique opportunity to understand social exclusion from access to health systems within a population, as well as quantify causes of death. When conducted sequentially in the same community, VA enables an understanding of how patterns of access to health systems change over time. This research presents both substantive findings to this end and develops the VA methodological approach. A three delays model can be used to examine barriers in accessing care. This approach categorises delays into three phases: seeking, reaching and receiving care , 18 and draws attention to the multifactorial nature of access to care. 19 It facilitates an overview of access to care for a particular context by breaking down the care pathway into a series of steps, and suggests targets for health planning. Use of the three delays model has been used extensively for care in obstetric emergencies 18 with recent broader applications, for example to sepsis, 20 trauma 21 22 and time-critical conditions. 23 It has not yet been applied to CVDs. This study aims to contribute to bridging the gap in the evidence needed to support prioritising and planning services for NCDs, which is especially important as South Africa works towards the goal of universal health coverage. 24 This study is the first known investigation combining both mortality trends and barriers to access to care for people who have died of CVD. Using VA data, this study documented trends in CVD deaths that can be described as 'downstream' deaths; cardiac conditions, strokes, and renal disease, rather than 'upstream' CVD risk factors . It aimed to understand differences over time, by condition, age and sex of those dying in a rural population in South Africa. This study also aimed to ascertain barriers to and patterns of access to care leading up to the death, using a three delays framework and examination of both CoM and free text components of VA data. 23 By combining quantitative and qualitative findings, this study aimed to develop a three delays model specific to CVD in the Agincourt population, providing a basis for development and application of the three delays model to the final stages of chronic illness in other settings. --- METHODS --- Study design This study presents a mixed-methods analysis of VA data. Quantitative analysis comprised of descriptive analysis of CVD mortality rates and a three delays approach to analysis of circumstances of mortality indicators and free-text narratives of the final illness. Free text narratives were also analysed qualitatively. --- Study setting and participants This study was done in the Agincourt Health and Socio-Demographic Surveillance System site, in Mpumalanga province, rural North East South Africa. The HDSS covers an entire sub-district population of ~117 000 persons and has been stewarded by the MRC/Wits Rural Public Health and Health Transitions Research Unit since inception. 25 The population has been surveyed at least annually since 1992, with VA conducted on every death in the study population occurring in each preceding year. Previous studies have shown recall to be good up to a year after a death. 26 At the time of research, VA data from 1993 to 2015 was available. --- Quantitative analysis Measures Deidentified VA data were electronically stored and three elements extracted: Cause of death: derived using InterVA-5 27 InterVA-5 is the most recent probabilistic modelling approach developed to assign cause of death, using responses to VA interview questions. Detailed explanation of how cause of death is assigned is provided by Byass et al. 28 Two authors reviewed VA cause of death categories to select deaths related to CVDs. Included categories were 'acute cardiac disease' , 'stroke' , 'renal failure' and 'other and unspecified cardiac disease' . 29 After discussion among the authors, renal failure was included in our analysis as a CVD outcome because the main causes of chronic kidney disease are CVD risk factors-hypertension and diabetes. These exceed other causes, such as chronic glomerulonephritis or HIV. 30 Chronic kidney disease is especially likely to be related to CVD in older adults in this population, where CVD risk factors are highly prevalent and under-treated. 31 To minimise capturing non-CVD causes, cases of renal failure deaths in people aged under 40 years old were excluded. 32 InterVA-5 gives a likelihood of a death being due to a given cause; greater than 50% was required for case inclusion. Each death appears once in the analysis. --- CoM indicators The ten CoM indicators are closed questions added to the VA tool in 2012, which examine social and health systems determinants of death and social exclusion from access to care. 17 These indicators are based on a framework of proximate , intermediate , and distal determinants of health outcomes. 17 An 11th indicator was included alongside the ten CoM indicators, which asks whether the deceased was discharged while still ill. While not part of the CoM indicators, this is thought to be important to include, especially for chronic conditions such as CVD, as these patients may experience multiple cycles of accessing care. It may suggest poor quality of care. Delays in access to care were examined using the CoM indicators, including cases where death occurred between 2012 and 2015 . A person was classed as having experienced a delay if they had a positive response to the relevant indicator, with the exception of three , where a negative response was used to capture experiencing a delay. --- Free-text descriptions of the death and its circumstances The free-text sections of each VA are recorded by a trained lay field worker at the time of interview. Based on discussions with the respondent, most often the primary care-giver, this enables more nuanced details about circumstances around the time of death to be captured. --- Analysis All deaths recorded within the Agincourt population were included in the dataset. Systematic house-to-house visiting and detailed inquiry of every household member limits the risk of missing deaths. Descriptive data were generated using SPSS V. 25. 33 Mortality trends CVD deaths for each condition category, 5-year time period, sex and 10-year age group are described as agesex-time-standardised mortality rates. --- Three delays quantitative analysis To identify inclusion of cases in the three delays analysis, cases where free text and CoM data were not both available were first excluded. Free texts were examined to identify cases which were found dead at the scene or had died suddenly, with therefore no opportunity to seek care. These cases were also excluded. The remaining cases are referred to as those 'able to seek care.' Responses to CoM questions are described as total cases indicating they had experienced at least one delay across any of the CoM indicators, shown as a proportion of total cases able to seek care. Total cases for each specific delay are then shown as a proportion of cases that had experienced at least one delay across any of the CoM indicators. In addition delays extracted from the free text are analysed quantitatively, in a similar manner for that done with the CoM indicators, using data on the three delay coding extracted during the qualitative analysis . The free text was also examined to determine the first careseeking act, where this information was available. --- Qualitative analysis The free texts were analysed qualitatively to ascertain delays in access to care, using a combination of deductive and inductive approaches. A deductive approach enabled causes of delays to be extracted using a set of a priori codes based on the three delays framework of seeking, reaching and receiving care. An iterative process 23 enabled emerging codes to be incorporated until saturation was reached. Initial coding was conducted by JNLV, with codes reviewed by JD and AF. This method has been previously applied to the same dataset. 23 NVivo V.12 software was used for the qualitative analysis. 25 34 Construction of a three delays model Finally, a three delays model, specific to CVD in the Agincourt population, was developed through aggregating delays derived from the CoM and free-text analyses. --- Patient and public involvement We did not directly include public involvement in this study. The database, from which this study derives data, is used to inform a public engagement programme at the Medical Research Council/Wits University Rural Public Health and Health Transitions Research Unit . --- Open access --- RESULTS --- Mortality trends A total of 15 305 registered deaths occurred over 1 851 449 person-years between 1993 and 2015. Of these deaths, 9% were attributable to CVDs, corresponding to a crude CVD mortality rate of 0.77 per 1000 person-years. Standardised mortality rates by sex, age group and time period are displayed in table 1. The CVD mortality rate increased from 0.34 in 1993 to 1.12 in 2015 . Across all time periods, stroke was the dominant cause of CVD deaths and suggested a linear increase, responsible for 41.0% of all CVD deaths across all years . Mortality contributions of acute cardiac disease and renal failure were low across all years, although 2010-2015 saw an increase in acute cardiac disease deaths. Over the entire time period, the CVD mortality rate was higher in women . Mortality was higher in women for stroke and other and unspecified cardiac diseases . Mortality was higher in men for acute cardiac disease , there was no significant difference for renal failure . CVD mortality rate increased with age . --- Access to care Between 2012 and 2015, 495 CVD deaths occurred; of these 41 cases did not have both free text and CoM information available and were excluded from analyses. When considering all delays, there was no significant difference in numbers experiencing any delay between sexes using the CoM indicators or free text . No significant differences were seen in numbers experiencing a delay by age group on either CoM or free-text analyses. Significant difference was seen by condition on CoM, with 18.1% acute cardiac diseases , 48.2% stroke, 46.9% renal failure and 37.8% of other and unspecified cardiac disease cases reporting any delay . No equivalent significant difference was seen on the free text . --- Qualitative analysis Themes identified in each delay category are summarised below, with a non-exhaustive selection of quotes which have been chosen to illustrate themes. Each quote has been given by the field worker recording each case. --- Delays in seeking care Of the cases reporting any delay, almost half reported a delay in seeking care. However, specific reasons or explanations were rarely given, and lack of care-seeking was commonly expressed through phrases like 'nothing was done' or a refusal to seek care. Only one of the CoM indicators on seeking care-traditional medicine-was explicitly mentioned in the free text. Traditional medicine was often sought either before or after having sought 'western' care. Care was sought from a traditional healer following perceived failure of treatment given at a 'western' facility and vice versa. The doctors didn't say what was wrong. Treatment was given but nothing changed… He was taken to the private doctor. The doctor didn't say what was wrong… He was taken to the traditional healer. He went to traditional healer who gave him traditional medicine… Nothing changed. He then went to private doctor. Refusal to seek care was common among cases who did not travel to a facility. However, refusal appeared to be targeted towards hospitals rather than clinics. She was taken to the clinic… Treatment was given but nothing [changed]. After few days, she was worse. She refused to go to the hospital, she died at home. --- Delays in reaching care Reaching care was very rarely discussed in the free text. Two explicit delays were mentioned, which related to ambulances and dying on the way to healthcare. Other cases indicated a delay occurred-such as in cases where the intention to seek care had been made-but further detail was not provided. In cases where reaching care was discussed, this mostly centred around ambulance transport. Cases which mentioned calling directly for an ambulance themselves from home frequently experienced delays. An ambulance was called… It was late and when they arrive, they said that she is dead. In several cases, there were suggestions of an intention to seek care but which were unfulfilled. While the free text did not provide further insight into the reasons for this delay, this suggests there may be additional delays following an intention to seek care which were not captured in the free text. 2) with those who refused to seek care or initially did not seek care but later did, and those who initially sought care and later refused in subsequent cycles of accessing care. CoM, circumstances of mortality. --- Open access --- Open access She collapsed at the gate… She died after an hour before they can take her to healthcare. --- Delays in receiving care Delays in receiving care were reported in the free texts of most cases. This tended to focus on issues of communication between the clinician and patient or family, and being discharged while still ill. A repeated and persistent theme on receiving care was that the patient, family or caregiver were not told by the clinician what the problem was. At hospital he was admitted, oxygen, waterdrips and tablets was given but all the symptoms didn't change and the doctors didn't tell the family what was wrong. In many cases, the patient was reported to have been discharged while still ill, frequently leading to re-entering the three delays cycle. He was discharged at the hospital after six days and he was still sick. He was taken to another hospital after a day and he was admitted again… He was not getting better… He was discharged being sick. Referral between hospitals presented an opportunity for delays, such as lack of admission at the second hospital, being admitted but reporting that nothing was done, and/or being referred back to the first hospital without any treatment or diagnostic tests being performed. This again highlights the non-linearity of accessing care for CVDs, as patients re-entered a process of making decisions, travel and readmission. While she was at hospital she was taken to another hospital… To check her chest. She told them that they didn't do anything at the other hospital. They didn't admit her, she was sent back to the first hospital the same day. Other cases indicated a lack of alignment between hospitals involved in inter-facility referral, particularly when patients were sent to another hospital for an operation. In several cases, such patients were taken to the referred hospital only to not receive the operation they had travelled for. The doctor said she need to be operated [and] referred her to another hospital for operation, but she was not admitted and they said they are fully booked… She must wait until they have space. Many cases referenced elements of patient decision making which negatively impacted the receipt of care, therefore, contributing to delays in receiving care. A key theme in patient decision making was self, or family, forced early discharge. She was admitted and given treatment… The family asked doctors to discharge her but not given treatment as they promise to bring her back to hospital… She was not taken back to hospital… She died at home. Refusal to be admitted to hospital also emerged as a patient-led delay in receiving care. He was taken to the hospital. He refused to be ad-mitted… He died at home after a day of consultation from the hospital. --- First care-seeking act The first care-seeking acts identified from the free text are outlined in table 3. Cases were also categorised by hospital, non-hospital and delayed first care-seeking act. A total of 391 cases were able to seek care and therefore were included in the analysis. There were no significant differences in first care-seeking act by sex, age group or condition. Base=total applicable cases . 'Delayed' includes cases where initially no care was sought but later was, and those who never sought care. This differs from table 2, where the free-text delay 'at least one delay during seeking care' also encompasses cases who later refused to seek care after having previously sought it. Open access Three delays model A three delays model for CVD outcomes, developed through combining qualitative analysis of the free text and the quantitative CoM analysis, is shown in online supplemental figure A2. --- DISCUSSION While trends in deaths due to CVD have been the subject of previous studies, 35 36 this is the first known investigation combining both mortality trends and barriers to access to care for people who have died of CVD. In this lowincome rural South African population, we found that CVD deaths are increasing over time, with an increase in mortality rate from 0.34 to 1.12 between 1993 and 2015. Most of these deaths were due to stroke, and women overall suffered significantly more CVD deaths than men. Women experienced fewer ischaemic heart disease events than men, contrasting with knowledge from other settings that women are as likely to have these conditions as men. However, the findings do fit with evidence from elsewhere that importance of CVD in women may be underestimated 37 and underdiagnosed. 38 This difference could also be due to differences in care seeking behaviours among men and women, although this analysis suggests there was no difference in reporting barriers to care in the final illness by sex. To ensure validity of our findings, we used two methods to examine access to care for CVD in Agincourt, one using the CoM indicators and the other, free text. Care access was found to be generally good, with most cases seeking, reaching and receiving care. Nevertheless, many cases did report experiencing multiple and reinforcing delays in seeking and receiving care. Both the CoM and free text highlighted seeking care as a key delay in accessing care. Although only 6% of respondents said that the deceased doubted the need for medical care when explicitly asked in the CoM, most cases did not present to hospital or call an ambulance as their first care act, suggesting that the urgent need for care was not recognised. Early hospital-based care is a critical determinant of most successful interventions for CVD outcomes; 39 that the first health-seeking act for people who died of acute cardiac disease was often prehospital rather than hospital is therefore troubling. Not seeking hospital care in the first instance can also lead to a re-entry into the three delays cycle, presenting opportunities for the patient to experience further delays. Health literacy campaigns around symptom recognition and presenting to facilities able to deal with these acute conditions could improve not only the current access to care cycle but also future iterations, and have worked well in other settings. 40 However, this study highlights multiple supply-side factors which also drive barriers to care seeking, which likely combine and converge with demand-side issues. The South African public health system aims to provide healthcare that is free at the point of use for all. 41 Despite this, perceived costs were a major barrier to seeking care in most cases who experienced delays. Consistent reporting of prohibitive costs in Agincourt has also been shown elsewhere. 17 This may be due to poor information on availability of healthcare, or that non-medical direct and indirect costs are driving this barrier. Overall cost burdens of care have been shown to be low in Agincourt, but costs associated with transport and difficulty in obtaining income-related exemptions at hospitals limit the efficacy of state-provided protection. 42 Work must be done to further delineate the reasons for high reporting of prohibitive costs and institute appropriate solutions. Reaching care appeared to be a fairly minimal delay. Although free text analysis highlighted a further delay not captured in the CoM indicators, ambulance delay, the number of cases reporting this was minimal. Shortages of ambulances has been a key challenge in this area, but the Department of Health has invested in Emergency Medical Services including increasing ambulance numbers. 43 44 Nevertheless, given that many patients who need this service are not using it due to prior barriers at the stage of seeking care, investment in ambulance services may need to be accompanied by measures to reduce these prior barriers. Receiving care was a common barrier in both the CoM and free text analyses. Free texts highlighted many barriers not captured by the CoM; it is likely that the free text allows expansion on the barriers experienced in receipt of complex facility-based care that are not possible to capture in the smaller number of CoM fields for this delay, pointing to the complementarity of the two methods. Phrases like 'nothing was done,' and 'the doctor didn't say what was wrong' were ubiquitous. It could be that healthcare providers are not communicating effectively, or at all, with the patient and/or caregiver, and therefore this is an element of clinician-patient interaction which requires intervention. However, it could also be an indication that the clinician was not able to determine a diagnosis, potentially as a result of health system weaknesses . Poor quality of care is now a greater barrier to reducing mortality in LMICs than insufficient access to care, 45 so understanding where these issues are is vital. With either interpretation, these free text phrases suggest a wider issue of poor perception of care quality. Low expectations of care quality are prevalent in LMIC settings, including South Africa, which may reduce pressure on systems to deliver quality care. 46 Perceptions of care quality may also inform future care-seeking behaviours, as evidenced here by the frequent use of traditional healers after visiting a 'western' facility and feeling nothing changed or nothing was done, or by otherwise refusing to seek or delaying care-seeking. Refusal of care was common, and more so for hospitals than clinics. It could be hypothesised that care received in hospitals is perceived as poorer than in clinics; further supported by Open access the number of cases who sought non-hospital care in the first instance. Patient experience and perceived quality of care can impact utilisation of healthcare services. 47 Changing perceptions of hospitals as places to receive care and be treated, rather than places to die, requires both improvement in the health system to enable hospitals to provide quality care 48 and positive messaging to individuals and communities. Negative perceptions are particularly important to remedy in chronic conditions, which most CVDs are, where patients should have multiple opportunities to seek and receive care and are likely to accumulate experience and perceptions of care quality. Interpersonal skills, improving drug availability and technical care have been identified as priority areas to improve perceived quality of care in the adjacent Limpopo province; 49 aligning with the free text analysis in this study. However, evidence on perceptions towards care in South Africa is limited and warrants further investigation. This study explores application of the linear framework of the standard three delays model to CVD outcomes. However, several findings have highlighted the complexity in access to care for chronic conditions: delays in receiving care-such as being discharged while still ill-influence future care-seeking behaviours, and many cases enter multiple cycles. Delays may therefore multiply and reinforce each other over the course of multiple iterations of accessing care, even during the final illness as examined here. A linear model is oversimplistic for application to chronic conditions, which could be better represented as a cyclical model. This has been raised previously in applying the three delays framework to access to care in Burkina Faso and Indonesia. 50 --- Study limitations The non-specificity of InterVA-5 categories was a key limitation, particularly 'other and unspecified cardiac diseases.' This reduces the ability to analyse trends in specific CVD conditions and means it cannot be determined exactly what is driving a large proportion of this CVD burden. Renal failure was included due CVD risk factors being major causes of chronic kidney disease, especially in older adults. While we have aimed to minimise the inclusion of non-CVD-related causes of renal failure, we acknowledge this analysis may capture deaths due to other causes. Other factors influencing equity, such as migratory status and socioeconomic status, cannot be included due to lack of available data within VAs. That this study could not capture these factors limits the utility of the study for health planning from an equity perspective. Delays in receiving care may be under-reported: respondents may have had more direct experience of issues in seeking and reaching care, and so be more likely to report these, and people who died outside of facilities are also less likely to have experienced issues with quality of care, but this does not mean the issues do not exist. 17 Comparing proportions of cases experiencing each delay as calculated by the CoM versus the free text must be given careful consideration given the differing methods of data collection, but they do provide complimentary information amenable to synthesis. In comparing the CoM and free text, this study raises the question of breadth vs depth. The CoM is an invaluable tool for gaining a broad overview of delays in accessing care and are feasible to perform at scale. The free text, while able to pick up delays not captured in the CoM, is time-consuming to analyse with challenges of scalability. However, it has been shown that extracted word frequency counts from free texts can be used by machine learning for automated classification. 51 Insight gained from the free text could inform expansion or refinement of the CoM to enable a greater level of detail to be captured, particularly for assessing delays in receiving care. The CoM is continually evolving 52 and this analysis provides an empirical basis for revisiting them. --- Future research The three delays model developed here presents a basis for development and application to other settings to deal with access and quality issues in the final stages of chronic illness. This study also presents an opportunity to revisit and refine the CoM indicators. It also highlights several avenues for further research. First, the greater burden of CVD death among women requires further investigation in order to develop strategies to tackle this inequity. Prominent delays in accessing care also require further examination, particularly around high reporting of prohibitive costs, and on perceptions of care quality. --- CONCLUSION This study is the first known investigation combining both mortality trends and barriers to access to care for people who have died of cardiovascular. It shows the increase in deaths due to CVD between 1993 and 2015 in a rural region of South Africa. Given CVD in South Africa is now the second greatest cause of mortality after HIV/AIDS in South Africa, 3 there is a vital need for prevention and management strategies for these conditions. As women suffer a significantly higher CVD mortality than men, they should be a key focus of such strategies. While care pathways appear to be generally good for CVD, this study has highlighted issues around seeking care, including that many CVD cases first seek care outside a hospital setting during the final illness and receiving care. Strategies to improve access to care are therefore needed. This study highlights the complexity in accessing care for chronic conditions such as CVD; many cases enter multiple cycles and may experience multiple and reinforcing delays. While a three delays model provides a valuable framework, the potential need for a cyclical model for examining access to care for chronic conditions should be considered. --- Competing interests None declared. Patient consent for publication Not required. Ethics approval Approval from the University of Witwatersrand Human Research Ethics Committee to use MRC/Wits Agincourt Unit VA data had been previously obtained . Approval to perform secondary data analysis on VA data was obtained from the King's College London Research Ethics Committee . Provenance and peer review Not commissioned; externally peer reviewed. Data availability statement Data may be obtained from a third party and are not publicly available. Deidentified verbal autopsy data from Medical Research Council/ Wits University Rural Public Health and Health Transitions Research Unit 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 Cardiovascular diseases are the second leading cause of mortality behind HIV/AIDS in South Africa. This study investigates cardiovascular disease mortality trends in rural South Africa over 20+ years and the associated barriers to accessing care, using verbal autopsy data. Design A mixed-methods approach was used, combining descriptive analysis of mortality rates over time, by condition, sex and age group, quantitative analysis of circumstances of mortality (CoM) indicators and free text narratives of the final illness, and qualitative analysis of free texts. Setting This study was done using verbal autopsy data from the Health and Socio-Demographic Surveillance System site in Agincourt, rural South Africa. Participants Deaths attributable to cardiovascular diseases (acute cardiac disease, stroke, renal failure and other unspecified cardiac disease) from 1993 to 2015 were extracted from verbal autopsy data. Results Between 1993 and 2015, of 15 305 registered deaths over 1 851 449 person-years of follow-up, 1434 (9.4%) were attributable to cardiovascular disease, corresponding to a crude mortality rate of 0.77 per 1000 person-years. Cardiovascular disease mortality rate increased from 0.34 to 1.12 between 1993 and 2015. Stroke was the dominant cause of death, responsible for 41.0% (588/1434) of all cardiovascular deaths across all years. Cardiovascular disease mortality rate was significantly higher in women and increased with age. The main delays in access to care during the final illness were in seeking and receiving care. Qualitative free-text analysis highlighted delays not captured in the CoM, principally communication between the clinician and patient or family. Half of cases initially sought care outside a hospital setting (50.9%, 199/391). Conclusions The temporal increase in deaths due to cardiovascular disease highlights the need for greater prevention and management strategies for these conditions, particularly for the women. Strategies to improve seeking and receiving care during the final illness are needed.

Introduction Hispanics are the largest racial minority group in the United States . In 2020, about 61 million people of Hispanic origin were living in the country . By 2060, it is projected that the Hispanic population will increase to over 111 million people, a significant factor in U.S. population growth . Hispanics experience high prevalence of most chronic diseases . For example, the prevalence of diabetes among Hispanic adults is 22.6% vs. 11.3% in non-Hispanic Whites . Risk factors such as obesity, sedentary behavior, and poor healthy eating, are a major concern for Hispanics and significantly contribute to chronic diseases like type 2 diabetes. Hispanic adults and youth aged 2-19 years have the highest prevalence of obesity in the US . Hispanics are typically less likely to be aware of their conditions, receive instructions from their physicians to adopt lifestyle modifications, or self-monitor their conditions . If Hispanics continue to experience poorer chronic disease outcomes, the projected demographic change in the U.S. population will magnify these health inequities. Pennsylvania is home to 1 million Hispanics . Vibrant Hispanic communities can be found in mid-size cities and towns like Lebanon in Lebanon County, and Reading in Berks County . Hispanics in both communities experience high levels of known socioeconomic factors associated with increased rates of chronic diseases . In particular, more than onethird of Hispanics live below the poverty line, over 15% points above the U.S. average, and between 13.6% and 25.2% in the target communities had limited English proficiency compared to only 8.5% in the US and 4.1% in Pennsylvania . Several community health needs assessments conducted in the region, including Hispanic and non-Hispanic populations . prioritize chronic disease, obesity, sedentary behavior, and inadequate nutrition as health topics requiring immediate action. For example, adults in Lebanon reported higher percentages of inadequate fruit and vegetable consumption, heart disease, high blood pleasure, and cholesterol than the corresponding state and national averages . Both communities also reported high percentages of adults with obesity and high rates for heart disease and cancer deaths. The size and rapid growth of the Hispanic population in these two communities, coupled with their high rates of socioeconomic disparities and chronic disease, offers considerable reason to focus on this population's chronic disease prevention and management. Established in 1999, Racial and Ethnic Approaches to Community Health is the cornerstone program aimed at reducing racial and ethnic health disparities at the Centers for Disease Control and Prevention . In 2018, CDC funded a new 5-year cycle of grant recipients to reduce health disparities among racial and ethnic populations with high levels of chronic disease . Through REACH, recipients and their community partners plan and carry out local, culturally appropriate programs to address preventable risk behaviors leading to chronic diseases in socioeconomic deprived communities. Since 2018, Better Together REACH, our initiative, has leveraged strong community collaborations to implement locally-tailored practice-and evidence-based strategies aimed at increasing physical activity opportunities, healthy nutrition programming, and diabetes prevention programs in Lebanon and Berks counties. The overall objective of this report is to describe the work-in-progress and lessons learned from our REACHsupported initiatives. In particular, we describe the implementation activities and initial results of three strategies to improve physical activity , three strategies to improve healthy eating , and one strategy to improve community-clinical linkages . For each program presented in this community case report, we also summarize several activities that we carried out to incorporate the needs and preferences of our Hispanic population in order to make these programs culturally appropriate . We also introduce our community-academic partnership and the conceptual model guiding our work. --- Context: community-academic partnership The Better Together initiative was first conceived on October 2016 when a community health advocate in Lebanon found that many local organizations were offering community health and social services but operating independently. With a longstanding history in prevention and community engagement, Penn State PRO Wellness partnered with these organizations to bring them together in a community-academic partnership. Since then, over 60 local organizations are collaborating to break down silos and share a common agenda: address health disparities, improve the quality of life and wellness of residents, and promote healthier environments where people live, work, learn and play. The original initiative was organized around four action themes, including healthy food access, physical activity, mental and behavioral health, and family and community engagement. These themes were perfectly aligned with the goals of the overall REACH program. Our team leveraged these existing collaborations, shared interests, and the expansion of Penn State Health into Berks County to apply for the REACH award. Lebanon and Berks counties are geographically close and share similar demographics and identified community needs, creating areas for cooperative integration of leading organizations and their programming within these communities. Currently, Better Together REACH benefits from the active involvement of many local and statewide organizations with established connections with the Hispanic community . Convening individuals and leaders from many layers of the community helped to ensure discussions were representative and inclusive of the overall community voice. --- Conceptual model Figure 2 shows our model, which fully incorporates the inputs and activities of Better Together REACH to evaluate the proposed outcomes related to physical activity, nutrition , and community-clinical linkages. Importantly, we have identified short, intermediate, and long outcomes that align with the overall REACH objectives, the Healthy People 2020 goals pertaining to chronic disease prevention, our coalition's capacity, and the needs identified from multiple CHNAs. Beyond the project period, our coalition expects to achieve the following long-term outcomes: increased purchasing of healthier foods and increased levels of physical activity , reduced health disparities in chronic conditions and risk factors , and improved health outcomes . --- Programmatic components Since 2018, our Better Together REACH partnership has implemented several locally-tailored practice-and evidence-based strategies aimed at increasing physical activity opportunities, healthy nutrition programming, and diabetes prevention programs in Lebanon and Berks counties . In this section we describe the work-in-progress and lessons learned from our REACHsupported initiatives. --- FIGURE Better Together REACH community-academic partnership. --- Physical activity City redesigning and master planning We have partnered with the City of Lebanon and the Lebanon Valley Conservancy to improve the walkability of the city. One of our first initiatives was the launch of a bilingual mobile app to report issues with the City's streets and recreational areas, including uploading pictures and pin locations. Electronic reports from residents were sent to the City's department of public works for their attention. We also partnered with the Community Health Council of Lebanon to seek WalkWorks designation for a walking trail in downtown Lebanon. WalkWorks is a program run by the Pennsylvania Department of Health and Pennsylvania Downtown Center to recognize local efforts that promote safe walking routes; offer social support through guided, community-based walking groups; help schools develop walk-to-school programs; and address local policies to increase safe walking routes. A 1.32-mile route was awarded WalkWorks designation, providing safe connection to six everyday destinations. This was the first route in Lebanon County with such a designation. Our team also conducted a survey to assess physical activity practices, use of community green spaces, and specific areas that residents would like to see improved. The survey was developed with input from community partners, including Hispanics, and was delivered in Spanish and English to better reach our community. We leveraged these survey results and partnered with the City of Lebanon to help design, enhance, and restructure and urban alley into the Liberty Trail Park, Wengert Memorial Park, and the Borne Learning Trail through bilingual interactive signage and trail restoration. One pedestrian mile was enhanced and another mile of multi-use trail was created as part of the Borne Learning Trail. The Wengert Memorial Park will serve as a trailhead and connector to 14 miles of the Lebanon Valley Rail Trail. In the park will be a bike safety traffic garden, one of its kind in Pennsylvania, an area for youth to learn traffic rules and practice these skills in a secure and fun environment. Furthermore, we have worked toward installing seven automated eco-counters to monitor pedestrian traffic patterns in Lebanon and Reading. The locations for the measurements were chosen with guidance from community stakeholders to assess the need for new or expanded crosswalks, sidewalks or trails. Our goal is to develop new or enhance existing 11 miles of routes that connect everyday destinations and improve walkability. Our team also developed an interactive map for community members to identify local resources for physical activity , indoor spaces where the community can exercise , and other local programming, including emergency food providers, food assistance programs, farmers markets, local orchards, and community gardens . --- Nature Rx We partnered with Berks Nature, a nonprofit conservation organization, to implement the Nature Rx program to motivate adult and young individuals to spend time outdoors engaging in some sort of physical activity, such as walking, hiking, or biking. We first developed a bilingual marketing campaign to increase awareness of Nature Rx including examples of fun outdoors activities. We also adopted Goose Chase , a scavenger hunt mobile app available in Spanish and English, to encourage enrollment in Nature Rx and reward participants for spending time outdoors. We invited those who downloaded the app to complete a brief survey to measure their levels of physical activity and engagement with outdoor spaces. Participants reported spending 5 days outdoors for ≥10 min and spent an average of 1 h and 56 min outdoors each week. In a typical week, participants reported spending around 4 days a week engaging in moderate physical activity with an average of 1 h and 41 min per week. On average, participants increased exercise duration by 7 min after completing the program. --- Lebanon Bicycle Recycle In partnership with the Lebanon Valley Bike Coalition, an organization that works to educate, advocate, and improve physical activity through cycling, we helped launching the LBR in 2021. Meetings with local leaders confirmed that cycling is important in Lebanon for physical activity and transportation to work and school. LBR is a program that collects used bikes, repairs them, and returns them back to community members who need them. To date, over 125 bikes have been given to community members with an emphasis on youth and families . Further, LBR has worked with volunteers to train them to repair bikes, including a cohort of middle-school students who came after school to receive training. Over 280 hours of volunteer time have been given to the program. Also, LBR has partnered with the Migrant Education Program and the Lebanon County Christian Ministries to provide bikes to their members to help with their local transportation needs. For instance, we developed a bilingual marketing campaign to promote this program and education to encourage physical activity through cycling, including how to do it safely in the city. Our REACH team is conducting a survey with bike recipients to learn about their levels of physical activity and social needs. --- Healthy eating Farmers Market Nutrition Program The Pennsylvania Department of Agriculture provides those who are eligible four vouchers of six dollars each to spend from June through November on seasonal fruits and vegetables grown by Pennsylvania farmers. Distribution and redemption rates of these FMNP vouchers are lower in Lebanon when compared to neighboring counties. Lebanon Family Health Services , a local Women, Infants, and Children office had a 36.7% redemption rate of vouchers in 2019; that number dropped to 25.6% in 2020 as a result of the pandemic. The goal of our FMNP initiative is to increase voucher redemption rates among WIC beneficiaries served by LFHS and senior residents. To promote FMNP awareness and voucher redemption, LFHS holds the Latino Health Fair, a one-day annual event that includes live demonstrations on healthy cooking and bilingual nutrition education. In 2019, we conducted a brief evaluation with WIC beneficiaries to survey barriers participants faced in using FMNP vouchers like transportation to voucher-eligible markets, variety of produce, and limited locations of voucher-eligible markets . The survey also gauged the participants' interest in additional offerings to improve voucher redemption such as healthy cooking or nutrition education. To adhere to COVID-19 safety protocols, our FMNP initiative was adapted into a "grab bag" program to limit physical contact between participating families and staff while offering the same quality produce in convenient packaging. Informed by our 2019 survey, healthy recipes for cooking with seasonal fruits and vegetables were included in grab bags in both English and Spanish. These recipes included common Hispanic dishes. To further evaluate this new initiative, in fall 2021, we started inviting FNMP participants to complete a brief survey --- FIGURE Top row: sta making inventory of repaired bicycles to be distributed to youth and families through the Lebanon Bicycle Recycle ; farm stand for FMNP and Veggie Rx participants in Reading's Penn State Health St. Joseph hospital ; and community health worker conducting screenings for diabetes prevention and management programs at a COVID-vaccination site in Reading . Bottom row: billboard displaying our marketing campaign for breastfeeding in Lebanon ; bilingual breastfeeding education delivered by a community health worker at a WIC o ce in Lebanon ; and interactive map showing local resources for physical activity , healthy eating , and diabetes prevention programs in Lebanon . within 24 h of purchasing the grab bag and a second survey was sent within 7 days. Also in 2021, we started evaluating FMNP voucher redemption among seniors in Lebanon to identify potential barriers and areas of opportunity to increase redemption rates and ultimately, the consumption of fresh produce among this vulnerable group. One of the indentified barriers to voucher redemption was low access to vendors who participate in FMNP. In 2022, we worked with a new farmer to obtain the registration as FMNP vendor and we then partnered with the Chestnut Street Community Center in downtown Lebanon to host a new farmers market. CSCC is the location for two Hispanic churches, two non-profit organizations, an overnight shelter, and a new community garden. This location is ideal for expanding FMNP as it is along a popular bus route and walking distance to a WIC office and a local elementary school. Importantly, this new farmers market is the only one in the City of Lebanon that accepts FMNP vouchers provided to both WIC and senior participants. --- Veggie Rx Veggie Rx is a fruit and vegetable prescription program that allows clinical staff at Penn State Health St. Joseph in Reading to prescribe vouchers for fresh produce to patients with chronic, diet-related health conditions . The program was launched in collaboration with the United Way of Berks and The Food Trust. The goal of Veggie Rx is to increase access to and consumption of local fresh produce among patients with diabetes. Clinic providers prescribe produce vouchers , to Supplemental Nutrition Assistance Program -eligible patients based on their patients' household size, that can be redeemed at local farmers markets and participating retailers. During 6-month follow-up and group education sessions with a dietitian, patients fill out a nutritional assessment for daily fruit and vegetable consumption and discuss situations that may have changed their nutritional habits. To improve opportunities to redeem vouchers, a farm stand by a local farmer is located within the St. Joseph's downtown campus. This initiative is also assisted by one of our REACHfunded bilingual community health workers . A pilot study with 97 adults with type 2 diabetes [hemoglobin A1c ≥ 7.0% and body mass index ≥ 25 kg/m 2 ] showed a reduction in HbA1c post-program . Reduced HbA1c was associated with higher voucher redemption rates and a change in diabetes medications . There were no associations with BMI, but changes in blood pressure was positively associated with Veggie Rx voucher redemption. Given the initial success of Veggie Rx in Reading, our team is expanding this initiative to Lebanon in partnership with Union Community Care , a local federally-qualified health center. We anticipate enrolling 30 UCC patients into the Veggie Rx program in the next year. --- Bilingual breastfeeding education and workplace breastfeeding policy We partnered with LFHS, the local WIC provider in Lebanon, to provide bilingual breastfeeding support to expectant or new mothers. With REACH funds we hired a full-time CHW and trained her in the provision of professional lactation care. Since 2019, our CHW has provided one-to-one and group education to 456 pregnant and postpartum women at the WIC clinic, local hospitals, or by phone in both Spanish and English. To abide by COVID-19 protocols and guarantee the safety of our CHW and WIC recipients, all counseling sessions were moved online. To better support this initiative, our REACH program developed and disseminated a culturally-relevant marketing campaign about breastfeeding support across local healthcare organizations and the community, including media advertising in billboards and public transit . Partnering with the Lebanon Valley Chamber of Commerce, we also conducted a survey to determine what breastfeeding policies are in place at Lebanon County's businesses. Our team also conducted a series of focus groups with women who have returned to work while breastfeeding to learn how they are or are not supported through workplace policies. We found that polices supporting breastfeeding in workplaces were not consistently implemented. --- Community-clinical linkages Community health worker-delivered diabetes prevention program We partnered with the Langan Allied Health Academy-CHW Training Institute to recruit and train bilingual CHWs for work in community and outpatient/inpatient settings. To date, the Institute has trained 11 class cohorts and over 120 CHWs. The training provides comprehensive education in 15 areas including cultural competency, health literacy, patient navigation, and chronic disease prevention and management. Our project hired three local bilingual CHWs to conduct diabetes screenings at community events and provide DPP classes. Our CHWs completed their training with the Institute as well as a CDC-certified lifestyle coach training. At present, our CHWs have provided CDC-approved DPP to 204 residents at three clinical facilities in Lebanon and Berks counties. We have worked with LFHS to provide bilingual DPP in Lebanon. Serving two cohorts a year, 60 people have participated in this program. In 2022, REACH also supported the offering of DPP at Union Community Care , a federally qualified health center in Lebanon. We also supported the CHW at Family First Health, another federally qualified health center in Lebanon, to become a certified lifestyle coach in order to offer CDC-approved DPP. Each of these locations, LFHS, UCC and Family First, offers bilingual DPP classes, including in-person and Zoom options. --- Discussion Better Together REACH has made significant progress in addressing health disparities related to chronic disease prevention in the Hispanic-majority communities of Lebanon and Reading. Of note, we are actively promoting existing walking and bike routes that connect everyday destinations and supporting the planning and designation of new routes , which are evidence-based strategies to improve physical activity . We are also expanding access to affordable and nutritious food to alleviate food insecurity among low-income families and creating bilingual breastfeeding education programming to support WIC beneficiaries in their preferred language. To address critical community-clinical linkages, we are expanding access and referrals to DPP offerings by training local, bilingual CHWs to connect at-risk individuals with new and existing programs. CHWs are trusted voices in the community who can support communityclinical initiatives for chronic diseases . Our initiatives are promoted throughout strong community networks with culturallyrelevant marketing campaigns. A good number of our annual objectives are being met but much work lies ahead as we recover from the pandemic . The pandemic impacted our coalition's ability to deliver activities as many were planned as in-person events or large community gatherings. However, the limitations created by the pandemic offered opportunities to innovate, evidenced by our grab bag program and the virtual breastfeeding education. Other REACH programs can employ similar approaches to support their priority populations with needed chronic diseases prevention programming during these challenging times. Through our established community-academic partnership, we will continue implementing locally-tailored practice-and evidence-based strategies to address chronic disease disparities in Lebanon and Reading. These strategies were also tailored to respond to the cultural preferences and needs of Hispanics in the two target communities. Our next program evaluations will incorporate data from novel techniques including automated ecocounters to monitor pedestrian traffic patterns. This approach is promising because it incorporates easy-to-use technology into community-level interventions and involves community members in making decisions about where to create new or expand trails. This level of community engagement and shared decision-making is critical to meet our programmatic goals related to physical activity, including the creation of a bicycle and safety garden and mobile bicycle repair unit, as well as a park assessment initiative. Future program evaluation data will also be collected from new initiatives impacting nutrition and breastfeeding goals. For example, in collaboration with the local Chamber of Commerce we will champion a "Breastfeeding is Welcome Here" campaign among Lebanon employers after assessing existing policies and practices. Findings will inform the creation of resources and training to support breastfeeding practices at local worksites. We will also increase the number of worksites with implemented health nutrition standards by working with local food pantries to develop and evaluate placement and behavioral design initiatives. Our team has always been looking for new community partners to innovate our programming. For example, with the lessons we have learned in the past 4 years, we look forward to expanding our DPP beyond healthcare settings and directly into local employers. We have partnered with the largest poultry producer in the region to offer an online version of DPP for their employers. Almost 90% of the employees of this company are Hispanics, have limited English proficiency, and reside in the counties of Lebanon and Berks. This partnership will offer expanded health services to those who need them most. If this online DPP is successful, it would be replicated with other employers in our catchment area. Also, our team is continuing to refine programmatic data collection and measurements, including using geographical information mapping to evaluate our activities and provide timely feedback to our community partners. We will continue monitoring our coalition's capacity and updated CHNAs to revise our projected outcomes, but always aligned with practical strategies for culturally competent evaluation . Although not considered in the present project, there are several promising chronic disease prevention and management interventions that others working with Hispanic populations could consider implementing. For example, employerbased walking challenge campaigns, obesity reduction education at community venues , and CHWled physical activity programs . These interventions have shown broad reach of Hispanic individuals, good implementation feasibility, and effectiveness to improve physical activity and reduce obesity . The national goal of achieving health equity cannot succeed without eliminating Hispanic health disparities and engaging communities to seek and disseminate culturally-relevant solutions. Together with its partners, Better Together REACH is working toward that goal and on capacity building with the aim of continuity and sustainability of this initiative post-award. Intervening in areas facing high chronic disease health disparities, like Lebanon and Reading, leads us to develop a communitycollaborative blueprint that can be replicated across Hispanic communities in Pennsylvania and the United States. --- 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 the Penn State Institutional Review Board. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual for the publication of any potentially identifiable images or data included in this article. --- --- --- 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. --- Author disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of CDC or NIH.

Background: Hispanics in Lebanon and Reading, Pennsylvania, experience high levels of socioeconomic and health disparities in risk factors for chronic disease. In , our community-academic coalition "Better Together" received a Racial and Ethnic Approaches to Community Health (REACH) award to improve healthy lifestyles. This report describes our work-in-progress and lessons learned to date from our REACH-supported initiatives in Lebanon and Reading. Methods: For the past years, our coalition has leveraged strong community collaborations to implement and evaluate culturally-tailored practice-and evidence-based activities aimed at increasing physical activity, healthy nutrition, and community-clinical linkages. This community case report summarizes the context where our overall program was implemented, including the priority population, target geographical area, socioeconomic and health disparities data, community-academic coalition, conceptual model, and details the progress of the Better Together initiative in the two communities impacted. Results: To improve physical activity, we are: ( ) creating new and enhancing existing trails connecting everyday destinations through city redesigning and master planning, ( ) promoting outdoor physical activity, ( ) increasing awareness of community resources for chronic disease prevention, and ( ) supporting access to bikes for youth and families. To improve nutrition, we are: ( ) expanding access to locally-grown fresh fruit and vegetables in community and clinical settings, through the Farmers Market Nutrition Program to beneficiaries of the Women, Infants, and Children (WIC) program and the Veggie Rx to patients who are at risk for or have diabetes, and ( ) providing bilingual breastfeeding education. To enhance community-clinical linkages, we are training bilingual community health workers to connect at-risk individuals with diabetes prevention programs. Conclusions: Intervening in areas facing high chronic disease health disparities leads us to develop a community-collaborative blueprint that can be replicated across Hispanic communities in Pennsylvania and the United States.

INTRODUCTION In her dissertation, the author analyzed the experience of four wives who committed homicide in response to domestic violence using the feminist radical perspective and Bourdieu's symbolic violence theory. The author's standpoint is that women who commit homicide against their partners in response to domestic violence are victims. In-depth interviews of four informants support this standpoint at women's prison in Bandung and the author's analysis on structure, culture, and process [1]. The feminist radical perspective as the overall theoretical umbrella discusses patriarchal society and sexuality. At the same time, Bourdieu's symbolic violence theory, with the concept of habitus, space, and modal composition, is used to explain different forms of social domination. Analysis of the experiences of four women shows that they are victims of domestic violence, even though they are perpetrators in the perspective of the criminal law. Radical feminists show that the domestic violence they experience is a result of the patriarchal society, where the husband holds higher position of power within the family than his wife. Victimization occurs because of domination. Variations in dominance occur, which is determined based on Bourdieu's Theory [2] . The results are Domination and Capital Typology Matrix, consisting of fully dominated-fully-capital; partially dominated fullycapital; partially-dominated non-capital; Non-Dominated Non-Capital; and Dominated -Capital Irrational. The predominant category observed is the Dominated-Capital Irrational form. This dominance occurs among informants who hold a space not to be dominated due to their ownership of capital. However, in certain situations, they still experience domination . This power dynamic may lead them to seek out new partners as a possible solution. There are Four Typology Models of Domination and Capital based on the experiences of women who have experienced victimization. In radical feminist studies and criminology, the punishments given to them are very diverse. For example, not all sentences use the Law on the Elimination of Domestic Violence No. 23 of --- Diagram 1. Typology of Domination and Capital Meanwhile, when referring to the husband's role , it is characterized by starting first with the perpetrator. Therefore, it is safe to state that this type of crime is typical, where the perpetrator is also the victim of the crime. Even though they will still receive punishment for the crime, the author argues that the form of punishment will be different, such as rehabilitation or social work. This research aims to provide input to the proposed policy on punishment for women who commit homicide in response to domestic violence. The policy should look at rehabilitation measures, lenient treatment, and protection programs. The typology of domination and capital provides the baseline data to develop the appropriate punishment policy for them, which is the novelty of this research. --- LITERATURE REVIEW --- Feminist Criminology Theory on Women Who Kill Feminist theory is developed as a new branch of science about women to provide a system of ideas about human life that depicts women as objects and subjects, as actors and the knowledgeable individuals. Feminism refers to ideas generated primarily by women to change their self-awareness and change society. Feminists offer a perspective on social life from the point of view of seeing women as a disadvantaged social group. For example, in the context of women who commit homicide in response to domestic violence, appropriate punishment is yet to be considered. Feminist criminology is concerned with the treatment of women in the police, courts, and prison systems. Contrary to popular stereotypes, the study of women's experiences in the criminal justice system is unfavorable. Therefore, feminist jurisprudence is looking for the truth behind the power of law formed in a positivistic way [3]. In her work, "Crime and Criminology: A Feminist Critique," Carol Smart criticizes the lack of treatment of female perpetrators in mainstream criminology and ignores the victim's experience as a victim [4]. Feminist criminology focuses on women as perpetrators of crime, women as victims of crime, and ` 2012 women as workers in the criminal justice system. Renzetti interprets feminist criminology as a paradigm of study and explanation of perpetrators of crime and victimization, including institutional responses to fundamental gender issues [5]. For this reason, to influence the making of public policy, we need scientific knowledge. Feminist legal theory exposes how the legal system and practices dismiss women's experiences with crime and fail to consider them during the legal proceedings that women endure. For instance, the sentences imposed on women who commit homicide in response to domestic violence may not receive fair treatment from the criminal justice system compared to men who commit similar offenses. The objectives of feminist legal theory involve comprehending women's experiences, investigating whether laws and legal establishments provide equal treatment, and identifying potential reforms. According to this theory, the law sees and treats women the way men see and treat women. Substantively, the perspective of state policy in formulating experience is the same as the way men formulate experience [6]. The feminist legal theory also seeks to analyze and improve traditional legal theory and practice. It focuses on how laws are structured to deny women's experiences and needs. In addition, feminist jurisprudence opposes and reflects the demand for women, regardless of race, class, age, ability, to be recognized as an equal group by a social contract supported by existing laws and legal systems [7]. This article examined the policy of punishing women who commit homicide as a response to domestic violence in Indonesia. However, this does not guarantee that all women in different social conditions are powerless to challenge the law, patriarchy, the legal context in politics in defending their cases. --- Patriarchy and Sexuality in Radical Feminism to Understand Husband's Domination over Wife For radical feminism, patriarchy and sexuality are the two main concepts in understanding male domination over women. The term patriarchy was initially used by Max Weber to describe a particular socio-political system in which a father, due to his position in the household, can dominate members of his extended family network and control the economical production of the kinship unit. Meanwhile, one of the first radical feminists, Kate Millet, argued that patriarchy was brought by the idea of control and culture by men, not limited to the arena of kinship, but all of human life, such as economics, politics, religion, and sexuality [8]. Meanwhile, in other radical feminist writings, they proposed three universal types: biological maternal care, marriage-based families, and heterosexuals. Shulamith Firestone's argument, for example, said that patriarchy is based on biological factors where only women can conceive and give birth [9]. While some argue that the universal phenomenon underlying patriarchy is not related to women being biological mothers, but rather a social institution of the family based on certain types of marriage. According to Bouchier, for example, marriage is an "institutional source of real exploitation" [10]. Meanwhile, Marilyn French, who researched the history of patriarchy, said that the root of patriarchy is humans living in harmony with nature. They see themselves as a small part of a more extensive system. She stated that early society was a matricentric society the roles that were passed on involved feminine traits, such as bonding, sharing, and harmonizing with nature [11]. In its development, which is characterized by an increase in population, the food supply becomes limited. As a result, humans slowly began to distance themselves from nature. Humans want to control nature, but in turn, it causes them to be alienated from nature. This alienation then encourages the traits of "divorce," "hostility," "fear," and "hatred." These negative feelings encourage men desire to control nature and women, as women are often viewed as a part of nature because of their reproductive function [12]. The use of power by men over women is found in the public-structural and ideological contexts in work, education, the media, and beyond. Besides, patriarchy is also significant at the personal level, , in the private world of intimate relationships between men and women. As Mary Marynard said: "Politics occurs ` 2013 within families and between individuals when one person seeks to control or dominate another, It is in the personal and personal context that women are especially vulnerable to male power" [13]. As Sylvia Walby points out, from this point of view, the slogan, "the personal is political" Applies. The point is that because male domination exists not only in the public arena but also in the very private arena of life, women struggle to abolish subordination not only outside the house but starting inside the house. For instance, the question of who is responsible for performing household tasks or who has the right to interrupt others during casual conversations is regarded as a component of the system of male dominance. [14]. Feminists have taken over some ideas about sexuality and patriarchy from other schools, but criticisms of the use of these concepts and the fundamentals of radical feminist thought have been raised. First, this view emphasizes the universality of women's conditions regardless of the context of time and place. Thus, various things are not considered, such as the differences among women themselves , which give a different context to the condition of women. Second, the fixation on the context of sexuality does not allow people to see other factors that affect women's lives, such as the development of capitalism or the dynamics of political economy, or the influence of all of them on women's work. Third, analytically, they do not explain why men need to dominate women. Generally, the explanations given tend to refer to male biological drives [12]. --- METHODOLOGY This research on punishment policy for women who commit homicide in response to domestic violence was feminist research. The British sociologist, Liz Stanley, said that feminist research is research done by women and oriented to defending the interests of oppressed women by male domination [15]. The partiality of the author's standpoint was firmly on the fact that women who commit homicide in response to domestic violence are victims, and therefore imprisonment is not appropriate for them. This research used a mixed approach of qualitative and quantitative methods. By adopting a qualitative methodology, the author investigated the experiences of women. This approach permitted an emotional connection, as it did not require the author to adopt a neutral stance, as was typical in positivist thinking to attain the "objective" truth. However, on the contrary, the author's personality was brought to the surface to attract information. The quantitative approach in this research was used in the early stages by conducting two surveys of respondents, the first to students and lecturers of Criminal Law at a State Universities, and then the second survey was conducted to members of MAHUPIKI , APVI , and ASWGI . As mentioned earlier, this was follow-up research of the author's dissertation, which proved that women who committed homicide in response to domestic violence were victims. Therefore, imprisonment was not appropriate, and a particular policy was needed. In the positive law, women who commit homicide regardless of the reason are considered guilty. However, from a radical feminist perspective, women who commit homicide in response to domestic violence are viewed as victims and require alternative sentencing. Then, to achieve the research objective, which is to provide input for the proposed legal policy related to the punishment of women who commit homicide in response to domestic violence, the author also collected secondary data using the online desk review method. With this method, the author conducted a literature study on textbooks, journals, or institutional reports that the author considered relevant. The data obtained were punishment policies for a similar topic in various countries and used as a comparison. The following is an illustration of data sources and research methods. ` 2014 --- FINDINGS Why women commit homicide in response to domestic violence? -Unstructured interview . Issue 1: A woman commits homicide in response to domestic violence is not a crime Informants' opinions on the above issue were grouped into two categories of opinions as follows: 1. A woman who commits homicide against a man in response to domestic violence shall receive no punishment. Why is this case? The majority of the informants who hold this view stated that killing was the most effective means of avoiding victimization within the family. By murdering their husbands or partners, they believe they can escape the violence inflicted upon them. --- 2. Women who commit homicide in response to domestic violence indicate that they are no longer strong enough to accommodate their suffering from violent treatment obtained from their husbands or partners. Public's perception of homicide by women in response to domestic violence -Survey UPN Veteran, Faculty of Law Students and Unstructured interview Several issues to describe how society perceives homicide by women in response to domestic violence are: 2) A woman who commit homicide against a man who abuses her is seen as a crime; 3) Many people argue that "a woman who commit homicide against a man in response to domestic violence does not deserve punishment"; 4) Women who commit homicide in response to domestic violence should not be solely held responsible for acting in self-defense. Issue 2: A woman who commits homicide against a man who abuses her is seen as a crime 34.4% of the respondents agreed with the statement. Meanwhile, only 29.3% disagreed, and the rest were hesitant. The following are the results of the author's interview with one of the informants: The statement implies that murder is a criminal act regardless of who kills and the reasons for killing; therefore, a woman who commits homicide against a man who abuses her is seen as a crime. --- Issue 3: Many people argue that "a woman who commits homicide against a man in response to domestic violence does not deserve punishment" There were 39% of respondents agreed with the statement. Meanwhile, 48.8% disagreed, and the rest were hesitant. The following are the results of the author's interview with one of the informants: The statement implies that women who commit homicide will be free of all charges, and I cannot fully agree with that. After all, a person's life was taken away, and the act of killing is morally wrong regardless of its reason. However, punishment should be appropriate, taking into account the reasons and context. Therefore, the statement shall be changed to, "a woman who commits homicide against a man in response to domestic violence does not deserve maximum punishment." In other words, the criminal justice system shall consider the socio-psychological aspects when giving sentences to women who commit homicide in response to domestic violence. ` --- Issue 4: Women who commit homicide in response to domestic violence cannot be solely blamed as they are acting in self-defense. 85.4% of respondents agreed with the statement. Meanwhile, 9.8% disagreed, and the rest were hesitant. The following are the results of the author's interview with one of the informants: I tend to agree with the statement that she cannot be solely blamed, there is also a plea mechanism to understand how she became a victim. Domestic violence does not happen once but throughout the marriage. When this violence continued, some women finally fought back, but some women did nothing about it. Therefore, we shall explore the self-defense plea because we understand the social and psychological context. While in some cases, extreme emotions may cause someone to unintentionally cause harm while defending themselves, leading to murder. --- Rehabilitation measures for women who commit homicide in response to domestic violence -Survey UPN Veteran, Faculty of Law Students and Unstructured interview An issue raised in this topic is whether rehabilitation measures are necessary for women who commit homicide in response to domestic violence. The given statement to the respondents is 5) Rehabilitation measures are a must for women who commit homicide due to domestic violence to recover from past trauma. It turns out that all respondents agreed with the statement. Issue 5: Rehabilitation measures are a must for women who commit homicide due to domestic violence to rcover from past trauma As mentioned earlier, all respondents agreed with the statement, and one of the informants said: Firstly, we have to see that she is a victim, which means protection and rehabilitation measures shall be in place because her act of murder is not premeditated. In this context, the victim usually has a psychological trauma, conflicting over being a victim, yet she committed a murder. Often, domestic violence has a layered psychological impact on the victims. ` 2017 The need for special treatment and protection for women who commit homicide in response to domestic violence -Survey UPN Veteran, Faculty of Law Students and Unstructured interview Several issues raised in this research were to describe the need for special treatment and protection for women who commit homicide in response to domestic violence. These issues are: 6) Judges presiding over cases involving women who commit homicide in response to domestic violence are required to include expert testimony and consider it during the proceedings. --- 1) Judges who decide cases in court against women who kill in response to domestic violence must adhere to the protection principle for domestic violence victims. 2) The State shall provide more protection and attention to women victims of domestic violence. 3) The positive legal system should prioritize the protection of women victims of domestic violence. 4) There is a high tendency to believe that women who commit homicide in response to domestic violence shall be considered victims and do not deserve punishment. Can this be accepted and realized in our criminal justice system? 5) The State should pay more attention and protect women victims of domestic violence. 6) Judges decide cases in court against women who commit homicide in response to domestic violence. 7) Should you agree that women who commit homicide in response to domestic violence must be given special legal treatment, what are the underlying reasons? Issue 6: Judges presiding over cases involving women who commit homicide in response to domestic violence are required to include expert testimony and consider it during the proceedings Of the respondents, 97.8% agreed with this statement. An interview with one of the informants: Oh, Yes. Having a witness expert present at a trial is very important, particularly expert witnesses who have experience accompanying women victims of domestic violence. --- Issue 7: Judges presiding over cases in court against women who commit homicide in response to domestic violence must adhere to the protection principle for domestic violence victims. Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. The following is the result of the author's interview with one of the research informants: Yes. It is important to consider the victim's perspective objectively. Therefore, the victim must receive protection, and the murder case should be intertwined with her status as a victim. This principle involves trusting the victim's statement as a primary factor, and subsequently considering other evidence, expert witness testimonies, and imposing punishment. --- Issue 8: The State shall provide more protection and attention to women victims of domestic violence. ` 2018 Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. One of the informants said: Yes, I think this issue has persisted for a considerable period of time. I understand that Indonesia has a Law on the Elimination of Domestic Violence in place. However, how does the implementation or practice of the law in reality? Although in each region, we have an Integrated Service Center for Women and Children Empowerment or Center for the Protection of Women and Children, it remains unclear whether the public facilities and services are sufficient. When implementing the protection program, we can all agree that victims will need paralegal assistance, a psychologist, maybe even a doctor if they experience physical abuse. Now, the question is, what is the perspective of law enforcement officers on the many domestic violence issues when the police and judges see women as victims. Although in this context, the state is obligated to make an effort to achieve these goals, it is clear that many things need improvement. --- Issue 9: The positive legal system must prioritize the protection of women victims of domestic violence Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. The following is the result of the author's interview with one of the research informants: Hm...Yes, it is still important because this positive law should have an orientation towards protection. In my opinion, the law is actually a reflection of social reality. Therefore, all aspects shall be considered, including the reason behinds the crime. But, unfortunately, I do not know how to change the way of this positive legal system. In the faculty, there is a sociolegal approach; maybe that is how the legal system includes it. Yes, it means that there must be such kind of training. --- Issue 10: There is a high tendency to believe that women who commit homicide in response to domestic violence shall be considered as victims and do not deserve punishment. Can this be accepted and realized in our criminal justice system? Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. One of the informants said: It would be difficult in our judicial system. It may prove challenging within our current legal framework to account for cases where murder is committed in response to domestic abuse. As the Criminal Code has yet to be amended, the ongoing discourse prioritizes the abolition of capital punishment. Even after this is achieved, addressing this issue remains a difficult task. I do not think they will not be completely free. --- Issue 11: The State should pay more attention and protection on women victims of domestic violence. Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. One of the informants said: Yes, but here is the relation with women who become victims, so what is clear is that first legal assistance is needed. Then, psychological assistance is also part of protection. Moreover, perhaps shelter should also be provided, for example, protection from the security side. It could also be from the husband's family, which is also from a security perspective. That is what the state must do, including protection from the media. ` 2019 Issue 12: Judges decide cases in court against women who commit homicide in response to domestic violence Of the respondents, 85.4% agreed, 9.8% disagreed, and the remaining were hesitant. In addition, many informants claimed that they had never seen an expert witness present at the trial. Issue 13: Should you agree that women who comit homicide in response to domestic violence must be given special legal treatment, what are the underlying reasons? --- 1. Some of the reasons for some of the informants are as follows: The position of women's status is indeed also subordinate in our society , so that women always experience discrimination and marginalization. --- 2. Sociological reasons are also important because understanding the context of why doing so needs to be considered. Premarital induction as prevention of domestic violence -Survey UPN Veteran, Faculty of Law Students and Unstructured interview Premarital induction as one of prevention measures to prevent domestic violence. Issue 14: The government should require all State Marriage Institutions to provide mandatory premarital counseling sessions to the prospective bride and groom. The sessions should explain the potential risks of domestic violence, its impact, and the procedures for reporting complaints. Of the respondents, 97.6%agreed that the KUA conducts premarital induction and explains the possibility of domestic violence. The following is the result of the author's interview with one of the research informants: Yes, I think this is important. Prospective marriage couples must have counselling session about domestic violance. In fact, according to the Law, domestic violence can be criminalized, right? If we think about the most appropriate service provider to place complaints, perhaps KUA would be best; hence this issue is important, mainly because the Law on eliminating domestic violence already exists. --- Punishment policies for women who commit homicide in response to domestic violence in various countries. In many countries, cases of homicide committed by women against their partners in response to domestic violence are mainly associated with Battered Women Syndrome , also known as the syndrome of women victims of violence or domestic violence syndrome. Initially, BWS was a psychological term to describe a subcategory of post-traumatic stress disorder where victims of domestic violence experienced significant trauma and believed themselves to be in danger even when they were safe. Over the years, the definition has evolved and entered the legal realm, where BWS is usually seen as an extension of self-defense for homicide, as the perpetrator's death usually occurs when there is no imminent danger, such as when the perpetrator is asleep. In this scenario, the victim chooses the only way to be sure of her safety from the tormentor, which is to murder him [16]. Research conducted by Penal Reform International in 2016 [17] showed a global awareness of the issue of BWS and a slow reaction to it. However, legislative and judicial attitudes towards women who retaliate or commit "criminal acts" against their abusers in response to violence, have varying degrees of leniency. In the countries that were the focus of PRI research, there was no specific legislative basis for a history of harassment/violence against ` 2020 women, which was considered a mitigating factor. Therefore, requests for lenient treatment were then brought into the existing criminal law framework. Usually, violators try to cover up their plea to get lenient treatment in the existing defense Attempts by victims of abuse to rely on self-defense, temporary insanity, and provocation have had different results in various countries. However, the PRI report also revealed that the existing defenses have proven to be unable to adapt to women with BWS conditions in many countries, and its reaction is prolonged. In a few countries studied, notably in several Australian states, the PRI found legislative amendments to the criminal code to facilitate more lenient treatment of female offenders due to the violence they experienced. These amendments take a variety of forms, from the introduction of new defenses specifically available to victims of domestic violence , to amendments to existing defenses so that they are better adapted to address victims of domestic violence . While some legal systems have been willing to adapt existing laws or even create new laws to deal with victims of domestic violence, other systems seem reluctant to expand beyond traditionally defined parameters. Their adapted legal system sympathizes with the view that the killing reaction may result from longstanding and accumulated domestic violence. The PRI report reveals that in certain countries, a record of domestic violence might diminish culpability or reduce penalties. Hong Kong, India, Japan, and Spain lack explicit sentencing rules or protocols, in contrast to Australia, Brazil, Mexico, Poland, and the United States . According to the PRI, these countries have significant latitude in imposing sentences, even if formal sentencing guidelines do not exist. --- • In Poland, courts have used their discretion to consider "general" factors in sentencing, such as the history of domestic violence. It led to the application of a reduced sentence or a reprieve. --- • In Mexico, courts may consider various factors when determining guilt , including the family relationship with the offense victim and other relevant circumstances. --- • In the United States, judges have broad discretion under the Federal Sentencing Guidelines and State Act, allowing various mitigating factors to be considered. In practice, it allows a history of past domestic violence to be used as a mitigating factor in some cases. --- • In Brazil, the rules on punishment can be applied broadly, allowing factors such as a history of domestic violence to be considered. --- • In Australia, sentencing guidelines and policies do not expressly allow the history of domestic violence to be considered. However, courts in all states usually rely on the court's power to consider all relevant factors in awarding sentences, taking into account the relevant case law. --- • In Spain, If there are mitigating circumstances, the court will impose a sentence that corresponds to the lower end of the penalty scale appropriate for the crime (unless one or two aggravating factors are also present --- • In Japan, a wide range of legal penalties are available for each crime, allowing judges and jurors considerable flexibility in determining the sentence to be imposed in each case. Therefore, this will allow a history of past domestic violence to be considered in sentencing. --- • In India, courts have recognized "continuous" provocation as a defense for murder. The acknowledgment has allowed a reduced sentence to be applied in the context of a history of domestic violence. In some cases, specific statutory reduced sentences may be applied if past history of domestic violence is considered a mitigating factor under one of the broader sentencing principles available under jurisdictional law. ` 2021 --- • In Brazil, if a crime is committed because of "social or moral values or excessive emotion," sentencing guidelines allow the sentence to be reduced by between one-sixth and one-third. As mentioned above, the history of domestic violence has been taken into account to determine that women's violations were indeed committed because of relevant "social or moral values." --- • In Australia, there are no laws or guidelines that expressly define the weight that should be given to a history of past abuse. However, particular examples have been identified to demonstrate the weight given to a history of domestic violence in sentencing. --- Current Policy in Indonesia Based on the research findings, women who commit homicide in response to domestic violence are sentenced using the Criminal Code or the Law on the Elimination of Domestic Violence No. 23 of 2004. Judges tend to use the KUHP, which is gender neutral. Moreover, victims undergoing police examinations sometimes experience multiple victimization in cases of domestic violence. This occurs due to inquiries being made about sensitive subjects and exerting pressure on the victim, who may be emotionally unstable. 1 --- SUGGESTED POLICY AND INTERVENTION PROGRAMS According to Babcock and Steiner, several studies have shown that detention alone will not help rehabilitate offenders from sacrificing their partners [18]. Meanwhile, Bennett et al. stated that programming in prisons that extends to the community after a period of detention is vital for the success of the rehabilitation of perpetrators [19]. Covington and Bloom [20] suggested that intervention programs for cases of domestic violence should be gender-specific since women who commit domestic violence and then commit homicide have different motivations than men. Identifying and addressing the particular risks and needs of both genders is crucial to effective intervention. They also suggested that intervention programs for domestic violence cases for women could include therapy related to why she committed the crime. Another program could also include an education program related to victim trauma because many women are victims of domestic violence. Many women who were previously victims of domestic violence also had experienced sexual harassment from their partners. When she is sexually abused by her partner, trauma informed therapy specific to sexual abuse is needed. Studies show that women are usually victims of domestic violence, but women can also be perpetrators of domestic violence. When providing a program for women who commit homicide in response to domestic violence, it is crucial to address the underlying issues related to why she made efforts to solve her problem of victimization by committing domestic homicide as an option. The Duluth model [21] suggests that programming for men should be different from programming for women because it must identify some gender-specific differences. For example, male domestic violence programs may include education about power and control, parenting classes, and dealing with substance or drug abuse. The role of men in domestic violence has been described as one rooted in the manipulation and power, and control of others. Therefore, programming for men should address their inherent desires for power and control by providing specialized work opportunities. Meanwhile, for women who murder to avoid victimization in the household, the next step is to focus more on problem-solving options such as filing a complaint with the police, seeking protection from neighbors and other parties. Several policies and procedural implications can help protect victims of domestic violence and provide the necessary programs for perpetrators in prison , among others, a consistent and cohesive program of legal guidance and awareness-raising interventions in prisons and the community. Alternatively, an intervention program for guidance and legal awareness to perpetrators in the community can help them mentally. Through these intervention programs, they can opt not to choose violence to avoid the possibility of further victimization within their family. ` 2022 Bennett, Stoops, Call, and Flett [19] have examined the effect of completing an intervention program of guidance and legal awareness on perpetrators of domestic violence and the rate of re-arrest. They determined that the average length of time between initial participation in the intervention program and recidivism was 2.4 years. Regarding re-arrests, 14.3 percent of program completers and 34.7 percent of non-completers were re-arrested for acts of violence to avoid victimization of domestic violence. Completing the intervention program reduces recidivism rates related to the selection of violence in order to avoid further domestic violence victimization. The findings from the research above show that behavior can be learned. For example, suppose violent behavior carried out by women to avoid domestic violence victimization has previously been believed to be a robust effort through previous social learning. In that case, it teaches them how to think about other options to prevent violence victimization through intervention programs and rehabilitation. Although this necessary procedure may encounter obstacles, it should be approached with optimism. Perpetrators need to be motivated to change, not forced to change. Motivation is the key for prisoners to complete the program successfully, and lack of motivation is often cited as a reason for failure to comply and re-offending. The suggestion for this is to allow inmates to decide if they want to get involved in the programs. Prisoners' motivation originates from the belief that they can effect positive change and that their future behavior hinges on their own actions and successful completion of the current program. [22]. When providing programs for men and women in prison settings, it is crucial to provide gender-specific programs to address their unique risks and needs. For instance, as mentioned earlier, a larger percentage of men have a record of domestic violence-related offenses compared to women. It is explained by the theoretical paradigm, which states that a man is dominant and aggressive in innate behavior. The Duluth Model, a popular domestic violence program, explicitly provides programming for men while purposely addressing the patterns men used to control and dominate their intimate partners. The Model also addresses the significant issue of societal acceptance of men using power and control to gain obedience from women. To address the specific needs of female offenders, in the context of avoiding victimization, Covington and Bloom [20] stated that the three most critical problems in women's lives are substance abuse, trauma, and mental health. Therefore, women's programming should include rehabilitation for substance abuse, opportunities to improve their current mental health, improve relationships with children and families, and cope with significant trauma. When female murderers in the context of family violence are released from prison , it is very important that there are parties in the community who help protect them from being perpetrators in responding to domestic violence if they are to rebuild their family life. Perpetrators must also be protected from feelings of revenge from the family of the husband or partner she murdered. Often, the family of the victim may seek revenge against the female perpetrator by attempting to contact them after their release, and possibly using violence. The Ohio Center for the Prevention of Family Violence 23 states that it is vital to have a security plan for female offenders whether or not they decide to remain in the relationship with the victim's family. In addition, the information and resources provided in the safety plan can be helpful in the future when the female offender finally decides to leave her victim's family. The Center also recommends three strategies when dealing with female offenders: providing protection if they stay, offering support if they choose to leave, or ensuring their safety if needed. First, protection strategies seek to prevent or respond to physical violence experienced by female offenders. These strategies may include leaving the victim's family, intervening in a third party, defending oneself, or temporarily leaving the relationship. Second, defensive strategies address the problems and risks that the female offender experiences as long as she remains in the relationship. The strategy includes: --- • Joining a support group for victims of domestic violence, ` 2023 • Maintaining a victim support system by staying close to family and friends, and; • Always trying to reconcile with the families of victims of domestic violence to avoid more severe treatment of violence. The final strategy is the leaving strategy, which addresses to the risks that female murderers may face when deciding to leave the relationship. Another way to help female perpetrators of murder is through thorough safety planning with the families of highrisk victims. There is a safety plan in place when they are released into the community. Before the female murderer is released from prison, the victim's community must develop a plan to implement specific safety measures to protect themselves. With a security plan in place, victims of domestic violence can have a plan for the worst-case scenario [23]. Resources are limited, particularly for women who commit murder. To ensure their success upon release, it is important to focus on planning for their safety and release. Therefore, prison personnel and members of the public who are willing to participate need to unite. It will be challenging when there is a lack of community or staff resources and advocacy programs for women who commit murder. In addition to providing resources for women who commit murder in the context of domestic violence, resources should be offered to prisoners upon their return to society. Some suggestions include community programs to continue to address core issues of domestic violence, therapy and programs to help teach better-coping skills, and resources to help inmates find housing, education, and jobs. Although these resources are necessary for successful reintegration into society, many prisoners are not linked with the right resources, or there are not enough resources available to prisoners on release. Training should be provided to professionals in healthcare and the criminal justice system to fully comprehend the extent of domestic violence. Services and safety planning can then be provided to the female murderer so that she has the resources she needs if she decides to leave her abusive partner's family. According to Kalra, Tanna, and Garcia-Moreno [24], healthcare providers are the first point of contact for many victims of intimate partner violence. Healthcare providers may be the only people the female murderers feel comfortable talking to about the abuse they experience at home. By identifying and providing care for women victims of domestic violence, healthcare providers can assist in reducing the violence experienced by them. Healthcare providers can also play an important role in gathering evidence for legal action against perpetrators. The human resource and its implication provide inputs for making better policies [25]. Implementation of performance accountability system for government institution provides inputs for making better regulation on implementation of performance accountability system for public officials and practitioners [26]. The Performance Accountability System For Government Agency provides inputs for making better regulation on performance accountability system for government institution policy [27]. Transformational Leadership, Technology Adoption, Public Service have a significant positive effect on Job Competency [28]. The leadership and service are useful for improving policy and practice and providing information to stakeholders [29]. The user satisfaction has a positive and significant effect on organizational performance [30]. Public Private Partnership policy and practice provide information to stakeholders related [31]. The role of Workplace Spirituality mediates effects of Information Technology on Innovative Work Behavior and mediates effects of Transformational Leadership on Innovative Work Behavior [32]. The collaboration must involve related institutions and should aim to improve regulations on disaster management [33]. Managing conflict strategy and its implementation are needed for providing information to stakeholders to provides inputs for making better regulation and policy [34]. Improving policy are needed for providing information and producing a clear input for public officials in making better regulations [35]. The community empowerment ` 2024 positively affects production capacity and social capital [36]. The Collaborative Governance in Digital Infrastructure Development and its implementation are needed for providing information to stakeholders [37]. It is necessary to formulate and implement a policy as inputs for making a better regulation in managing the implementation of the policy [38]. The Content of Policy and Context of Implementation are positively affected by the Program. The Collaborative Governance positively affects the Program [39]. The social leadership and social capital positively affect community empowerment and social capital fully mediates the relationship between social leadership on the community empowerment [40]. It is essential to implement community empowerment to provide stakeholders with information and inputs for making better regulations and policies [41]- [44]. --- CONCLUSION AND SUGGESTIONS FOR FUTURE RESEARCH In Indonesia, based on the research findings, two policies are used for sentencing, the penal code and the Law on the Elimination of Domestic Violence No. 23 of 2004. However, most judges tend to use the penal code, which is gender neutral. In addition, victims undergoing police examinations sometimes experience multiple victimization in cases of domestic violence. This occurs due to inquiries being made about sensitive subjects and exerting pressure on the victim, who may be emotionally unstable. Against the backdrop, the author concludes that a specific policy is needed in Indonesia to regulate alternative intervention programs for women who murder their partners in response to domestic violence. The intervention programs shall include social work penalty and rehabilitation measures as well as implementation procedures, success indicator measurement, and evaluation criteria to see the implications of the policy. On the rehabilitation measures, the policy can adopt the Duluth model, where women who commit homicide in response to domestic violence can receive treatment for substance abuse, trauma, and mental health which ultimately increase opportunities to improve their current mental health, improve relationships with children and families, and cope with significant trauma. Having a specific policy will help protect victims of domestic violence and provide the necessary programs for perpetrators in prison , among others, a consistent and cohesive program of legal guidance and awareness raising interventions in prisons and the community. Alternatively, an intervention program for guidance and legal awareness to perpetrators in the community can help them mentally. Through these intervention programs, they can opt not to choose violence to avoid the possibility of further victimization within their family. This research suggests that further research is needed to better understand the criminal history of domestic violence among prisoners, including women who commit homicide in response to domestic violence. In addition, since this research only analyzed a selected sample of individuals encountered in several correctional institutions, future research shall include a larger population sample. Future research should also include additional demographic and social factors to examine the prevalence of domestic violence by race, age, socioeconomic status, and sexual orientation. These demographics will make it possible to specifically focus on each population's specific risks and needs through specialized programming. In addition, future research should analyze the prevalence of domestic violence by sex to see if there are significant differences between men and women, but further research could complement a comprehensive examination of gender differences and look at the relationship between prisoners and their victims, which will provide insight into who is the victim of a crime related to domestic violence. These further assessments are critical because someone who abuses their intimate partner will require a different program from someone who abuses their young child. Also, future research can specifically look at the population of women who had domestic violence-related offenses and analyze the situation as to why these women were indicted and or punished for domestic violence. For Complaint analysis could offer insight into whether the

The findings from the dissertation showed that, in a normative manner, a wife who committed homicide against her husband was guilty and violated the criminal law regardless of the motive behinds it. Using feminist research and qualitative approach, the author argued that women (wives) who killed their partners in response to domestic violence were also victims. Therefore, imprisonment was not appropriate for them. This article discussed the public's perception on women committing homicide in response to domestic violence by exploring the reasons why women engaged in such acts, rehabilitation measures, lenient treatment, and protection programs for women who committed homicide in reaction to domestic violence. Based on these discussions, the author concluded that a specific policy was necessary in Indonesia to regulate alternative intervention programs for women (wives) who commit homicide to their partners in reaction to domestic violence. The intervention programs shall include social work penalty and rehabilitation measures as well as implementation procedures, success indicator measurement, and evaluation criteria to see the implications of the policy.

Introduction --- Background Eating disorders , such as anorexia and bulimia, are complex mental disorders defined by extreme obsessions with body weight or shape and unusual eating behaviors [1]. These diseases have the highest mortality rate of any mental illness [2], intractable comorbidities, [3] and worldwide prevalence [4], having become a major public health concern. Although a variety of treatment options have emerged over recent years [5], populations affected by EDs are often hard to reach through traditional health care services. This is mainly because of fear of stigma or a feeling of shame; many sufferers conceal their ED symptoms and never seek professional treatment or support [6,7]. To keep struggles with illnesses private, people often seek health-related information and support through online peer-to-peer communities, particularly via social media sites such as Twitter and Facebook. Participation in online communities is common in ED populations [8] and has been suggested as a screening factor for EDs [3]. This provides an opportunity for health care professionals to deliver health support to these hard-to-reach populations through online communities [9][10][11][12][13]. Moreover, as online communities present a relatively anonymous environment for individuals to naturally self-disclose and socialize [14], user-generated data online provide a large amount of records about individuals' concerns, thoughts, emotions, and social interactions [15][16][17], which can complement traditional data sources in understanding risk factors of EDs. Hence, growing research has focused on characterizing individuals' behavioral patterns in online communities [15][16][17][18][19] so as to better understand EDs and promote population-level well-being. One notable characteristic of online ED communities is their participants have widely different stances on EDs [8,20,21]. Some communities encourage members to discuss their struggles with EDs, share treatment options, and offer support toward recovery from EDs, so called prorecovery communities [20][21][22]. There are also many anti-recovery or pro-ED communities in which members often deny an ED to be a disorder and instead promote EDs as a healthy lifestyle choice [8,23]. These pro-ED communities can negatively affect health and quality of life among people with and without EDs, through reinforcing an individual's identity around EDs [24], promoting thin ideals [25], and disseminating harmful practices for weight loss [8]. Recent studies have shown that individuals' language use online strongly indicates their pro-ED or prorecovery stances [15,17,20], as well as emotions of depression, helplessness, and anxiety that reflect their mental disorders [16]. Other studies have also examined interactions between pro-ED and prorecovery communities on Flickr [21], anorexia-related misinformation [18], sentiments in comments on ED-related videos on YouTube [26], characteristics of removed pro-ED content, [27] and lexical variation of pro-ED tags on Instagram [19,28]. Yet, prior studies have largely focused on examining how people engage in and maintain an online ED community, whereas little is known about how people drop out of such a community. As a dynamic process, people who join and actively engage in a community at earlier stages can have less participation and leave the community at later stages. Understanding the attrition processes of online communities can enhance our knowledge of the dynamics in these communities. Studying the attrition process of an online community can also have practical implications for disease prevention and health interventions. Given the ease of accessibility of social media for many individuals , increasing attention has focused on using online communities to deliver health interventions [9][10][11][12][13]29,30]. One of the most popular approaches is to deliver health lessons and behavior-changing instructions via online communities [9][10][11][12][13]29]. Although pilot studies based on small samples have demonstrated the effectiveness of these approaches in reducing body dissatisfaction and disordered eating [12,13], evidence from interventions for a variety of health behaviors suggests that attrition is one of the most common challenges in online interventions [10,29]. This is known as the law of attrition of online interventions [31]. A recent study has shown a high attrition rate in an online intervention for EDs [32], though this intervention is delivered via a purposely designed website rather than a general social media site. Thus, an important goal in conducting successful interventions via online communities is to improve members' retention, as members who remain longer are more likely to receive these interventions and have more opportunities to promote a target behavior change. To achieve this goal, a critical first step is to understand what factors influence members' retention in an online community. Previous studies have shown that people's decisions of retention or dropout in online communities are associated with a variety of factors [33,34], including personality traits [35,36], interests [37], recognition in a community [38][39][40][41], and support from others [42,43]. However, such an association is not adequate to conclude the presence of a causal relationship [44,45] between an individual's attributes and her or his online participation. This is because an association can arise from non-causal relationships. For example, most previous studies focus on the use of self-reported surveys and rely on participants' reports of their own personality, concerns, and behaviors [35,39,46]. This can introduce considerable retrospective bias and measurement errors, leading to a coincidental association between 2 unrelated variables, particularly in small samples. Even if variables are measured rather than self-reported [43,47], participation in an online community is inherently self-selected and members can drop out for many different reasons . Thus, unobservable factors may affect both a main predictor and participation outcomes, causing a spurious association. Moreover, in some cases reverse causality can lead to an association. For example, previous studies suggest that feelings of social isolation are linked to frequent social media use [35,48] whereas recent studies indicate that social media use is linked to increased feelings of social isolation [49]. Technically speaking, the issues of measurement errors, confounding variables, and reverse causality can cause endogeneity, which refers to an explanatory variable of interest being correlated with the error term in a regression model [44]. In these cases, traditional methods such as ordinary least squares give biased and inconsistent estimates of the effect of interest. It is therefore not surprising that mixed results exist in previous studies. For example, a positive association between individuals' expertise and online participation was found in a study by Tausczik and Pennebaker [46] whereas a negative association was found in another study by Cook et al [38]. --- Objectives This study aimed to estimate determinants of dropout in an online ED community, while addressing the endogeneity issues by using an instrumental variable approach [44]. Specifically, we analyzed tweeting activities for a large set of individuals who self-identified with EDs on Twitter for over 1.5 years and identified the presence of dropout if a user ceased to post tweets in the observation period. We explored determinants of a user's dropout based on the incentive theory [34,50], which argues that people's engagement in an activity can be driven by intrinsic motivation which refers to doing something because it is interesting or enjoyable and extrinsic motivation which refers to doing something because it earns an external reward. We focused on intrinsic motivation captured by personal emotions and extrinsic motivation captured by sociometric status in an online peer-to-peer community. Rather than using self-reports [35,39,46], we measured users' emotions based on their emotional expressions in tweets using sentiment analysis techniques [51] and quantified users' sociometric statuses by network centrality [52] in the social network of an ED community on Twitter. On the basis of these measured variables, IV estimators both for the decision to drop out and for the time to drop out were implemented to achieve consistent estimates of the effects of personal emotions and network centrality on dropout in an online ED community. To better understand the estimation results, we further examined heterogeneity in tweeting interests of users with differing levels of characteristics and dropout outcomes. To our knowledge, this study is the first to systematically characterize the determinants of dropout behaviors in online ED communities. A total of 3 research questions were examined: what are the general characteristics of the attrition process in an online ED community? how do intrinsic and extrinsic factors affect the decision of an individual to drop out of the community? and how do these factors affect the duration of time until the occurrence of dropout? --- Methods --- Data Collection Our data are collected from Twitter, a microblogging platform that allows millions of users to self-disclose and socialize. As many social media platforms such as Facebook and Instagram have taken moderation actions to counteract pro-ED content and user accounts [28], Twitter has not yet enforced actions to limit such content [53]. This makes Twitter a unique platform to study the attrition process naturally happening in an online ED community and allows us to examine individuals' behaviors in a nonreactive way. Our study protocol was approved by the Ethics Committee at the University of Southampton. All data used in our study are public information on Twitter and available through the Twitter application programming interfaces . No personally identifiable information was used in this study. Our data collection process included 3 phases: 1. First, we collected a set of individuals who self-identified with EDs on Twitter using a snowball sampling approach. Specifically, we tracked the public tweet stream using "eating disorder," "anorexia," "bulimia," and "EDNOS" from January 8 to 15, 2016. This resulted in 1169 tweets that mentioned EDs. From the authors of these tweets, we identified 33 users who self-reported both ED-related keywords and personal bio-information in their profile descriptions . Starting from these seed users, we expanded the user set using snowball sampling through their social networks of followees and followers. At each sampling stage, we filtered out non-English speaking accounts and finally obtained 3380 unique ED users who self-reported ED-related keywords and bio-information in their profile descriptions. Note that our focus in this work is studying individuals who are affected by EDs rather than those who are related to EDs. The inclusion of bio-information in user sampling allowed us to filter out ED-related therapists, institutes, or organizations, as these users often displayed ED-related keywords but did not show bio-information in their Twitter profile descriptions. Details about the data collection of ED users can be found in our previous study [54]. 2. Then, we collected all friends of each ED user, leading to a large social network consisting of 208,063 users. For each user, we retrieved up to 3200 of their most recent tweets and obtained 241,243,043 tweets in total. The data collection process finished on February 11, 2016. 3. Finally, we opened a follow-up observation period for all users on August 17, 2017, to obtain measurements on users' activities online. In the second observation, we only collected users' profile information which includes users' last posted statuses. To verify the quality of our collected sample, 2 members of the research team classified a random sample of 1000 users on whether they were likely to be a true ED user based on their posted tweets, images, and friends' profiles. Users were classified as "disordered" if they frequently and intensively posted their body weights, details of their dietary regimen , struggles with eating , pictures of themselves, and self-reports of being disordered or in recovery in tweets and followed ED-related friends . The process revealed a 95.2% match between the identified ED individuals in the data collection stage and those classified as ED during inspection. Although it is impossible to diagnose individuals' disorders based on their online behaviors, this inspection provides a strong indication that the collected users are likely to be affected by EDs rather than those who merely talk about EDs online. See [54] for details of data validation. --- Estimation Framework There are 2 different models specified to estimate the effects of emotions and network centrality on dropout. First, we specified a linear probability model on the whole sample to estimate the effects of individuals' characteristics observed in the first observation period on the probability of dropping out in the second observation period. Finally, we estimated survival models to explore the effects of individuals' characteristics observed in the first observation on the time to dropout in the second observation . However, similar to all social media studies, only a limited number of individuals' characteristics are available for our estimations and these are mostly observed through user-generated data online. This leads to confounding variable bias, as unobservable factors can be correlated with both the main explanatory variables and dropout outcomes. For example, undergoing hospital treatment can simultaneously affect a person's emotional state and the use of social media. Furthermore, previous studies have shown that social media use is associated with increased depression [55], social anxiety, [49] and body dissatisfaction [56,57], implying an effect of online participation on individuals' emotions . Both confounding variables and reverse causality result in biased estimates of the effects of emotions and network centrality on dropout. This problem can be addressed by using a randomized controlled trial in which emotions or network centralities are randomly assigned to users by researchers [58]. Such a trial, however, is not feasible, because of ethical and practical limitations [59]. Here, we utilized an alternative approach for estimating the effects of interest that is based on IV regression, an econometric technique to infer causal relations from observational data [44]. This technique has been applied to a variety of contexts, from identifying the causal effect of education on earning [60], the effect of a health treatment [61], to estimating social contagion effects on both online [59] and offline behaviors [62]. Formally, consider a model Y=Beta1*X1 + Beta2*X2 + u, where X1 is endogenous, X2 is exogenous, u is a random error term, and Betas are effects to be estimated. IV methodology uses an instrument Z is not equal to 0, and [iii] uncorrelated with u, ie, cov=0, conditional on the other covariates such as X2) and runs a first stage reduced-form regression X1 = B1*Z + B2*X2 + v, where v is a random error and B s are coefficients. The causal effect of X1 on Y is then given in a second stage regression Y=Beta1*X3 + Beta2*X2 + u, where X3 is the predicted value of X1 from the first stage. See [44] for more details. --- Measures A number of variables are needed for estimations. All independent variables and IVs are measured in the first observation period , whereas dependent variables are measured in the second observation period. --- Dropout Outcomes as Dependent Variables Following previous studies [42,43], we identified the presence of dropout if a user ceased to post tweets. Specifically, in the linear probability models, we encoded the dropout status of a user as 0 if the user had updated posts in our second observation and 1 otherwise. In the survival models, each user has a 2-variable outcome: a censoring variable denoting whether the event of dropout occurs and a variable of survival time denoting the duration of time until the occurrence of dropout. We censored the occurrence of a "dropout event" in 2 ways. First, users are said to drop out if they have not posted tweets for more than a fixed threshold interval π before our second observation . As people use social media platforms with different activity levels , our second censoring method further accounts for personalized posting activities of individuals . In this method, users are said to drop out if they have not posted tweets for more than a variate threshold interval λπ + Ii before our second observation, where π is a fixed threshold, Ii is the average posting interval of individual i in our first observation period, and λ is a tunable parameter to control the effects of individual activities. We tuned the parameters by maximizing the agreement between the estimated dropout states based on users' activities in our first observation and the observed states in our second observation. See Multimedia Appendix 1 for details. For users who were censored as dropped-out, we set their survival times as the durations from our first observation to their last postings in our second observation. For those who were censored as non-dropped-out, we set their survival times as the whole time period between our 2 observations. --- Emotions and Network Centrality as Main Explanatory Variables Individuals' emotions were measured through their language used in tweets. There is a variety of sentiment analysis algorithms to measure emotional expressions in texts [51,63]. In this study, we used SentiStrength [51] as it has been used to measure the emotional content in online ED communities and has shown good interrater reliability [26] and it is designed for short informal texts with abbreviations and slang and is thus suitable to process tweets [51]. After removing mention marks, hashtags, and URLs, each tweet was assigned a scaled value ranging from -4 to 4 by SentiStrength, where negative and positive scores indicate the strength of negative and positive emotions, respectively, and 0 denotes neutral emotions. We quantified a user's emotional state by the average score of all tweets posted by the user. All retweets were excluded, as retweets reflect the emotions of their original authors more than those of their retweeters. To obtain robust results from the language processing algorithms, we only considered users who had more than 10 tweets and posted more than 50 words. Network centrality measures the importance of a person in a social network; people well-recognized by their peers often have high centralities in a group [52]. To measure a user's centrality in the ED community, we built a who-follows-whom network among ED users and their friends, where a directed edge runs from node A representing user A to node B representing user B if A follows B on Twitter. Although there are various measures of network centrality, we focused on coreness centrality [64] as it has been shown to outperform other measures such as degree and betweenness centrality [52] in detecting influential nodes in complex networks [65] and cascades of users leaving an online community [66,67]. We measured the sociometric status of a user in the ED community by the in-coreness centrality [68] of a node in the generated network using the package igraph version 0.7.0 [69]. --- Aggregated Emotions and Network Centrality of Friends as Instrumental Variables As IVs for a user's attributes, we used average emotions and network centrality over all followees of the user, namely people who are followed by the user. The choice of these IVs is based on the following considerations. First, we considered the relevance assumption of our instruments, requiring that the characteristics of followees be correlated to the user's characteristics, namely cov is not equal to 0. We expected the followees' updates to act as information sources for a user and followees' behaviors as well as emotions manifested in their tweets can influence the user. Previous work [54] has shown the presence of homophily among ED users on Twitter suggesting that users who share similar emotional and network attributes tend to follow one another. Furthermore, the empirical existence and strength of the relevance property are tested in a first-stage regression and presented along with the structural estimates of the models. Finally, we examined the exogeneity requirement =0), where followees' emotions and centrality must not have a direct effect on the dropout decision of the user other than through their effect on the user's emotions. Although we have taken such assumptions to be reasonable, we identified a pathway through which direct links could arise. Followees' attributes could affect a user's dropout through their effects on followees' own dropouts. For example, followees' emotional states may affect their own dropouts, and a feeling of loneliness because of friends' leaving may then drive the target user to drop out. To control for this channel, we measured the proportion and durations of followees that remained active in our second observation . Furthermore, we changed the definition of followees to those who are followed by a user but do not follow the user back . In this setting, the reverse causality of a user's dropout on followees' attributes was nullified, which strengthened the exogeneity assumption on IVs and controls. --- Estimation Covariates Our estimates control for several covariates that may affect users' tweeting activities, as listed in Table 1. First, we measured users' social capital on Twitter to capture the fact that people with different levels of popularity may have different tendencies to share content online [70]. Note that, although the numbers of followees and followers can be regarded as the in-and out-degree centralities of a user in the whole social network on Twitter , we were interested in the "local" network centrality in the ED-specific communities. Second, as previous studies show an association between social media use and depression [55], we measured historical activity levels of users to capture effects that previous engagement may relate to both users' emotions and their future engagement. We also measured users' activity frequencies to capture their patterns of Twitter usage. Third, the covariates on observational bias were used to control for effects caused by incomplete observations. For example, a limited number of tweets were retrieved and used to measure emotions for a user. All variables on social capital, activity level, and observational bias were measured from users' profile information and tweets collected in our first observation. Finally, as discussed above, we included the proportion and average durations of followees that were active in our second observation to capture the channel that followees' emotions affect a user's dropout through their effects on followees' own dropouts. --- Model Estimations --- Instrumental Variables Estimation in Linear Regression Model We use standard 2-stage least squares estimators for linear probability models. In the first stage, we ran an auxiliary regression and predicted the endogenous variables based on IVs and exogenous covariates. In the second stage regression, we substituted the endogenous variables of interest with their predicted values from the first stage. Estimation was conducted through the package of applied econometrics with R [71], and robust standard errors were computed. --- Instrumental Variables Estimation in Survival Model We used a Kaplan-Meier estimator [72] to estimate the survival function from data. Aalen's additive hazards model [73] was used to estimate the effects of users' attributes on the time to dropout. Compared with the proportional hazards models in which the ratios of hazard functions for different strata were assumed to be constant over time [74], the additive model was more flexible and applied under less restrictive assumptions. To compute an IV estimator in an additive hazards model, we used a control function based approach which was proposed by Tchetgen et al [61]. The timereg package [75] was used for the implementation of the estimation algorithm. Standard errors were obtained through nonparametric bootstrap. --- Results --- Descriptive Statistics We obtained 2906 users who posted more than 10 tweets and 50 words in our data, where 84.61% of users had no posting activities during our 2 observation periods. Among the 357 users who self-reported gender information in their Twitter profile descriptions, 84.0% of them were female. The mean age was 17.3 years among ED users who self-reported age . On the basis of the timestamps of account creation and last posting, we used the Kaplan-Meier estimator to estimate the "lifetime" of a user on Twitter, namely the duration from account creation to the last posting. The estimated median lifetime of these users on Twitter was 6 months. That is, one half of the entire cohort drops out at 6 months after creating an account. Figure 1 visualizes the social network between dropouts and nondropouts among ED users, laid out by the Fruchterman-Reingold algorithm [76]. We noted that users with the same dropout states tended to cluster together. Computing Newman's homophily coefficient r [77] of this network by users' dropout states, we found r=0.09 , suggesting that users with the same dropout states tended to befriend one another. See Multimedia Appendix 1 for details of data statistics. --- Estimation Results of Linear Probability Models Table 2 shows estimated results in the linear models with 2 different IV specifications. In the first specification, we used all followees of a user to create IVs for the user's attributes. The results are given in columns 2 to 3, in which both OLS and IV estimators show that positive emotions are associated with a higher probability of dropout , with largely comparable coefficients for covariates. Compared with the OLS estimator, the IV estimator of the effect of emotions on dropout was remarkably stronger. The Wu-Hausman test further showed a significant difference between the OLS and IV estimators , suggesting the presence of endogeneity. These results indicated that ignoring endogeneity in the OLS estimation leads to an underestimation of the effect of interest. Moreover, the F statistics in the first stage regressions showed that the relevance of IVs exceeds the conventional standard of F=10 [78], indicating the validity of our IVs. Columns 4 to 5 show results of the second IV specification in which only single way followees are used to create IVs. Users who have no single way followees are excluded as instruments for these users' attributes were not available. Thus, the number of observations decreased as compared with that in the first IV specification. Moreover, as data on a smaller number of friends were used in the second IV specification, the relevance of IV became weaker but still passed the conventional test in the first stage regression. Despite such changes, the 2 specifications produced largely similar results. Computing Wald tests of equality of coefficients between the 2 IV models, we found that the estimated effects of emotions on dropout were statistically the same across different IV specifications , potentially suggesting robustness of the results. Note that network centrality was excluded from the linear models. This is because many users had dropped out long before our first observation , and the social networks of such users might largely change from the dates of their dropouts to our first observation. For example, a user might be followed by new followers when these followers were unaware of the dropout of this user. In these cases, network centralities in the future are used to explain dropouts in the past, which can produce misleading results in the linear models. In fact, including network centrality in the above linear models produces statistically insignificant effect of centrality on the dropout decision, confirming our argument that network centrality is irrelevant to the binary decision to drop out or not. --- Estimation Results of Survival Models In the survival models, we only considered users who were active past our first observation period, so as to examine the effect of network centralities in our first observation on users' activities in the second observation period. Table 3 shows mean coefficients of emotions and network centrality in the survival models. All models are estimated controlling for the full list of covariates but are omitted from the tables because of space concerns. The complete results are available from the authors. Following [61], the effects of all covariates are assumed to be time dependent in estimations. Both the standard and IV models on the identical interval censored data show that positive emotions lead to a shorter survival time and a core position in social networks is associated with a longer survival time . Estimations on the personalized interval censored data and using different IV specifications gave similar results. The strong relevance of IVs in the first stage regressions confirms the validity of IVs across different models. A comparison of results between the linear and survival models further shows that these models have consistent estimators for the effect of emotions on dropout, namely positive emotions increase the likelihood to drop out. --- Identical interval censoring -0.061 -0.018 -0.043 -0.018 Emotions 0.001 0.001 0.001 0.001 Centrality Personalized interval censoring -0.056 -0.015 -0.038 -0.016 Emotions 0.001 0.001 0.001 0.001 Centrality a F-statistic tests the joint significance of the 2 instruments from a first-stage regression of a user's emotions on followees' emotions and followees' centralities plus the rest of the covariates, where F=66.11 when all followees are used to build instruments and F=34.99 when only single way followees are used. b F-statistic tests the joint significance of the 2 excluded instruments from a first-stage regression of a user's centrality on followees' emotions and followees'centralities plus the rest of the covariates, where F=27.85 when all followees are used to build instruments and F=12.62 when only single way followees are used. c The numbers of samples that are used in estimations. d CIs for coefficients are obtained from 1000 bootstrap replicates. A coefficient is significant at P<.05 if 0 is not in 95% CIs. e IVs stands for instrumental variables. --- Underlying Connection Between Emotions and Dropout To better understand the relationships between emotions and dropout, we examined posting interests among users with different dropout statuses and emotional states based on hashtags used in users' tweets . We found nondropouts were interested in advocating a thin ideal and promoting a pro-ED identity . In contrast, dropouts engage in discussing their health problems and offering emotional support for others , which implies a tendency of these users to recover from disorders [20][21][22]. Similarly, we split all ED users into 3 equal-size sets based on their emotional scores and examined hashtags used by each set of users. We found that users with negative emotions often engage in promoting thin ideals , showing largely overlapping interests with the nondropouts. In contrast, users with neutral and positive emotions were more interested in discussing their health problems , opposing pro-ED promotions , and encouraging healthier body image and behaviors , showing similar interests with the dropouts. See Multimedia Appendix 1 for more detailed lists of hashtags. Measuring the Spearman rank correlation ρ between pairwise lists of hashtags posted by users with a given state , we found a positive correlation between negative users and nondropouts in hashtag usage , indicating similar posting interests among these users. A similar pattern occurred between positive users and dropouts. In contrast, users with other pairs of states showed a negative correlation or noncorrelation in hashtag usage, indicating their discrepancies in posting interests. Note that all tags in 2 lists Di and Dj are considered in computing the correlation ρ ; tags in each list are ranked by TF-IDF scores [79] and the TF-IDF score of tag t in list Di is 0 if Di does not contain t. These results revealed a possible underlying connection between positive emotions and dropout. Compared with users with positive emotions, those with negative emotions had more similar interests to active members in the ED community. Finding similarities with other members in a community can enhance a sense of belonging to the community and positively increase intention to engage in community activities [33,37]. Therefore, it is not surprising that negative users are less likely to drop out than positive users in our estimations. --- Discussion --- Principal Findings This study provided the first estimates of the effects of personal emotions and interpersonal social networks on dropout in online ED communities. This study has several strengths. First, we based our analysis on the incentive theory to explore determinants of users' online behaviors , allowing us to study users' behaviors in a more systematic way than most previous studies that often focus on a single type of determinant . Second, we used automated sentiment analysis techniques to measure users' emotions and network analysis methods to quantify users' sociometric statuses in an online community, leading to higher efficiency than traditional research methods such as surveys [35,37,39,41,48]. Finally, we applied an IV approach to both linear probability and survival models, which enabled us to achieve a more consistent estimate of human behavior in online settings than traditional methods used in previous studies [39,41,47]. Overall, we found that positive emotions increased the likelihood of dropout in ED individuals and accelerated the dropout process on Twitter. In contrast, a central position in the social network of ED individuals at an earlier stage was associated with prolonged participation of an individual at a later stage. These findings were verified across a variety of robustness checks. Despite differences in methodology, our findings aligned with previous studies in psychological and social media research [5,33,35]. Our results suggested that ED users with negative emotions had high levels of participation on Twitter. This aligned with previous survey studies on social media use , where people with social anxiety and shyness were found to spend more time online [35,48,81]. An explanation for this is the online disinhibition effect [82]. Specifically, because of anonymity in online interactions, people with social inhibitions might be more willing to share personal feelings and reveal themselves in online interactions than offline interactions to meet their social and intimacy needs [48]. Additional analyses on users' posting interests revealed that users with negative emotions share similar interests with active users. This allowed us to confirm the validity of our results via the social capital theory [39,40], namely sharing common attributes with other members can enhance a sense of belonging and positive feeling toward a community, which drives people to actively engage in the community. Consistent with positive associations between network centrality and active participation in other online communities [67,70], we find that central users in the social network of an ED community tend to have a longer-lasting participation in the community. This result is expected for several reasons. First, users who are centrally embedded in a group have a relatively high number of social ties with other members, which can lead these users to feel being socially accepted and approved, as well as a strong sense of belonging to the group. Previous studies have consistently shown that recognition from other members and identification within an online community increase an individual's commitment to the community [34,[39][40][41]. Finally, information shared by central users is likely to spread to the majority of a community through social ties, and their central positions in the community may promote other members to trust such information [70]. This implies that central users have a greater potential than peripheral users in influencing members' opinions, emotions, and behaviors in online communities [84]. Thus, compared with peripheral users, feeling influential may provide an additional incentive for central users to continue participating. In line with previous studies on online ED communities [15,17,21], we found that ED users on Twitter have different stances on EDs, where users with negative emotions often share pro-ED content and those with positive emotions often share prorecovery content. As pro-ED content often contains thin-ideal images and harmful tips for weight loss and control [8,24,25], this result aligned with clinical evidence on ED treatment showing that more emotional distress is associated with a higher risk to learn and develop dysfunctional coping behaviors among ED sufferers [5]. Thus, as suggested by previous studies [85], engaging in pro-ED content may serve as a coping mechanism to deal with emotional pressures and stress of EDs. A possible explanation for the association between engaging in harmful online content and coping with stress is sensation seeking [86], a basic personality trait defined as the seeking of varied, novel, complex and intense sensations and experiences, and the willingness to take risks. Several studies have shown that sensation seeking is prominent in adolescence and closely related to pathological internet use, such as use of violent sites [87] and internet dependence [88]. Our study also offered new insights into online ED communities. First, ED users have a high dropout rate and a short lifespan between an account creation to lost posting on Twitter . This aligns with views of online ED communities as hidden, secretive groups [30], but also indicates the dynamic characteristics of these communities. Second, users who discuss their health problems and share prorecovery content have lower levels of posting activities than those who share pro-ED content on Twitter. This can be explained as follows. Owing to common interests in EDs, prorecovery and pro-ED groups are likely to be connected in the same social networks, and content shared within a group is hence likely to be visible to the other group. However, exposure to content from the antagonist group can have distinct effects in pro-ED and prorecovery groups. Exposure to pro-ED content is harmful for prorecovery users and can impede their recovery process [3,24], whereas exposure to prorecovery content can instead stimulate harmful behaviors in pro-ED users [21]. Thus, prorecovery users might tend to leave such an online community to avoid a risk of further deterioration or relapse. Our finding may also explain why pro-ED content is found being more pervasive than prorecovery content across social media sites [15,17,21], for example, almost 5 times in terms of unique publishers on Tumblr [15]. Finally, ED users tend to connect with others with the same dropout states on Twitter. This implies that whether an individual drops out from online communities depends on whether others in the individual's social networks drop out. In other words, dropout in online ED communities is not only a function of individual experience or individual choice but also a property of group interactions, such as homophily [89] and social contagion effects [59]. --- Implications Our findings are of practical relevance to the promotion of public health over social media. First, the decision to maintain active participation in an online community can be caused by intrinsic and extrinsic characteristics or traits of the participants, such as personal emotions, interests, and social networks. Such self-selection bias can lead to the sample not being representative of the whole population, and hence researchers need to consider both active and dropped-out users for a well-rounded picture of online health communities. This is particularly important for public health officials to make special efforts to reach these dropouts and offer more intensive support when they are trying to recover. Second, high attrition rates are often regarded as negative outcomes in online interventions, particularly in those delivered over a purposely designed website [11,31,32]. However, this may or may not be the case in interventions over general social media sites depending on how targeted populations use these sites. For example, when an intervention is delivered in an online community in which members often shared harmful content, a high attrition rate may not be a negative outcome. Using automated data-mining techniques to track users' behaviors , as used in this study, can provide more detailed information about people's use of online health communities and improve our understanding of attrition in online interventions. Third, interventions that recommend content containing positive emotions to ED users may reduce their engagement in a harmful online community. This aligns with Fredrickson's broaden-and-build model which argues that cultivating positive emotions is useful to prevent and treat mental health problems [90]. Finally, intervention strategies could be tailored for different individuals depending on their positions in the social network of an online community. For example, identifying central individuals as change agents might enhance the efficacy and cost-effectiveness of an intervention because of their greater influence potential through larger numbers of social ties [91] and also their longer-lasting effects through longer-term participation in the community. --- Limitations First, we recognized that self-diagnosis information on Twitter might be itself self-censored by users to align with their personality traits and perceptions of their audience on the platform. People may not use tags such as "eating disorder" to self-report their experience of illness and would be excluded by our collection methods. Also, although over 208,000 users and over 241,000,000 tweets are studied in this study, a small sample of rich social media data is used to explore the attrition of ED communities on Twitter. Thus, generalization of our results to all ED-related online communities should be cautious. Second, our measures of dropout are based on posting activity, whereas some people primarily use Twitter to receive outside information but rarely post their own information. We have little activity data on these users and hence less understanding of the characteristics of their dropout. This thus raises important issues that need further research to enhance our understanding of attrition in online health communities, such as consensus and clarity about the definition of dropout. Third, our study focused on the Twitter platform, without validation on other platforms. However, stopping using a platform can be related to the attractiveness of the platform. Hence, future research is also in need to examine many other factors that we did not explore but can affect dropout on social media, such as individual personality, physical health states, perceptions, and purposes of using a particular social media platform. Fourth, user accounts on a social media site are often not unique-an individual may have multiple accounts on the same site. Thus, we cannot be certain whether individuals who stopped using an account will engage in the same or similar online communities through other accounts. In other words, the dropout of a user account may not necessarily imply that an individual abandons a specific identity shared within a community. Finally, although our methodology allows us to establish a causal link from emotions to dropout behavior, it offers limited insights into the pathways through which this link exactly operates, and future work is needed to explore such issues in detail. --- Conclusions This study presented a systematic characterization of attrition in an ED community on Twitter. Our analysis offered the first attempt toward the estimation of the effects of personal emotions and network centrality on dropout behaviors in individuals affected by ED on Twitter. Our results provided new insights into the trajectories that ED communities develop online which can help public health officials to better understand individual needs in using online ED communities and provided tailored support for individuals with different needs. --- Conflicts of Interest None declared. --- Multimedia Appendix 1 Supplementary information for estimating determinants of attrition in eating disorder communities on Twitter: an instrumental variables approach. [PDF File , 700KB

The use of social media as a key health information source has increased steadily among people affected by eating disorders (EDs). Research has examined characteristics of individuals engaging in online communities, whereas little is known about discontinuation of engagement and the phenomenon of participants dropping out of these communities. Objective: This study aimed to investigate the characteristics of dropout behaviors among eating disordered individuals on Twitter and to estimate the causal effects of personal emotions and social networks on dropout behaviors. Methods: Using a snowball sampling method, we collected a set of individuals who self-identified with EDs in their Twitter profile descriptions, as well as their tweets and social networks, leading to 241,243,043 tweets from 208,063 users. Individuals' emotions are measured from their language use in tweets using an automatic sentiment analysis tool, and network centralities are measured from users' following networks. Dropout statuses of users are observed in a follow-up period 1.5 years later (from February 11, 2016 to August 17, 2017). Linear and survival regression instrumental variables models are used to estimate the effects of emotions and network centrality on dropout behaviors. The average levels of attributes among an individual's followees (ie, people who are followed by the individual) are used as instruments for the individual's attributes. Results: Eating disordered users have relatively short periods of activity on Twitter with one half of our sample dropping out at 6 months after account creation. Active users show more negative emotions and higher network centralities than dropped-out users. Active users tend to connect to other active users, whereas dropped-out users tend to cluster together. Estimation results suggest that users' emotions and network centralities have causal effects on their dropout behaviors on Twitter. More specifically, users with positive emotions are more likely to drop out and have shorter lasting periods of activity online than users with negative emotions, whereas central users in a social network have longer lasting participation than peripheral users. Findings on users' tweeting interests further show that users who attempt to recover from EDs are more likely to drop out than those who promote EDs as a lifestyle choice.Presence in online communities is strongly determined by the individual's emotions and social networks, suggesting that studies analyzing and trying to draw condition and population characteristics through online health communities are likely to be biased. Future research needs to examine in more detail the links between individual characteristics and participation patterns if better understanding of the entire population is to be achieved. At the same time, such attrition dynamics need to be acknowledged and controlled when designing online interventions so as to accurately capture their intended populations.

Introduction In the 21st century, we are facing large scale collective action problems that may have catastrophic consequences -from vaccine hoarding in a global pandemic to the ongoing refugee crises around the globe . It is clear that, in order to solve these crises, we need compassionate and coordinated action from the global community -that is, we need to act as one. However, reality tells us that we do not always adopt superordinate goals such as these, particularly when our interests conflict with those who are different from us. Indeed, as classic social psychological research has shown, strong more resources than another . In the current work, we study two unequally resourced groups that have the option to work together to achieve a superordinate goal. Across three studies, we explore whether diverse interventions can overcome barriers to compassionate and cooperative action across group boundaries. --- Superordinate goals help to overcome the intergroup divide Muzafer Sherif was one of the first to discuss the importance of superordinate goals -goals where two or more groups need to cooperate to achieve a particular outcome . In the Robbers Cave experiment, Sherif coordinated a summer camp with several young boys where they were separated into two groups. Unbeknownst to the boys, the summer camp was a disguise for a larger goal -to explore how individuals compete and cooperate when they are members of different groups. The two groups quickly fell into conflict, and cooperation could only be achieved when the experimenters introduced a superordinate goal. The findings show that when people are divided into groups, intergroup conflict can arise, and individuals often pursue the goal of their ingroup at the expense of their outgroup counterparts . Since these original studies, research has reinforced that an intergroup divide can be bridged when groups adopt superordinate goals . However, there are many socio-structural factors that prevent group members from embracing such superordinate goals. One obstacle might be the magnitude of the resource gap between the groups. If this is perceived to be too large, group members at opposite ends of the wealth spectrum may see the groups as too different and this might affect coordinated action . Group resource inequality not only enhances the perceived difference between groups . Indeed, resource inequality enhances 'us' versus 'them' divisions , and there is evidence that inequality, therefore, leads to less compassionate behavior toward others . Failing to overcome the ingroup-outgroup divide can have negative consequences when groups need to cooperate to achieve superordinate goals. We are interested in investigating potential interventions to reduce intergroup competition in this contextwith the ultimate aim of increasing compassionate action in line with a superordinate goal. We focus on three possible interventions. The first targets individualistic solutions and we test the power of compassion training aimed at making individual group members feel and act more compassionately. Our reasoning builds on past research that has shown that compassion training , can foster compassionate action . Second, interventions that change the structural context may offer an effective method to elicit compassion by altering or shifting group dynamics. We reason that if, relative to low economic inequality, high economic inequality exacerbates the ingroup-outgroup divide and erodes prosocial behavior , reducing the degree of inequality may foster greater adoption of superordinate goals. Third, another potentially effective intervention focuses on informal leaders altering group norms that promote acting for the common good and breaking down 'us' versus 'them' barriers. Leaders that focus on achieving desired superordinate outcomes might be particularly influential when there are no established group norms and when people therefore look to others for appropriate actions on how to behave . Past research has demonstrated that group norms can have a significant impact on the behavior of other group members , and may influence the adoption of superordinate goals. --- The present research If we are to understand how to foster compassionate action in the face of large-scale problems, we need to understand what individual, structural and normative factors may shift group behavior. The current study aimed to gain a better understanding of effective ways of orienting individuals toward superordinate, compassionate behavior when they are embedded in an unequal intergroup context. In particular, we assessed the efficacy of compassion meditation training , lower levels of structural inequality , and the introduction of a pro-sharing group norm by an informal leader . If we are to tackle some of the pressing issues of the 21 st century such as climate change and future pandemics, we need to establish which interventions are most likely to lead to collaborative action that gives priority to superordinate goals. --- Study 1 Several scientific studies have demonstrated that compassion meditation practices can promote prosocial behavior and compassionate responding . To date, most of these have demonstrated an effect of long-term compassion meditation practice , but there is less research about the effect of short-term compassion meditation on behavior. Brief interventions, such as 10-min meditations , as well as self-compassion re-framing have led to changes in self-report levels of motivation and feelings , but the researchers did not assess how behavior might be impacted. Past work has also shown that cueing individuals to the needs of others results in more prosocial behavior in dictator games . In each of these studies however, an individual is typically asked to make a compassionate decision individually, yet little is known about whether compassion meditation can affect decisions made by groups, and specifically, group contexts where there is an unequal distribution of resources. This is important to understand as many compassionate acts in the real world need to be made by groups rather than individuals, and also occur in a broader ecological context. Here we aimed to assess the effect of a brief compassion meditation on the adoption of a superordinate goal when groups are unequal. To achieve this, we exposed participants to one of two 10-min meditation exercises: compassion meditation or a focussed imagery meditation. Participants were then divided into groups and asked to complete a food assembly task to make food for starving people. As exposure to compassion meditations appear to enhance compassion toward others , we hypothesised that all participants would behave more compassionately by working collaboratively after being exposed to a compassion meditation compared to a focussed imagery meditation. --- Method Ethical clearance of the study was obtained in line with the ethical review processes of the Human Research Ethics Committee . --- Transparency and openness This study, including the hypotheses and analytical approach, were preregistered on the Open Science Framework . We note in the method and results sections below which measures were confirmatory and exploratory. All data, materials and R script has also been made available on OSF. See the following link for these resources: https://osf.io/fjp6b/?view_only =b51cca6755654a76b8d1c8d77a3cfa53. 1 For each study, we report all data exclusions , all manipulations and how the sample size was determined. 1 We pre-registered additional hypotheses regarding a dependent variable: entitled behavior as measured by more taking of sweets . However, this variable falls outside our primary research concerns and we have elected not to include any information about it here. However, descriptions of this variable and findings have been included in Supplemental materials 1. --- Design summary Our study design was inspired by an activity used with high school students in a non-scientific setting . Compassion It-an organization that aims to invoke more compassionate action in the world-ran compassion workshops with students in grade 10 in San Diego, USA. At the conclusion of these workshops, the students were assigned to groups that represented high and low income countries. All countries were tasked with the same goal, to produce as much 'food as they could' using LEGO bricks in a specified time. The countries either had an abundance or not enough LEGO bricks. Critically, how they should go about achieving this goal was ambiguous; the teenagers were not told whether they should compete or cooperate with the other nations. The students were free to move around the room during the exercise and observe the other countries and their resources. During the exercise, none of the high-income countries spontaneously shared any LEGO bricks with the low-income countries even when those low-income countries asked for help. This suggests that inequality may be a suppressor of behavior in line with a superordinate goal. However, even though these findings are noteworthy, the activity was designed as a learning opportunity rather than a scientifically valid study and these findings should be interpreted with caution given the lack of experimental control. In the current study, we transformed the activity by Compassion It into a rigorous and highly controlled experimental design . Participants were randomly assigned to groups that represented either a high or a low resource country and each country was presented with LEGO bricks, whereby the high resource group had an abundance of LEGO bricks and the low resource group had very little. Participants were asked to assemble LEGO bricks into food items to ensure that "no one will starve. " Critically, the means by which participants should achieve this goal was ambiguous; we did not tell the groups to compete or cooperate. The compassionate, superordinate goal was to maximize the food creation to ensure no one will starve, and any behavior that contributed to this goal was considered an indication of compassion. To achieve this, we coded for three forms of compassionate behavior: initiating sharing of items, resulting in a transfer of LEGO bricks from the high resource group to the low resource group, the amount of food pieces made, and the efficiency of food making. We further included several self-report measures that broadly measured competitive and cooperative attitudes toward the other group, group cohesion and fears of showing and receiving compassion. --- Participants and design The sample was comprised of 283 participants and were 21.36 years-old on average . Based on a sample size calculation on Pangea, we required a minimum sample size of 152 to detect a medium effect size with 80% probability. This effect size was deemed appropriate based on findings regarding competitive sentiments under situations of high inequality . We aimed to achieve this sample size at a minimum and collected larger numbers until the participant pool was exhausted. The data was not analyzed until data collection ceased. Participants were recruited from either a first-year pool of psychology students from a large urban university or from a paid-pool and were reimbursed $10 per half an hour of participation. Participants reported several demographic variables including their age, gender, level of education, ethnicity, total pre-tax income and subjective SES. Subjective SES refers to where one feels they fit into society relative to others in terms of job prestige, education, and income on a 10-rung ladder . On average, participants reported having middle class backgrounds . See Supplementary materials 2 for the full demographic description of the sample used. The current study employed a 2 by 2 between-subjects design, and participants were randomly assigned to a condition and resource group. We were interested in the effect of these independent variables on outcomes described in greater detail below. --- Procedure Before the study began, the table and chairs were arranged in a way to clearly separate two groups: the high resource group and the low resource group. Each session contained between 4 and 12 participants. In the event of odd numbers, the extra participant was assigned to the high resource group. The table held two transparent containers with LEGO bricks, with one container assigned to each resource group. Images of these containers can be provided to readers upon request from the corresponding author. The high resource group container held 500 colored LEGO bricks and 100 non-colored bricks . In contrast, the low resource group container contained only 100 colored bricks and 500 non-colored LEGO bricks. Importantly, the valued resource in this context was colored LEGO bricks, whereas non-colored bricks held no value. As participants entered the room, they were randomly assigned to sit at the high or low resource group side of the table . After consent was obtained, participants were asked to listen to a 10-min meditation audio track which was played aloud to the entire group. Each session was randomly assigned to engage with a compassion meditation track or a focussed imagery track. The compassion track began with basic meditation instructions, before telling participants about the definition of compassion, asked them to contemplate the definition and imagine engaging in compassionate behavior. Our focussed imagery meditation condition served as an ideal control task, as it contained basic meditative practices but did not contain any information about compassion. In line with past approaches , this allowed us to isolate the effect of reflecting on being compassionate from the practice of general mindfulness. See Supplementary materials 4 for full scripts of each meditation. The experimenter then told participants they were separated into two 'countries': Nasherland and Lindithia . We chose fictitious countries as real countries may prime stereotypes in participants' minds about behavior in that specific culture and the use of fictitious nations has been successful in past experiments . Participants read a basic description of their country which contained demographic information such as the local delicacy, the population and the climate . The experimenter then told participants the aim was to create as much food as possible within a 5-min period to prevent starvation. Participants were told one piece of food could be created by assembling LEGO bricks in the following order : blue, green, yellow, red. Participants were then shown an image of a correctly assembled food item. This image can be provided to readers upon request from the corresponding author. These instructions were purposefully ambiguous; we did not tell the groups to cooperate or compete as we were interested in how they would interpret the ambiguous situation. As such, if participants asked if they could share bricks, they were told "The aim is to make as much food as possible so no one will starve. " We pilot tested the LEGO brick distribution to ensure the high resource group could not finish assembling their LEGO bricks in the time given whereas the low resource group would always finish assembling their pieces with excess time left. Implicitly, it was clear that the only way to maximize food creation was for the groups to work together. The groups then had 5 min. to assemble food, and the participants were then asked to complete a questionnaire at the conclusion of the task .2 Participants were then debriefed and thanked for their participation. --- Measures Compassionate behavior We defined compassionate behavior as actions that would contribute to the superordinate goal of creating food for the starving people more broadly. Here, any action that results in maximizing food creation preventing starvation was counted as compassionate behavior. First, we assessed whether individuals initiated sharing as well as the amount of LEGO bricks that were transferred from the high to low resource group per individual. Importantly, sharing Frontiers in Psychology 05 frontiersin.org could be initiated by the high or low resource group. Table 1 demonstrates the kinds of behavior that were counted as initiating sharing per resource group, and whether that instance of sharing was initiated by the high or low resource group. This coding system meant that both the high and low resource group could engage in the sharing of LEGO bricks between the groups. In addition, we assessed the number of correctly assembled food pieces made per individual. Finally, we assessed the food making efficiency of each resource group. 3 --- Fears of compassion The questionnaire contained a fears of compassion scale , as past work has found people can be fearful of being compassionate to others because it could result in resource loss and be fearful of receiving compassion from others due to obligations to return the help . We included these measures as exploratory additions to the study, as fears of giving and receiving compassion may be a significant barrier to coordinated action. Participants were asked 10-items that reflected fears of giving compassion and 13 items that gauged fears of receiving compassion . Responses per item were scored from 0 to 4 , and the responses were added together for each participant to achieve a total score. For fears of giving compassion, the total score could range from 0 to 40 , and for fears of receiving compassion, the total score could range from 0 to 52 . The fears of giving and fears of receiving compassion scales both yielded acceptable reliability . --- Group dynamics and cohesion In the current study, the means by which the groups should achieve the goal , we have reported the overall efficiency of the resource group. food so no one starves) was purposefully ambiguous. To assess the participants' interpretation of these ambiguous instructions, we included three exploratory questions to ascertain whether they interpreted the task as competitive or cooperative. Participants were asked "To what extent did you feel this task was a competition between the two countries?, " "To what extent did you feel this task was a cooperative task between the two countries?" and "To what extent did you feel the context was one of "US" versus "THEM" . " Responses were scored on a scale from 1 to 10 . We further asked two exploratory questions to gauge how participants felt about the cohesiveness of their group, as strong ingroup unity may act as a suppressor of coordinated action. Specifically, participants were provided with the following statements: "I felt a sense of unity within my group" and "I felt that people in my group seemed to be on the same wavelength. " Responses were scored on a scale from 1 to 10 , and an average score of these two items was created . --- Attention checks Finally, participants were asked several questions probing their attention to the inequality as well as a manipulation check to assess feelings of compassion. First, inequality salience was measured with the following question: "During the activity, to what extent did you notice the groups were unequal?. " This question was scored on a scale from 1 to 10 . Second, we included a measure to ensure the high resource group felt like they had a greater capacity to complete the task compared to the low resource group: "My group had enough LEGO bricks to complete the task. " Responses were scored from 1 to 10 . Participants were also asked "To what extent did listening to the audio track make you feel more compassionate?, " and responses were recorded on a scale from 1 to 10 . --- Analytical approach In our design, individual behavior was potentially impacted by the behavior of their group members. For example, if one group member decided to share, this may have influenced other group members to share as well. To adjust for this non-independence of data, all individual level measures were analyzed in Linear Mixed Models with 'group' as the random intercept. --- Results See Supplementary materials 8 for the full results for each analysis, including mean differences between conditions and resource groups. An independent samples t-test showed that there was no significant difference in the sizes of groups randomly allocated to the compassion meditation and focussed imagery conditions, t = 1.32, p = 0.187. This variable was thus not considered further. The conditions and resource groups also did not differ in terms of age and gender. See Supplementary materials 9 for the means and standard deviations per condition, per resource group for each of the dependent variables. --- Attention checks Overall, participants were highly cognizant of the unequal resources between the two groups . An LMM was conducted on the effect of resource group and condition on the extent to which the participants noticed the inequality. There were no differences between resource groups, F = 0.50, p = 0.482, or conditions, F = 2.28, p = 0.137, in the extent to which participants noticed the inequality. Likewise, there was no significant interaction between resource conditions and compassion manipulation conditions, F = 3.45, p = 0.069. We further assessed the extent to which participants felt they had enough LEGO bricks to complete the task. An ANOVA4 revealed a significant effect of resource group, F = 313.62, p < 0.001, where the high resource group indicated more strongly than the low resource group that they had enough LEGO bricks to complete the task. However, there was no significant effect of condition, F = 0.03, p = 0.852, and no condition by resource group interaction, F = 0.04, p = 0.846. Finally, we assessed whether participants felt more compassionate after listening to the compassion compared to the focussed imagery meditation as a manipulation check. An ANOVA revealed participants in the compassion meditation condition felt more compassionate compared to those in the focussed imagery condition , F = 18.29, p < 0.001. Moreover, there was no significant effect of resource group on feelings of compassion, F = 0.06, p = 0.805, nor was there a significant interaction between resource group and condition, F = 0.42, p = 0.516. --- Compassionate behavior In total, 15.1% of participants initiated some form of sharing. A GLMM was conducted to establish the effect of resource group and condition on whether an individual initiated sharing . Results revealed no significant effect of resource group, X 2 = 0.56, p = 0.456, or condition, X 2 = 0.29, p = 0.590, nor a significant interaction between the two variables, X 2 = 0.69, p = 0.407. See Supplementary materials 10 for the number of times each category of sharing behavior was observed. Altogether, individual participants initiated the sharing of 2.93 LEGO bricks on average. A GLMM assessed the effect of resource group and condition on the number of LEGO bricks transferred when sharing was initiated. For this model, we used a Poisson distribution and the square root link function due to the exponential nature of the dependent variable. Results revealed no significant effect of resource group, IRR = 0.82, p = 0.204, or condition, IRR = 1.08, p = 0.647, on the number of LEGO bricks transferred when sharing was initiated. Additionally, there was no significant interaction observed between the two variables, IRR = 1.24, p = 0.184. Collapsed across conditions and resource groups, participants assembled 9.35 food pieces on average. The effect of condition and resource group on the number of food pieces made was assessed using an LMM. More food pieces were made by the high resource group compared to the low resource group , F = 18.36, p < 0.001. However, there was no significant difference observed between the conditions and the number of food pieces assembled, F = 1.17, p = 0.284, nor was there a significant interaction between the two variables, F = 0.07, p = 0.789. On average, the groups made approximately 6.93 food pieces per minute. We assessed the effect of condition and resource group on the efficiency of LEGO brick assembly using a two-way ANOVA. The high resource group worked faster compared to the low resource group , F = 41.83, p < 0.001. However, there was no significant difference in work rate based on condition, F = 0.03, p = 0.857, nor was there a significant interaction between condition and resource group, F = 0.08, p = 0.773. --- Exploratory analyses We conducted several exploratory Linear Mixed Models examining the effect of condition and resource group on fears of compassion as well as group dynamics and cohesion. As demonstrated in Table 2, those in the low resource group felt their groups were more cohesive relative to those in the high resource group . --- Discussion Study 1 assessed the effect of compassion meditation on working toward a shared and superordinate goal when groups . The finding also suggests that while brief compassion training may increase feelings of compassion, this may not translate into more compassionate behavior. Study 1 demonstrated that a brief standalone compassion meditation did not result in greater collaboration across boundaries of groups that are unequal. It appears that this individualistic approach may not be effective when individuals are members of groups. Here, the dynamics of the group may have a strong influence on an individual's behavior, and interventions that target structural elements may instead be more effective. Lower inequality, for example, is thought to reduce "us" versus "them" dynamics between different resource groups , potentially paving the way for greater coordinated action. --- Study 2 In line with classic social identity theorizing , Study 2 examined whether the structural context shapes group behavior. We proposed that structural factors, as opposed to individual factors , may be a more important determinate of whether groups adopt superordinate goals. Previous research has shown that intergroup competition is lower when individuals or groups have more equal resources than when inequality of resources is high , and cooperation declines when inequality in resources is highly visible . Following from this research, we explored whether lower inequality would result in more behavior that is in line with a superordinate goal. To examine this, we placed participants in two groups where the difference in group resources was either moderately or extremely unequal. 5 Since lower inequality reduces competition, we expected that participants in groups that experienced moderate inequality would be more likely to act in line with a superordinate goal relative to groups in extreme inequality. 6 --- Method Our methods and analytical approach were identical to that described in Study 1, apart from the deviations detailed below. --- Participants and design The sample was comprised of 173 participants and were 20.98 years-old on average . Our approach to sample size and recruitment was identical to that described in Study 1. On average, participants reported having a middle-class background . See Supplementary materials 2 for the full demographic description of the sample used. The current study employed a 2 by 2 between-subjects design, where participants were randomly assigned to a condition and a resource group. 7 5 To test this, we heightened inequality relative to Study 1 . This decision was made because, in our paradigm, groups that possess equal resources would have no superordinate reason to collaborate. Hence, heightening the differences between groups is the only feasible way to test the role of inequality in this paradigm. 6 We originally preregistered a null hypothesis by mistake , and believe this does not accurately reflect the body of research existing at the time. We have changed our hypothesis to better reflect the state of the literature prior to creating this study. 7 It is worth noting that this manipulation creates a natural confound. That is, when the inequality changes between the groups , we also change the level of resources each group has; the low resource group in moderate inequality has more resources than the low resource group in extreme inequality. However, this is unavoidable as manipulating the gap between the low and the high resource group also varies the absolute amount of wealth each group has. We were interested in the effect of these independent variables on several outcomes including compassionate behavior, fears of compassion and group cohesion. --- Procedure We followed an identical procedure to that described in Study 1 with a few exceptions. First, participants did not listen to an audio meditation and instead were given the task instructions immediately after giving their consent. Second, participants experienced one of two LEGO brick distributions. In line with Study 1, participants in the moderate inequality condition were in a context where the high resource group was given 500 colored and 100 non-colored bricks and the low resource group was given 100 colored and 500 non-colored bricks. We increased the magnitude of this inequality in the extreme inequality condition, where the high resource group had 560 colored and 100 non-colored bricks and the low resource group was given 40 colored and 500 non-colored bricks. In addition, those in the extreme inequality condition were given additional information about the wealth of their country . --- Measures All measures were identical to Study 1 . However, we did not include the manipulation check measure that assessed how compassionate participants felt in response to the meditation. --- Results In total, 31 experimental sessions were used for the final sample and group sizes ranged from four to nine. The full results for each analysis from Study 2 can be found in Supplementary materials 8. An independent samples t-test was conducted to establish whether the conditions differed in the size of the groups, and results revealed no significant difference between the extreme and moderate inequality conditions, t = 0.38, p = 0.708. Group size was thus not considered further. The conditions and resource groups also did not differ in terms of age and gender. See Supplementary materials 9 for the means and standard deviations per condition, per resource group for each of the dependent variables. --- Attention checks Overall, participants were highly cognizant of the unequal resources between the two groups . An LMM was conducted on the effect of resource group and condition on the extent to which the participants noticed the inequality. There were no differences between resource groups, F = 0.72, p = 0.401, or conditions, F = 2.26, p = 0.139, in the extent to which participants noticed the inequality. Likewise, there was no significant interaction between the two variables, F = 0.02, p = 0.879. We further assessed the extent to which participants felt they had enough LEGO bricks to complete the task. An LMM was used to assess the effect of resource group and condition on this variable. A significant effect of resource group was found, F = 68.23, p < 0.001, where the high resource group felt more so than the low resource group that they had enough LEGO bricks to complete the task. However, there was no significant effect of condition, F = 0.002, p = 0.969, and no condition by resource group interaction, F = 2.39, p = 0.128. --- Compassionate behavior In total, 28.3% of participants initiated some form of sharing. A Generalized Linear Mixed Model was conducted to establish the effect of resource group and condition on whether an individual initiated sharing . Results revealed no significant effect of resource group, X 2 = 0.02, p = 0.896, or condition, X 2 = 0.37, p = 0.543, nor a significant interaction between the two variables, X 2 = 0.02, p = 0.878. See Supplementary materials 10 for the number of times each category of sharing behavior was observed. Altogether, individual participants initiated the sharing of 5.27 LEGO bricks on average. A GLMM assessed the effect of resource group and condition on the number of LEGO bricks transferred when sharing was initiated. For this model, we used a Poisson distribution and the square root link function due to the exponential nature of the dependent variable . The low resource group initiated the sharing of more LEGO bricks compared to the high resource group , IRR = 0.50, p = 0.004. Likewise, more Frontiers in Psychology 09 frontiersin.org LEGO bricks were transferred in instances of sharing in the extreme compared to the moderate inequality condition , IRR = 1.80, p = 0.012. However, there was no significant interaction observed between the two variables, IRR = 0.81, p = 0.365. Collapsed across conditions and resource groups, participants assembled 11.14 food pieces on average. The effect of condition and resource group on the number of food pieces made was assessed using a Linear Mixed Model . More food pieces were made by the high resource group compared to the low resource group , F = 25.89, p < 0.001. However, there was no significant difference observed between the conditions and the number of food pieces assembled, F = 3.43, p = 0.070, nor was there a significant interaction between the two variables, F = 0.04, p = 0.849. On average, the groups made approximately 6.78 food pieces per minute. We assessed the effect of condition and resource group on the efficiency of LEGO brick assembly using a two-way ANOVA. The high resource group worked faster compared to the low resource group , F = 25.70, p < 0.001. However, there was no significant difference in work rate based on condition, F = 1.73, p = 0.193, nor was there a significant interaction between condition and resource group, F = 0.34, p = 0.561. --- Exploratory analyses We conducted several exploratory Linear Mixed Models examining the effect of condition and resource group on fears of compassion as well as group dynamics and cohesion. As demonstrated in Table 3, those in the low resource group reported lower fears of receiving compassion relative to those in the high resource group . Those in the extreme inequality condition felt there was a greater cohesion in the resource group compared to those who experienced moderate inequality . --- Discussion In Study 2, we analyzed the effect of the degree of inequality on the adoption of behavior directed toward a superordinate goal. We found no consistent support for our hypothesis and results show that participants in the moderate inequality condition did not behave in line with a superordinate goal more than those in the extreme inequality condition. Overall, we found that there were no differences in whether sharing was initiated between conditions. However, when sharing was initiated in the extreme inequality condition, more LEGO bricks were involved in that transfer compared to moderate inequality. Importantly, we found no differences in the extent to which participants noticed the inequality, and both conditions yielded ceiling effects; inequality was highly salient to participants in both conditions. Low resource participants also initiated sharing more than high resource participants and reported reduced fears of receiving compassion. Moreover, our manipulation did not result in differences in feelings of competitiveness or cooperativeness. However, extreme inequality did result in greater cohesion with the group, suggesting that when differences between groups are enhanced, participants feel more united with their ingroup . These results in combination suggest that our manipulation of the degree of inequality did not promote the adoption of compassionate behavior and attitudes. Instead, interventions that alter the normative structure, such as an informal leader promoting a pro-sharing group norm, may result in more compassionate action when groups are unequal. --- Study 3 The behavior of others around us, and in particular, the members of our ingroup, can have a dramatic effect on how we choose to act . In particular, highlighting norms about what individuals should do tends to enhance prosocial behavior . We also typically favor members The four LEGO bricks shared by the confederate at prompt 4 were not counted in the sharing score. of our own group over members of other groups, even if the group membership is dictated by something as arbitrary as a similar colored shirt . However, when group members promote a norm that helps outgroup members, ingroup favoritism can be overridden . Likewise, when participants are prompted to reflect on what they should do, they are less likely to favor the ingroup over the outgroup . Ambiguous situations present a particular challenge for groups . Because of this, an individual who introduces a pro-sharing group norm can become an informal leader and guide their group toward superordinate action. Past research has shown that informal leaders who offer cognitive alternatives -that is, alternatives to the current reality -can have a powerful impact on the behavior of other members . Study 3 aimed to explore the influence of a pro-sharing group norm on the emergence of superordinate, compassionate behavior when groups are unequal. To assess this, we utilized the same design from the moderate inequality condition in Study 2. This time, a confederate was planted in the high resource group. In our pro-sharing group norm condition, the confederate gradually prompted sharing between the groups with increasing intensity over the five-minute task period. This was compared to a control condition where the confederate instead discussed their enjoyment of LEGO bricks. While the confederate was acting as a high resource group member, they had the potential to sway the behaviour of members from both the high and low resource group. In line with prior research , we hypothesised that more compassionate behavior would be exhibited by all participants in the pro-sharing group norm condition compared to the control condition. --- Method Our methods and analytical approach were identical to the moderate inequality condition in Study 2, with exceptions outlined below. --- Participants and design The sample was comprised of 160 participants and were 20.36 years-old on average . Our approach to sample size and recruitment was identical to that described in Study 1. On average, participants reported having a middle-class background . See Supplementary materials 2 for the full demographic description of the sample used. The current study employed a 2 by 2 between-subjects design, and participants were randomly assigned to a condition and a resource group. We were interested in the effect of these independent variables on a number of outcomes including compassionate behavior, competitive sentiments, fears of compassion and group dynamics. --- Procedure The study followed an identical procedure to the moderate inequality condition in Study 2 with a few exceptions. Participants either experienced the implementation of a pro-sharing group norm or a control condition, and this was achieved by including a confederate in the high resource group. The confederate took on an informal leadership position in the group and spoke only during the LEGO brick assembly task. In both conditions, they spoke at one-minute intervals and were instructed to only speak to group members when spoken to. In the pro-sharing group norm condition, the prompts escalated in their intensity. The confederate first pointed out the LEGO brick inequality, then created an injunctive norm where they suggested sharing. Eventually they themselves physically shared bricks. In the control condition, the confederate spoke about their enjoyment of LEGO bricks at each minute interval. The specific prompts are outlined in Table 4. The confederate was instructed to work at a similar rate to the other group members. We chose to have a control confederate rather than a no confederate condition to control for any effects consistent discussion might have on participant behavior. That is, a confederate who speaks frequently -regardless of what they speak about -might promote a different group dynamic and this may change how participants behave. --- Measures All measures were identical to Study 2 . --- Results An independent samples t-test was conducted to examine whether the size of the groups were identical across the Frontiers in Psychology 12 frontiersin.org --- Exploratory analyses We conducted several exploratory Linear Mixed Models examining the effect of condition and resource group on fears of compassion as well as group dynamics and cohesion. As demonstrated in Table 5, there was a significant interaction between resource groups and conditions on fears of giving compassion . Follow up simple effect analyses revealed a significant effect for the high resource group only, F = 7.64, p = 0.008, such that the high resource group members in the compassionate norm condition had reduced fears of being compassionate compared to high resource group members in the control condition . Moreover, those in the low resource group felt more like the context was competitive compared to those in the high resource group . Participants in the pro-sharing group norm condition felt the context was more cooperative compared to those in the control condition . Likewise, participants in the pro-sharing group norm condition felt the context was less one of "us" versus "them" compared to those in the control condition . --- Discussion Study 3 revealed the effect of a confederate who took on an informal leadership role in the group and changed the status quo by introducing a pro-sharing group norm. We found that a fellow group member who promotes sharing can have a significant influence on the behavior of other groups members-participants were more likely to initiate sharing, transferred more LEGO bricks between the groups, worked faster and made more food pieces when a confederate suggested sharing. Importantly, participants in this condition felt that the context was more cooperative and less one of 'us' vs. 'them' , and the high resource participants had reduced fears of being compassionate. Together, these findings suggest that a member who imbeds a pro-sharing group norm within a group can have a powerful influence on the behavior and mentality of other group members-their leadership behavior decreases intergroup competition, and compassionate action can be achieved. --- General discussion When Covid-19 surged, wealthy countries hoarded vaccines, and poorer countries -who were not only battling poverty but also the devastating impact of the virus -were left without . How do we promote compassionate action under these conditions? Here we aimed to understand effective ways of orienting individuals toward a superordinate, compassionate goal when they were embedded in unequal groups, via an individualistic intervention , structural intervention , and a normative intervention promoting a different way to respond to the status quo . We found little evidence that compassion meditation and varying the degree of inequality enhanced the adoption of compassionate action. However, when a confederate took the lead by introducing a pro-sharing group norm, participants engaged in more compassionate behavior and adopted a collaborative approach to the task. Frontiers in Psychology 13 frontiersin.org The introduction of a pro-sharing group norm resulted in enhanced compassionate action and attitudes, and this finding is in line with past work suggesting norms can have a significant shift on group behavior . Additionally, past research shows that individuals who offer a cognitive alternative to the current status quo can become informal leaders and sway the behavior of their group . Here participants transferred more LEGO bricks between the groups, interpreted the task as cooperative, worked faster and, critically, created more food for 'starving people' . Moreover, the high resource group had reduced fears of being compassionate relative to the control condition, suggesting that the introduction of a pro-sharing group norm paved the way for group members to feel more positive about behaving compassionately toward others. It remains unclear however if the groups would adopt superordinate behavior if the confederate was instead a member of the low resource group, and this is a promising direction for future research. It is also unclear whether the confederate introduced a norm of sharing as intended or whether their comments drew attention to the unequal resources, and this instead prompted sharing. While participants reported noticing the inequality more when the confederate introduced the sharing norm, they were still highly aware of the inequality in the control condition. Nonetheless, future research should include questions about how participants view the norms of the group and assess whether this altered by condition. On the other hand, our individualistic intervention -a compassion meditation -did not promote compassionate action or attitudes. While participants reported feeling more compassionate, there was no evidence that this translated to behavior -a phenomenon that is in line with past research demonstrating a gap between attitudes and behavior . This further suggests that while compassion meditations may alter attitudes and behavior in some settings , such interventions may be too individualistic to affect compassion in a group setting. However, we only assessed the effect of short-term interventions and longer-term interventions may instead prove fruitful. Likewise, our structural intervention -varying the degree of inequality between the groups -also did not result in any meaningful changes in compassionate actions or attitudes. While enhanced inequality did lead to more LEGO bricks being transferred between the groups , this did not result in more food pieces being made. This intervention also did not impact attitudes, and this may have been because the situation invoked two competing motivations; while the need for sharing was more tangible under extreme inequality, unequal resources suppress compassionate action . Together, these three studies have revealed several insights about human behavior in a previously unexplored context. We assessed the effect of three different interventions from diverse literatures to establish which approach is most effective. The efficacy of these interventions was measured across a variety of behavioral and self-report outcomes, giving us greater certainty of the effects. In addition to theoretical contributions, these studies also have significant practical applications. In a world of increasingly complex social dilemmas, there have been growing discussions about how to promote a more compassionate world -for example by getting rich countries to assist poor countries in their acquisition of Covid-19 vaccinations . Our work suggests that leadership by one individual, whether it be an individual person or possibly an individual country, may set a norm that can have a positive domino effect on compassionate actions more broadly. --- Limitations and future research Despite these strengths, our work has produced several questions that warrant future research. While our experimental approach allowed us to gain a high degree of control, the assembly of LEGO bricks is distantly related to the acquisition of real-world resources. Thus, future work is needed to assess effects of these kinds of manipulations in real-world settings. Additionally, the endowments were windfall gains, and people tend to be less generous with resources when they are instead earned . To test this possibility, future work should compare the effect of windfall versus earned resources on intergroup interactions in this context. Moreover, future work may wish to also vary the degree of inequality within-groups and explore how this interacts with betweengroup inequality. We have also defined compassion as any action that aims to maximize the food creation for 'starving people' and placed behavior such as sharing and food assembly under this definition. However, this may not be the only motivation that is driving participants to engage in sharing and food creation. For example, high resource participants may feel pity or awkwardness directed toward the low resource group due to their lack of LEGO bricks. Future work should directly assess the motivations that drive participant sharing behavior and establish whether these are compassionate in origin. We also did not directly compare interventions across studies, and future research may wish to test which manipulations yield the largest effect size. Finally, participants were disproportionately female, largely comprised of first-year students and came from a W.E.I.R.D. population . Frontiers in Psychology 14 frontiersin.org Future work should extend upon these findings in more representative and culturally diverse samples. --- In conclusion The human capacity for compassion is one of our most extraordinary traits, yet we do not always help those who are suffering. Here we aimed to establish how to foster compassionate action and promote the adoption of a superordinate goal under situations of group inequality. We assessed the effect of three interventions: compassion meditation, altering the degree of inequality, and implementing a pro-sharing group norm. Compassion meditation and changing the degree of inequality had no meaningful effect on compassionate action. The introduction of a pro-sharing group norm instead had a marked influence on the behavior and attitudes of the unequal groups. This work offers new insights into the feasibility of different interventions to foster compassionate behavior, which may be critical in promoting a more unified world. --- Data availability statement The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number can be found at: Open Science Framework: https://osf. io/fjp6b/?view_only=b51cca6755654a76b8d1c8d77a3cfa53. --- Frontiers in Psychology 11 frontiersin.org conditions. Results revealed a significant difference such that groups were smaller in size in the pro-sharing group norm condition compared to the control condition , t = -6.15, p < 0.001. This difference emerged despite careful random allocation procedures, and we thus used group size as a covariate in all analyses. The full results for each analysis from Study 3 can be found in Supplementary materials 8. The conditions and resource groups did not differ in terms of age and gender. See Supplementary materials 9 for the means and standard deviations per condition, per resource group for each of the dependent variables. --- Attention checks Overall, participants were highly cognizant of the unequal resources between the two groups . An LMM was conducted on the effect of resource group and condition on the extent to which the participants noticed the inequality. Those in the pro-sharing group norm condition noticed the inequality more compared to those in the control condition , F = 6.21, p = 0.016. However, there were no differences between resource groups in the extent to which participants noticed the inequality, F = 0.80, p = 0.374. Likewise, there was no significant interaction between the two variables, F = 0.51, p = 0.480. We further assessed the extent to which participants felt they had enough LEGO bricks to complete the task. An ANCOVA revealed a significant effect of resource group, F = 370.15, p < 0.001, where the high resource group felt they had enough LEGO bricks to complete the task more so than the low resource group . However, there was no significant effect of condition, F = 2.51, p = 0.115, and no condition by resource group interaction, F = 0.66, p = 0.418. --- Compassionate behavior In total, 40.6% of participants initiated some form of sharing. A GLMM was conducted to establish the effect of resource group and condition on whether an individual initiated sharing . Results revealed that participants were more likely to initiate sharing in the pro-sharing group norm condition compared to the control condition , X 2 = 18.32, p < 0.001. There was no significant effect of resource group, X 2 < 0.01, p = 0.945, nor a significant interaction between the two variables, X 2 < 0.01, p = 0.972. See Supplementary materials 10 for the number of times each category of sharing behavior was observed. Altogether, individual participants initiated the sharing of 6.02 LEGO bricks on average. A GLMM assessed the effect of resource group and condition on the number of LEGO bricks transferred when sharing was initiated. For this model, we used a Poisson distribution and the square root link function due to the exponential nature of the dependent variable. More LEGO bricks were transferred in instances of sharing in the pro-sharing group norm condition compared to the control condition , IRR = 2.88, p < 0.001. Results additionally revealed no significant effect of resource group, IRR = 1.26, p = 0.375, and there was no significant interaction observed between the two variables, IRR = 1.02, p = 0.926. Collapsed across conditions and resource groups, participants assembled 10.58 food pieces on average. The effect of condition and resource group on the number of food pieces made was assessed using an LMM . More food pieces were made by the high resource group compared to the low resource group , F = 18.78, p < 0.001. Furthermore, more food pieces were assembled in the pro-sharing group norm condition compared to the control condition , F = 6.30, p = 0.015. However, there was no significant interaction between the two variables, F = 0.19, p = 0.661. On average, the groups made approximately 6.53 food pieces per minute. We assessed the effect of condition and resource group on the efficiency of LEGO brick assembly using a two-way ANCOVA. The high resource group worked faster compared to the low resource group , F = 88.50, p < 0.001. Those in the pro-sharing group norm condition were also more efficient at making food relative to those in the control condition , F = 5.96, p = 0.018. Accounting for group size, the estimated marginal mean for efficiency in food assembly was higher for the pro-sharing group norm condition compared to the control condition . However, there was no significant interaction between condition and resource group, F = 1.36, p = 0.248. --- Ethics statement The studies involving human participants were reviewed and approved by Human Research Ethics Committee at the University of Queensland . 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. --- Author Disclaimer LEGO ® , the LEGO ® logo, the Brick and Knob configuration and the MINIFIGURE figurine are trademarks and/or copyrights of the LEGO Group of Companies, which does not sponsor, authorize or endorse this article. --- Supplementary material The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2022.1021093/ full#supplementary-material

Amid a global pandemic and the climate crisis, there is an increasing need to understand how to promote largescale, coordinated action between different groups. Yet certain factors such as inequality can hinder cooperation. We aimed to establish how to orient groups toward a superordinate goal when they have unequal resources. Participants were divided into two 'countries' and asked to assemble LEGO bricks into food (by building pieces in a certain order) to prevent starvation among 'the people'. One 'country' had few LEGO bricks whereas the other had an abundance, and the only way to maximize food creation was for the groups to work together. We assessed the efficacy of three diverse interventions on superordinate behavior and attitudes: compassion meditation training (Study 1), lower inequality (Study 2), and the introduction of a pro-sharing group norm by a confederate (Study 3). Compassion meditation training and altering the degree of inequality between groups did not have a clear effect on collaborative action. Only the introduction of a pro-sharing group norm enhanced sharing behavior, made participants feel more cooperative and reduced fears of being compassionate toward others. Our findings speak to the importance of leadership in promoting coordinated action to address challenges that face the superordinate group.

The lives of millions of displaced individuals are characterized by exclusion, xenophobia, and global apathy. Displaced persons may live in formal or informal camps, or in other settings with few resources and unreliable access to essential commodities and services. Forced displacement, both in the short term and protracted, therefore has complex impacts on health. 1,2 Factors that affect the health of forcibly displaced individuals can be split broadly into two realms: the physical and the social. Displaced communities, both during migration and when they are not moving, are often in locations that have limited access to essential health services. The geographic locations also often increase exposure to toxic and unstable environments, which can have immediate and chronic health impacts. The social attributes of displacement include legal status, nationality, and other socioeconomic factors that can likewise influence access to health services , and force communities to live in a state of fear and anxiety. They may also affect the geographic settings in which displaced persons reside. As humanitarian crises grow around the world and new humanitarian crises appear in South Asia, the Americas, Africa, and Europe, so do the camps that are sometimes home to displaced individuals for decades. [3][4][5] In many camps across the globe, there are multiple generations that have been born and spent their entire lives in camps. The populations of these camps are neither fixed nor static, which can strongly influence epidemiological dynamics. [6][7][8][9] Intergenerational factors create heterogeneity that leads to environmental changes, including those at microbial scales, not excluding changes in the gut microbiome. Changes in demography and environment are rarely studied from an integrated ecological and/or evolutionary lens, and are rarely part of the displaced individuals' health landscape. This absence of scholarship and research allows for the continuation of siloed approaches and creates blind spots in understanding disease dynamics, emergence of new health challenges, provision of long-term care, and inefficient delivery of health services. We propose a more integrated approach to understand health challenges of the forcibly displaced-one that is multidisciplinary and multi-scale, and analyzes the communities and their environment from an ecological framework. We define ecology as the interaction between organisms and their environment at multiple spatial scales. This includes interactions among microbes; among humans, disease vectors, and microbes; and between the physical landscape and humans, disease vectors, and microbes; and does not exclude noncommunicable diseases. Current approaches, even in the context of a One Health framework, can sometimes be siloed or at a single spatial or temporal scale. The common siloed approach through which many services are currently provided leads to inefficiencies in the provision of services, varying temporal coverage based on funding cycles and incomplete pictures of health. Environmental, ecological, and evolutionary considerations are often not even considered, posing further gaps in our understanding of the health of populations within such camps. Though we have information on a higher, general level, we know few specifics about disease dynamics and the relationship of people and their environment in complex humanitarian emergencies. 10 Siloed approaches neglect the reality that all components of humanitarian management have important health considerations when using an ecological lens. We therefore aim to understand the dynamic ecologies that connect vertically and horizontally in forced displacement camps. Understanding the ways that upstream factors can influence human health and microbial communities is useful in diagnosing and preventing poor health outcomes. Likewise, this type of big-picture lens is useful from a public health standpoint. A public health system that focuses on all health outcomes-communicable and noncommunicable, including mental health-and likewise considers the multidirectional health implications of human-biophysical environment interactions is useful for preventing disease. For those who are planning or administering camps, having a holistic view of the linkages between humans and their environments may lead to better planned and managed camps. Camp locations could often be better chosen, with health implications for the dwellers, for other communities in the same region, and for the environment. An approach that prioritizes local knowledge will lead to better understanding of the local ecological systems. To achieve this, we argue for considering displaced individual environs as adaptive ecosystems, with interactions at multiple temporal and spatial scales. An ecological lens challenges the siloed approach and allows for benefiting from a new understanding and new tools in the discipline. It helps us better understand the evolutionary and environmental pressures in the camps, and enables us to improve the lives of those who live in them. --- ENVIRONMENTAL AND SOCIAL ECOLOGIES Humans, animals, and the environment influence each other in a codependent manner that varies with time and space, and is affected by both local and global factors. This multiscale dependence, although well recognized by those who are forced to live in the camps-has not been fully understood in research circles. To illustrate the importance of a multiscale dynamic ecological model in understanding the health of displaced individuals in camps, we offer Figure 1, which demonstrates and visualizes some of the ecologies and interactions occurring between the environment and health of displaced populations. Social factors also play a critical role in the camp ecosystem and include the camp's and individuals' legal status, discrimination, mobility, employment, and host community relations. The camp's legal status affects the provision and availability of services throughout the camp, as well as its development and care. The legal status of displaced individuals influences movement out of the camp, including being unable to seek external health care or obtain employment. Employment that is obtained may be in environments where certain diseases are more common. Differences in race, gender, legal status, and other demographic variables can lead to discrimination, inequality, and violence. These have direct and indirect effects on the health of displaced persons, and are compounded further by the individual's vulnerable social and legal position, and lack of legal accountability for perpetrators. Studies have also found an association between gender inequality and environmental degradation. 11 These social factors interact with the broader environment and work in tandem to influence health outcomes, including those related to mental health. Table 1 summarizes a sampling of these factors, though it is by no means an exhaustive discussion of these factors and interactions. --- DISCUSSION The United Nations High Commissioner for Refugees views camps as "temporary solutions of last resort ," yet the formation of camps is so common that it warrants better planning. 39 Although we fully acknowledge that displaced persons should not be confined to camps and strongly argue for dignified living conditions for those who have been forced to leave their homes, we argue that for those who are in camps, an expanded understanding of the Built environment 12 Infrastructure, including dwellings and water/sewage systems, that are poorly constructed and unsafe can increase exposure to toxic and unstable environments and disease. Physical landscape 13 The physical landscape informs disease ecology, affects resource availability, and increases vulnerability to weather events. Camp location 12,14,15 Camps are often located in suboptimal environments that would otherwise be sparsely inhabited and affect the provision of humanitarian aid. Camp creation leads to landscape changes to accommodate the population and built environment. Fuel demand and supplies 13,16,17 Availability of energy has impacts at the individual level, such as through the ability to cook or have electricity, and at the camp level through provision of health care and waste management. An insufficient energy supply often results in fuel poverty, increased deforestation, and malnutrition. 13 Deforestation 13 Deforestation changes the physical environment of the camps, including vector landscapes, often quite dramatically. It also contributes to climate change, and increased air pollution and irritants for chronic health issues. --- Uncontrolled fires The use of firewood has immediate dangers related to uncontrolled fires, which can spread rapidly in tightly packed camps. Climate change and extreme weather events [18][19][20] Temperature, climate, and extreme weather events can trigger displacement and affect individuals who have already been displaced. 18 These events can also alter the physical landscape, exacerbate the spread of infectious diseases, and affect the provision of humanitarian aid. Water and sewage systems 20,21 Unsafe drinking water and stagnant contaminated water can be sources of infectious diseases and long-term health impacts related to the consumption of naturally derived compounds. [22][23][24][25] Clean water can also affect the provision of health care. Groundwater quality 22,26,27 Insufficient sewage systems may introduce or reintroduce viruses and bacteria into groundwater and surface water, disrupting the natural ecosystem and contributing to disease spread. 22,27 Land clearing and burning, food production, and development or land hardening also release pollutants into these water sources. Livestock [28][29][30] Livestock serve as reservoirs for diseases and can attract disease vectors and multidrug-resistant bacteria, especially if practices are unregulated. Animal waste can also contaminate the water supply. Livelihood production 28 Means of livelihood, especially those that rely on water-intensive agricultural practices or livestock, can affect water resources. 28 Such means are also vulnerable to changes in temperature and climate. Watershed health and viability 26,31 Long-term viability of water sources depends on watersheds around camps, which are affected by the removal of native vegetation, changing hydrology, alteration of the microbial communities, and the addition or increase of pollutants in runoff. 31 Camp legal status 32 The legal status affects the provision and availability of services throughout the camp, as well as the camp's development and care. --- Political and conflict environment Political and/or conflict-related challenges cause variation in camp services over time, especially when camps are in areas of strategic importance. Discrimination, inequality, and violence [33][34][35] Differences in race, gender, legal status, and other demographic variables affect how an individual is treated upon entering and living within the camp. Women and girls are particularly vulnerable to gender-based violence, and lack an established justice system and access to appropriate physical and/or mental care. [33][34][35] Mobility 36,37 Individuals may have their movement restricted or may be completely unable to leave a camp legally, which limits access to health services and their ability to seek employment or education. Occupational opportunity 38 Displaced persons are often limited with regard to occupational opportunities as a result of legal and structural barriers to accessing work. Because of this, many are forced to work in environments where certain diseases are more common or they may work illegally. --- Host community relationships Host community relationships may be strained by actions such as deforestation, unsustainable land use, and employment outside of camps. These relationships can have marked effects on the longevity of the camp. --- Demography Demography has impacts across all sectors. Changes in demography can affect disease burden, required services, livelihood, physical landscapes, fuel demands, and other important components of the camp ecosystem. interacting ecologies will help us improve lives as other policies that provide a dignified existence out of the camps are prioritized. We have thus far focused on situations in which displaced persons aggregate in camps, though many are instead living in urban environments. These environments likewise tend to be unsafe and unhealthy. An ecological lens allows us to understand more fully the living conditions in areas such as these, though other factors are needed to address the needs of this specific population. 40 We recognize that this lens has its limitations. Individual camps create unique contexts and ecosystems, meaning that ecological interactions may vary greatly among camps. The adoption of this approach may have numerous logistic barriers, though these may be largely a result of the siloed approach this lens attempts to deconstruct. Developing a better understanding of how different components of this ecosystem interact, including at different space-/timescales, could lead to better planning and interventions to improve the health of both populations and environments in these difficult settings. --- RECOMMENDATIONS We need more partnerships that are cross-cutting between silos and that bring together humanitarian aid providers, camp planners, ecologists, public health professionals, health-care providers, environmental practitioners, lawyers, policymakers, and other experts of factors that contribute to the ecology of camps. Integrated policy needs to be developed based on assessing connections among systems so that evidencebased decisions can be made about how interventions may influence outcomes in multiple sectors. Sharing resources and taking an integrated approach will assist in buffering risk and creating more effective and proactive governance. We recommend research and collaboration that increases our understanding of interactions among camps, their environments, and broader ecosystems to incorporate this knowledge into future camps. This includes improving data gathering and analysis to incorporate local knowledge and partners with an emphasis on actively engaging women. 11 We recommend having the data ecosystem be considered more seriously and prioritized in these settings. Making decisions through an ecological lens does not need to preclude immediate aid provision. Through our recommended collaboration, we foresee a pathway to integrate ecological considerations into camp planning, development, management, and sustainment. However, we must first learn how we can adapt this lens in a way that is practical in real-time camp settings. As the first step in this process, we recommend a series of workshops that brings together different actors across silos to make concrete pathways for the inclusion of ecological considerations in camps. This piece is but the first contribution in what we anticipate, and indeed hope, will be much larger conversations around improving the care and services we provide displaced persons.

Health care among forcibly displaced persons is frequently driven by siloed approaches. Aspects of the built environment, social factors, and the bidirectional relationship between the changing ecosystem and residents are often ignored in health policy design and implementation. While recognizing factors that create a preference for siloed approaches and appreciating the work of humanitarian agencies, we argue for a new data-driven and holistic approach to understand the health of the forcibly displaced. It should be rooted in the realities of the emergence of new diseases, dynamic demographics, and degrading environments around the displaced communities. Such an approach envisions refugee and internally displaced camps as dynamic, complex ecosystems that alter, and are altered by, spatial and temporal factors. At the root of this approach is the necessity to work across disciplines, to think holistically, to go beyond treating single ailments, and to develop ethical approaches that provide dignity to those who are forcibly displaced.

Introduction Law enforcement agencies have adopted social media strategies for expanding their scope of communication and benefiting from the breadth and depth of its influence . Social media is a powerful tool that provides platforms for people to establish interact and also provide information about the individual's social life. Tang and Liu classify the different forms of social media as blogs , forum , media sharing , microblogging , social networking , social news , social bookmarking and wikis . Law enforcement agencies are increasingly using social media to interact with communities and also collect information required for criminal investigation. Social media is known for its four major strengths that include collaboration, empowerment, participation and time . Organisations and local governments have been increasingly using Facebook sites for official purpose to reach out to customers as they live online . Police departments acknowledge the potential of social media and are using it to enhance citizen input to police investigations, establish and strength its image, use it as a tool to control crowds and crisis and encourage input in policy-making processes . Mergel identifies four different social media strategies being adopted by the government agencies which include: push, pull, network and tactic strategies. Under the push strategy, the governments only broadcast information through the web content and there is no interaction. The pull strategy demonstrates the government attracts citizens to or users to provide information and usually limiting the level of interaction with them. The networking strategy exhibits high interaction whereby the government actively or passively participates in networking with the diverse constituencies. The final tactic strategy encompasses government services with actual transactions being implemented through social media applications. Thus, when government agencies choose to apply a social media strategy it also needs to anticipate citizen's social behavior and also be proactive in adopting technological trends. Sultanate of Oman has begun its smart government initiatives under the hegemony of the Information Technology Authority. 'Omanuna' has been introduced as the official Oman eGovernment Service Portal which maintains all the eParticipation platforms for government agencies. It manages 64 social media accounts of various government entities. In comparison to the Arab region and the world, Oman is ranked 2 nd and 43 rd respectively in the 2018 eParticipation index . The ranking was a significant improvement to the 2016 position where it stood at 33 rd position in comparison to the world reflecting that the government placed a special emphasis on the effective utilization of eParticipation channels and establishing contact with the public. Meijer and Thaens point out that the police departments are distinct from the other government operations. The police are in contact with all citizens and have diversity in their communication. The use of social media has been extensively investigated however very limited research exists that explores how police use social media and interact with the public . One of the noteworthy extant literature that explored the usage of Facebook and Twitter by seven police departments in the USA suggested that the contents of the information posted had an impact on how citizens interacted with them . Lieberman et al. investigated Facebook posts of 23 largest US police departments over three months in 2010. The findings suggested that departments with a high posting frequency shared more crime-related information whereas those with a low posting frequency shared more public relations information. Another research investigated Canadian police Twitter usage as a response to the 2011 Vancouver Stanley Cup riot . The Canadian police used media to garner citizen cooperation, gather evidence and mediate communication for community involvement. There is a dearth of research that explores the social media content from the police perspective. World over police departments are adopting social media strategies to connect with the community, thus there needs to be a systematic evaluation on the extent and purpose of why social media is being used by the police. To have a grasp on the potential of social media for the police department, this research specifically focuses on the Facebook page of Royal Oman Police. The central research question that this research investigates is: what type of information does the Royal Oman Police send through its Facebook and what is the extent of citizen engagement through the Facebook page. The article is structured into the following sections. Section 2 addresses and discusses the literature review related to e-participation related studies for the government entities and police department. Section 3 describes the research methodology adopted to investigate the research question. This is followed by section 4 that presents results and discussion of the study. Finally, section 5 is on the discussion, implications, and recommendations related to the results of the research. --- Literature Review As the world is moving to a more internet-based environment where people are becoming more connected, the law enforcement agencies in many of the countries have started using social media as a powerful platform for interacting with communities and collecting information for criminal investigations. Social media is used frequently by police departments to interact with the public and to get and provide information. Social media is helping the police departments to gather information for the effective deployment of resources and existing practices. The police agencies can get a feel of community sentiment about their policies, procedures, and practices. Good police-community relations are vital for developing trust between citizens and police. The use of various social media helps the police departments to build community ties which in turn improves the people's trust in police. A study was carried out by Dai et al. to systematically examine and analyze the varieties in the use of social media by traditional American local police departments and their interactions with citizens. The study collected data between October 1, 2013 and March 31, 2014 from Facebook and Twitter accounts of seven city police departments in the area of Virginia in the USA. Content analysis and text mining were used to identify the patterns of social media posts by police departments, and then statistically analyzed citizen interactions on social media with the police. The authors have applied a commonly used automated text analysis approach to analyze textual contents in Facebook and Twitter posts . The authors also observed that during the sixmonths of study there were a total of 1,293 posts and tweets on Facebook and Twitter for the seven agencies under observation and every agency had at least one post on Facebook or Twitter a day. The study used a number of Facebook likes and Twitter followers as indicators of community interaction with the police agencies. The findings of the study proved that there was a lot of variation in the way citizens interact with police on social media. It was observed that people on Facebook did more on networking as they visited the Facebook page of police agencies to like posts about the agency and police officers and provided some comments. However, people on Twitter tried to get access to information they like from the police Twitter sites such as general information required for everyday life, weather updates etc. The study finding had important implications for the police agencies to attract more citizen interaction for effective functioning. It was observed in the study that the contents of information play an important role in citizen engagement and interaction. The authors had recommended police agencies to adopt a problem-solving approach and analyze which category of information on their social media would likely to promote citizen interactions. The authors have suggested future research on how the citizen interchanges view with the police departments through social media will give more insight into the understanding of people's expectations and perceptions on policing. Recently, the use of social media by governmental agencies has been considerably increased. For example, Mossberger et al. state that 92.4 percent of local governments in USA has a Facebook page and 87 percent of the municipalities in the USA were active on Facebook by 2011. A study conducted by Bonson et al. in 2015 shows that all German local governments have a Facebook page . The study by Ellison and Hardy in 2014 shows that the majority of local governments in UK uses Facebook and Twitter to interact with people . Bertot et al. state that the use of social media by governments began as a way to improve transparency and openness and also highlighted collaboration, participation, empowerment and time as major potential strengths of social media . Bellstrom et al. conducted a study to identify the type of information exchange that happens between a local government and its citizens using social media. The research involved a qualitative single case study of the Facebook page of the Karlstad municipality in Sweden. The authors have used content analysis on Facebook data collected between May 2015 and July 2015 to find out different categories that determine the analysis. The frequency of content category proves that the municipality has used the Facebook page to promote different happenings in the municipality while the page users were asking questions to the municipality. The research involved analyzes of the type of information disseminated through the Facebook page of local government, the type of information received through the Facebook page of local government and the relation between the type of information communicated on Facebook page and the engagement between the government and citizen. The authors have suggested further study on motivational factors of posting comments and community engagement. Al-Aufi et al. examined the perceptions of Omani citizens towards the use of the government's social media for participatory and interactive relationships. Mainly the study investigated and explained the views of social media users on the transparency, engagement, responsiveness, and trust about the use of social media by the Omani Government. The study found that even though the citizens are increasingly accessing the governmental social media, governmental departments are not utilizing social media effectively and failed to engage citizens to design and deliver more collaborative and efficient services. The major emphasis of the study was to strategically frame the use of participatory social media by the governmental departments for interactive governance. According to Warren et al. , the governmental departments have been obliged to listen and engage with citizens as the citizens are empowered by the use of social media. The study states that governments are now expected to regulate and professionalize their engagement with citizens through social media platforms. Gohar Feroz et al. explored Twitter usage by Korea's central government by classifying Twitter-based networking strategies into G2C and G2G which are Government to Citizen and Government to Government respectively [10]. The authors investigated the nature of networking and social media interaction strategies of the Korean government by analyzing the tweets of 32 ministries. The results found that Korean government institutions have made extensive use of Twitter in their daily interactions with the public, but their networking strategies did not necessarily motivate the public to participate in their social media activities. The research findings indicated that the ministries focused more on addressing the needs of the citizens rather than an increase in the number of followers. Police departments in many countries have started using social media sites widely as the main method of communication with people. Liberman et al. performed a study on Facebook messages posed by 23 police departments in USA over a three month period during 2010. The researchers used content analysis and the result finding indicates that the content pattern was related to the frequency of department's Facebook postings to some extent. Crime-related messages were posted by the police departments who have used Facebook more frequently and the departments who have used Facebook less frequently have posted mostly public relations messages. The researchers also analysed the policy implications of these trends in posting messages. A study conducted by Sadulski states that Twitter and Facebook are frequently used by police departments to inform the public about crime incidents. The study also informs that police department and other law enforcement agencies can improve community trust through transparency in communication with citizens. The police departments use Facebook and or Twitter for informing the public about progress in case investigations. According to the Law Enforcement Use of Social Media Survey, Law Enforcement Agencies use social media for a wide variety of purposes . The survey results show that 91 percent of enforcement agencies use social media for notifying the people about safety concerns, 89 percent use social media for community outreach and engagement with citizens and 86 percent use social media for public relations and managing reputation. Unlike the developed countries, those have utilized the advantage of social media in increasing the trust and transparency of governments, the governments in Arab countries use social media mainly for news broadcasting and updating information . As per the reported information from the Omani National Center for Statistics and Information , almost 50 percent of the public uses social media regularly in Oman. The report also states that people with Higher Educational Qualifications use social media more frequently than people with lower education levels. According to the report, even though the Information Technology Authority , Oman has been promoting e-government and e-services through "Digital Oman Strategy", the use of social media by Omani Governmental departments remains unregulated. --- Research Methodology This study uses the case study approach that involves content analysis of the Facebook usage of the Royal Oman Police , Sultanate of Oman. This approach is suitable as it explores the understanding of a complex and contemporary social phenomena. --- Data Collection and Coding This study uses the manual data collection of Facebook content. A web content analysis is used to explore the Facebook posts for the period 1 st January 2019 to 11 th June 2019. This period is chosen to investigate the communication by the ROP during the year that includes the month of Ramadan. During the holy month of Ramadan, Muslims abstain from eating and drinking from dusk till dawn. Sultanate of Oman is one of the Arab countries that has mandated reduced working hours for certain categories of employees . The period of study makes it unique as the activities are different for the citizens, residents, and tourists in comparison to other times of the year. The steps observed for investigating the posts are illustrated as follows: 1. Identification of the official Facebook page of ROP; 2. The capture of the Facebook posts in the period 1 st January to 11 th June 2019; 3. Codes being assigned to the posts based on the content and media type 4. Exploration of the engagement rate of the posts 5. Analysis of the post types, purpose and engagement rate In line with the prior literature, Netvizz service is being used to collect data from the Facebook page . Netvizz is a Facebook application that can archive the page activity by delimiting it based on the number of posts or a time interval. This research uses the latter approach to cover the specified timeframe as identified above. The page consisted of 43 posts by the ROP, 37 comments and 738 reactions. The data was collected on 13 th June 2019. The research followed the approach advised by Miles and Huberman to use a 'first list' of codes generated by the extant literature . The first list of codes is used before reading the actual data and is modified and expanded based on content analysis. The coding took place using seventeen classifications that included: crime information, law enforcement, asking for tips, crime prevention, safety advice, community interaction, community events, traffic updates, agency-related information, recruitment, other government services, information sharing, holiday greetings, non-police related, advertising and road conditions. The data coding and analysis were performed using three steps. Step 1 involved coding by the first author using the first list of codes and any subsequent inductive modification. Step 2 involved coding by a second co-author and also two graduate students not affiliated with the study to ensure the reliability of the coding process. Step 3 as the final analysis led to clarification of category names. A Cohen's Kappa inter-rater reliability score of 80 percent was found for the co-author's results for the categorization. This reflects a substantial agreement between the two coders and is also consistent with the findings of Williams et al. . The framework of Bonsón and Ratkai was adopted to investigate the stakeholder engagement on the ROP's Facebook page. The framework is evaluated using three metrics that reflect popularity , commitment and virality . All three metrics are aggregated to calculate the overall engagement score. The metric components are discussed in Table 1. The findings of ROP Facebook posts collected from 1 st January to 11 th June 2019 are presented in Table 2. The Facebook page has a total of 29,558 followers as of 11 th June 2019. A total of 185 posts are found in this period and based on content analysis these were classified various categories. A total of eighteen categories were identified to capture the information posted on the Facebook page. The highest number of posts were made by the ROP under the safety advice category that accounts for 39 percent. This suggested that ROP is using Facebook with a community orientation tactic and disseminate related information. This is an interesting finding in comparison to the study by Williams et al. that suggested its two most common categories were traffic and accidents. Table 2 highlights that the second most important category of posts was related to ROP awareness sessions for schools which accounted for 14 percent. Also, a cumulative of 28 percent is assigned to all awareness session categories that reflect that ROP is continuously involved with various sections of the community to educate them. --- Table 1: Stakeholder engagement metrics for Facebook The posts also actively gave information on law enforcement and agency-related information. On average there were at least 34 posts per month which almost accounts for one post per day. Consistent with the findings of prior studies, the ROP tended to have no postings on information that was not related to the police unless it was to do with safety advice which was driven on religious account . The safety advice posts were highest during the month of Ramadan. The post included a verse from the Quran that advocated safe driving practices. This strategy by the police was being adopted to reduce the number of accidents in the month of Ramadan as fatigue is attributed during fasting and it was one of the main reasons for fatal accidents. The safety advice posts also advocated child safety belts, car maintenance checks, tips on driving at night, avoiding phones while driving and tips on driving during rainfall. The school, university and other institution awareness posts communicated the awareness sessions conducted by the ROP to propagate traffic safety. Law enforcement category posts included information on the penalties imposed for wrongly overtaking, rash driving, speed limits, intentionally risking lives by crossing flooded valleys and allowable vehicle load. The agency-related information included information on their work timings, holidays, participation in events and meetings. The ROP posts also covered diversions on the roads and traffic updates in various other categories however at a lesser frequency. Crime information category did not have many posts considering that it is one of the first job priority of the ROP. The avoidance of posts on crime-related information may be due to the sensitivity associated with it and a strategy to avoid creating a bad image. It might also be a source of unrest and the possibility of creating fake news in the future. This is consistent with the approach followed by the police in the USA on their social media as reported in by Williams et al. . --- Citizens Interaction on ROP Facebook Social media is a ubiquitous tool that allows its users to interact with one another in the virtual world. The ROPs Facebook page is also a tool for the agency to allow community interactions related to their posts. A follower gets the opportunity to 'like' the posted information, 'share' with other users or 'comment' on the posted information. Table 3 presents the citizen interaction on the ROP Facebook for the period 1st Jan to 11th June 2019. The likes, shares and comments were classified for each of the eighteen categories. There were a total of 2506 likes, 136 shares and 127 comments . The safety advice posts received the highest number of likes. This is also reflective of the fact that the highest number of posts were for safety advice. The law enforcement category received the next highest record of likes. This was followed by school awareness, agency related information and other agency-related awareness information. The findings of this study are partially consistent with Dai et al. which also reported the highest number of likes for agency-related information and law enforcement categories. Table 3 reflects the highest number of shares were again for the safety advice followed by the law enforcement categories. The remaining shares were negligible for the posts. The posts also garnered comments which were in the order of safety advice, law enforcement and crime information. In comparison to the number of followers, the citizen engagement has not been too strong with the posts on the ROPs Facebook page. The overall engagement levels of the citizens with the ROPs Facebook post is represented in Table 4. The average number of likes per post per 1000 fans score is 0.46 which reflects on weak popularity. The average number of comments per post per 1000 fans is 0.02 considering that very few comments were made in the period of study whereas the average number of shares per post per 1000 fans was 0.02. The overall engagement level was only at 0.51. This indicator is useful to motivate the ROP in encouraging more citizen engagement. --- Types of Facebook posts Table 5 presents the statistics on the type of posts found on ROP Facebook page. Photos were used extensively for the posts and it accounted for a total of 96 percent of the 185 posts found over the investigated period. There were only about six videos and two links posted on Facebook. --- Conclusion and Recommendations This research contributes to the scarce literature on how police are using social media strategies and interacting with the public. This study uses content analysis of ROP's Facebook page. The findings suggest that there are 18 categories under which Facebook posts can be classified. ROP is active in posting information daily by using more photos than links or videos which is consistent with the recommendation offered by He et al . The content analysis reflects that the posts were heavy titled towards the well-being of the community, therefore, there was a high percentage of posts that advocated safety. The ROP has also been active in community awareness programmes that catered to the schools, universities and other organisations. Citizens engagement in terms of was limited to the posts however in comparison to its followers the engagement is considered weak. Currently, the ROP Facebook posts can be termed as a push strategy as described by Mergel . The practical implications of this research suggest that ROP can do better in improving its interaction with the community on social media. By reaching out to more people, the messages on public safety can be more effective. In line with the recommendation by Dai et al. , the ROP needs to adopt a more problem-solving approach to increase community engagement. This study has certain limitations. The research is restricted for a few months and can be extended to cover a longer period. The investigation is limited to the Facebook account only and needs to also consider other social media such as ROPs Twitter account. The use of social media by the ROP and other government agency are set to promote Oman's commitment to building smart cities and thus it requires that strategies are set to encourage e-participation by the community.

Royal Oman Police (ROP) began its virtual engagement with the citizens by enforcing its presence in social media such as Facebook and Twitter. The purpose of this paper is to explore the type of information shared through Facebook by the Royal Oman Police and analyse the extent of citizen engagement through the ROPs Facebook page. This study collected data from 1 st January 2019 to 11 th June 2019 from the ROPs Facebook account and supported through content analysis and statistical tests. The findings reflect that ROP had posts under 18 categories and the highest number of posts were on safety advice for the citizens, followed by the awareness sessions organized for schools, universities and other institutions. The citizens interacted virtually on the Facebook page by demonstrating likes, shares, and comments for the posts. This study contributes to providing detailed insights on the type of information exchanged between ROP and citizens. The results of this study can be used for developing strategies in increasing citizen engagement as part of Oman's development plan of building smart cities.

Introduction The Women's March in January 2018 was a worldwide protest to advocate legislation and policies regarding human rights and other issues, including women's rights, immigration reform, health-care reform, racial equality, freedom of religion, and workers' rights. Most of the rallies were aimed at Donald Trump, largely due to statements he had made and positions that he had taken which were regarded as racist, anti-women, or otherwise offensive. To vent this anger, many protest posters featured offensive jokes at the expense of Trump's body, mocking his "comb over" hairstyle, his small hands, his orange taint, and so on. Such posters were often spotted at protests and shared widely online, much to the amusement of the movement's supporters. While some people suggest that such charged political online humor can mobilize people and serve as "a pre-political gateway to future civic engagement" , there is also concern that it remains inefficient or even antithetical to meaningful sociopolitical change . Thus, in the context of social media, offensive political humor advances the so-called echo chambers where people only speak to like-minded individuals . Others argue that routine online searches for pleasure and entertainment "entrap[s] us within the circuits of neoliberal communicative capitalism-a process that continuously replaces political action with political feeling, forever turning activity into passivity" . By drawing on the literature that explores and adapts Bakhtin's concept of the carnivalesque , this article argues for a more nuanced understanding of offensive political humor as a flexible affective resource that is complexly intertwined with social change . Based on our analysis of 400 social media posts from Instagram and Twitter, we argue that the online circulation of humorous protest posters creates forms of "polysemic undertow" that both contest and confirm normative assumptions about white masculinity and the political public sphere. For this reason, the meanings of these protest posters are not so coherent as to reflect either transgression or backlash politics exclusively. Rather, the contradictory nature of offensive humor holds these circulating online images in tension, thereby enabling what Reilly and Boler call "prepolitization"-a novel form of civic participation that can mobilize citizens who would not otherwise explicitly participate in civic life, thereby creating new political sensibilities and desires. We argue that the Women's March provides unique insight into how offensive humor can function as a mobilizing force, without glossing over its limitations in the realm of civic engagement. Offending those in power does not replace rational political debate. Nonetheless, it can be an effective tool for drawing attention to situations of injustice, for binding people together against formal power structures of authority, and for carving out a space for empowering feelings of countercontrol, which are necessary ingredients for social and political change . In this sense, this article contributes to the work of contemporary scholars of social movements and media who rethink traditional understandings of politics and participatory democracy. --- Literature Review: Offense, Online Humor, and Mediated Protest The cheerful vulgarity of the powerless is used as a weapon against the pretence and hypocrisy of the powerful. Giving and taking offense on social networking sites is a contested topic. While some celebrate the interactive architecture of social media as a democratizing and diversifying force, others warn that these seemingly antihierarchical affordances invite offensive behavior, such as cyberbullying or the production and circulation of offensive material that more traditional media outlets would have censored or regulated. Thus, social media and other user content hosting companies are increasingly under the ethical and legal responsibility to make their network a "positive" and "safe" space where offense is avoided. Offense, however, is an affectively charged, slippery subject that escapes clear definitions. Although offensive material is, in principle, distinguished from that which is illegal , it remains difficult to define the boundaries in a robust and consensual fashion. Generally, media content is judged to be offensive when it is too graphic or explicit in style and content . Intrusive images of suffering, or racist, classist, or sexist depictions that contribute to stereotyping, or bias and inaccuracy in the media are often reported as offending audiences . In public discussions, "offensive" media content is often equated with "harmful" content. This equation is based on simplistic theories of media effects that conceive offense as a monolithic "bad" thing that can be pinned to certain media representations and eliminated through censorship. Such understandings fail to see the contextual, relational nature of offense as well as the emotional messiness of offense. Offense is far from a monolithic, clear-cut emotion but contains a wide range of contradictory feelings and emotions, such as pain, anger and frustration, alongside joy and titillation . Furthermore, these approaches overlook the potential for the so-called negative emotions to push us into new critical directions, as it has long been theorized by feminist scholars such as Audre Lorde , Sara Ahmed , andSianne Ngai . Taking offense and "getting angry" is here often conceived as an affective mobilizing force for social and political transformation. Offensive joking, in particular, has been theorized as offering temporary relief from oppressive social norms and conventions . Pickering and Littlewood argue that what remains crucial in this context is whether the humor kicks socially upwards or downwards; whether comic aggression is directed "at those who are in positions of power and authority, or at those who are relatively powerless and subordinated" . Such an understanding implies that offense is not in and of itself wrong and that, depending on its direction, it can have a positive or negative impact. The affirmative and liberating possibilities of grotesque, offensive humorous transgressions are often associated with "the carnivalesque": A general mood of liberation, mocking of hierarchies, and temporary suspension of rules . For Mikhail Bakhtin , the popular tradition of carnival has the potential to suspend social hierarchies through mostly bodily and bawdy humor, which finds expression in the celebration of bodily grotesqueness and excessiveness, fooling around, and profanities. These markers of indecorum are strictly policed during "normal" times, but during carnival, they can be animated and enable comic reversals. For instance, a jester might be crowned in place of a king, and, as a result, the authoritative voice of the dominant discourse momentarily loses its privilege. Bakhtinian carnival theory has been criticized for its neglect of carnival violence against women and Jews, its failure to consider social relations of gender, and its failure to deal with the consistence of dominant culture . We nevertheless see Bakhtin's concept as a valuable starting point and draw on the productive ways in which it has been extended through the work of Stallybrass and White and Rowe . First, Stallybrass and White argue that the carnivalesque should be situated within a wider pattern of transgression, in order to "move beyond Bakhtin's problematic folkloric approach to a political anthropology of binary extremism in class society" . They maintain that this broader focus on "transgressive symbolic domains" enables us not only to examine cultural hierarchies and binary social structures that underlie the carnival but also to "operate far beyond the strict confines of popular festivities." Second, Kathleen Rowe ) builds on Mary Russo's work on the female grotesque to adapt Bakhtin's concept for thinking about female unruliness. As Rowe argues, the transgressive figure of the "unruly woman" can help "sanction political disobedience" but is also associated with dirt and pollution . She threatens "the conceptual categories which organize our lives," and this liminality evokes intense, contradictory feelings: "Her ambivalence, which is the source of her oppositional power, is usually contained within the licence accorded to the comic and the carnivalesque. But not always." Our study, then, draws on these two key extensions of Bakhtin's work to examine how transgression spills over the confines of the temporary, local contexts of the Women's March through the online circulation of offensive protest humor. Here, the carnivalesque functions as a malleable resource that can provide spaces for disruption and rebellion, without glossing over cultural differences. However, while some scholars in critical humor studies have argued that offensive humor can operate as a powerful social corrective as well as a strategic and effective commentator on political issues , others highlight that its uniting-and-dividing function draws a sharp boundary between those who laugh and those who are not "in on the joke" . From such a perspective, bawdy political humor that predominantly works by deriding and offending those in power is merely: further convincing those who agree with it while alienating those who don't agree. Thus, the satirical mission to "make laugh, not war" only serves to polemicise the gap between those who agree and disagree with its political message, suggesting its transformative worth is limited. This so-called echo chamber phenomenon is often discussed in the context of social media. Critics argue that, rather than enabling debate and deliberative compromise essential for creating political change, our social media practices of "posting," "liking" and "sharing," along with algorithms, generate filter bubbles and echo chambers with restrictive partisan sentiments, where only likeminded people speak to each other . Nevertheless, Bivens and Cole maintain that "the prevalence of social media use, like Instagram, Facebook, and Twitter, provides a method through which individuals can push back against the legislative structures in the United States." They illustrate in their work on "grotesque protest" that social media provides individuals with opportunities to resist attempts to control bodies and to reinsert individuals' voices in political discourse that is aimed to exclude those bodies . In a similar vein, Tufekci and Wilson found that social media use greatly increased the odds of being involved in a protest, and that it "represent[s] crosscutting networking mechanisms in a protest ecology" . Thus, although commonly understood as like minds speaking to like minds, social media can also be seen to diversify protest networks and encourage debate. --- Methodology This article builds on our previous study , which examined the social media circulation of images from the 2017 Women's March. One of the key themes we identified was the prevalence of images featuring placards that mocked Trump's body. We want to explore this tendency further within the context of the 2018 Women's March, to consider how offensive humor might function as an affective protest strategy. We collected our sample by using the #WomensMarch2018 and #WomensMarchNYC hashtags to search for public posts of images that were shared on Instagram and Twitter between January 20 and 21, 2018. We chose these two platforms because they are associated with different affordances and cultures. Twitter is reportedly used by 24% of the U.S. adults , and, although it facilitates the sharing of imagery, it is primarily associated with text content and has often been used for political communication and activism . Instagram is reportedly used by 35% of the U.S. adults . This platform foregrounds imagery and is often considered a feminized online space that is preoccupied with celebrity, beauty, and style . We are interested in images of protest signs as a form of visual and affective political communication, and about how protesters and social media users can grab our attention and encourage circulation through the use of offensive humor and spectacle. Our data collection followed a three-step process. On the day of the march, we followed the #WomensMarch2018 and #WomensMarchNYC hashtags on Twitter and Instagram and observed recurring images of individuals and groups of protesters holding protest signs, many of which were designed to offend Donald Trump through bawdy and bodily humor. This trend confirmed that offensive humor was once again a prevalent protest strategy. On January 22, we then used the platform tools to collect the 200 "top" posts from each of the two platforms for thematic analysis. We identified three recurring themes: The ridicule of Trump's body, the association of Trump with excrement, and name-calling and violence targeting Trump. Finally, we selected one illustrative post from each of these three themes for close analysis. We include screengrabs of the images here but have removed social media usernames and profile pictures. The three-case study images were all widely shared on social media. This approach facilitates reflection on how the reiteration and circulation of images "invite polysemic undertow" that can unsettle Trump's intended personae as serious public official and thereby animate political engagement and social change. Having outlined our theoretical framework and our methodological approach, we will now move on to our three-part analysis. We begin by exploring the tendency to mock Trump's skin color. --- Orange Skin, White Masculinity, and Carnivalesque Countercontrol Trump's body is often the target of ridicule. His "orange" skin tone has inspired large numbers of Internet memes where the president is mocked as "Agent Orange" or "Cheeto Trump." Equally popular targets are his supposedly "tiny" hands. Merchandise includes t-shirts that read "Keep your tiny hands off my rights" and coffee cups with extra small handles, just two of the many physical and digital artifacts through which Trump's opponents publicly ridicule his masculinity. Here, we focus on the recurring degradation of Trump as failed White masculinity and use Bakhtin's concept of the carnivalesque to consider the ambiguous workings of this offensive humor in political protest. The above poster draws attention to the Deferred Action for Childhood Arrivals act that the Trump administration has tried to rescind since September 2017. DACA, an Obama-era protection scheme, allows those who entered the United States illegally as children to receive a renewable, 2year period of deferred action from deportation and to be eligible for a work permit. The scheme is now closed to new entrants and puts 800,000 registered recipients in danger of deportation. Trump's attack on DACA offended not only many DREAMERS but also protesters at the Women's March. Thus, numerous protest posters in the 2018 March focused on DACA and the above placard is one such example. Under the pink headline "DACA DACA TINY COCK-A," we see a cartoonlike drawing of Trump. He is naked, showing off his "orange" skin, and wearing only a blue jacket and a red tie. His signature comb over hair-do is exaggerated and his arms are wide open. The lines around his small hands make it look as if he is "flashing" the onlooker, showing off his small penis, or his "tiny cock-a," as the poster reads. The poster criticizes the imminent changes in DACA policy and aims to provoke laughter by offending and shaming Trump's body through the use of "carnivalesque" humor. Drawing on Bakhtin's work, we employ the concept of the carnivalesque to think about vulgar, grotesque, bodily humor that is commonly intended or experienced as offensive, and that is used to challenge privileged positions and reframe public and political discourse. The rhyme "DACA DACA TINY COCK-A" mimics the ways in which small children try to offend each other in the playground. It relies on the shared understanding that there is a comic incongruity between our expectations for "presidential" behavior and the "childish" and unconsidered ways in which Trump presents himself publicly and politically. Trump's child-like behaviour violates dominant assumptions about the rational, male agent in the political public sphere, and it can be argued that it is exactly these kinds of transgressions that Trump's opponents experience as offensive, and which in turn mobilize them to protest and give offense back. The poster also makes reference to the running joke of Trump's "tiny hands," and the popular myth that a man's hand size is indicative of his penis size. Small hands here suggest a lack of masculinity and a lack of gendered attributes, such as strength and leadership. The link between hands and gender performance is underlined by Janice Winship's influential work about the relationship between the positioning of hands and sexuality in advertising. According to her analysis, male and female hands are part of an entire message system of representation signifying appropriate gender behavior. In other words, hands allow us to tap into familiar ideologies of masculinity and femininity because the big and strong hand of the "leader" is "naturally" the hand of a man, whereas the small and delicate hand of the homemaker and caretaker is "naturally" the hand of a woman. By repeatedly mocking Trump for his "small" hands, opponents do not only offend his masculinity on a personal level, but they undermine his presidency by insinuating that he is not a "real" man, he is not a "leader" and therefore not someone we should fear, trust, or follow. Furthermore, the poster constructs the naked, overly tanned Trump as the butt of the joke because orange skin is commonly perceived as a funny tanning "accident" rather than a desired skin hue. As Graefer argues, "orange" skin invites ridicule and offense giving as it symbolizes excessiveness, lack of taste, and the pollution of "proper" whiteness. Regarded as "ugly" and "tasteless," this skin tone stands in stark contrast to the White hue that the proper White, middleclass subject should embody. The White, middle-class subject is controlled and rational in its desire to darken its skin, making tanning in this case an acceptable and positive habit. Orange skin, on the other hand, is taken as visible evidence of a subject's inner out-of-controlness and illustrates that Trump does not have the supposedly innate cultural tastes and decorum that wealthy White people should have. His highly visible over-and misuse of tanning products also marks him as vain and overly concerned with his appearance, characteristics that are commonly associated with femininity rather than masculinity. The DACA DACA protest poster then uses offensive, bodily humor to produce Trump as a figure of ridicule, but this kind of humor is riddled with both transgressive and conservative tendencies: One could, for instance, argue that offensive humor works here to undermine the powerful White man via emasculation. Yet it ironically also works to restore dominant assumptions of an idealized White masculinity that is free from feminine traces, such as tiny hands or vanity, and immaculately White, rather than orange. Furthermore, the vulgar ridiculing of Trump's body can also be interpreted as conservative because women have historically been silenced and policed through these same mechanisms of body shaming. However, we should not reject offensive humor as a tool in mediated social protest altogether. Rather, we suggest that this online sharing of offensive humor aimed at the powerful can be seen as a contemporary expression of Bakhtin's carnivalesque. Despite the fact that online practices in the context of Instagram and Twitter defy the circumscribed spatial and temporal specificity of "carnival," it still provides a useful tool for understanding the transformative potential of offensive, vulgar humor because it illustrates how the transgression of social boundaries can be a productive act of resistance. This potential is grounded in the collective experience of transgression: Carnival is not a spectacle seen by people; they live in it, and everyone participates because its very idea embraces all the people . . . It has a universal spirit; it is a special condition of the entire world, of the world's revival and renewal, in which all take part. Notwithstanding the inherent problems of Bakhtin's celebratory universalism, it can be argued that posting offensive placards against Trump generates new forms of collectivity because it serves as a public act of stance taking , where people align with others through the stances they take towards a particular idea, object, or person. Offending Trump through humor works, then, not only to vent the protesters' anger and frustration but to create a sense of superiority and belonging, through the affective experience of shared laughter: Laughing at faulty behaviour [and bodies] can also reinforce unity among group members, as a feeling of superiority over those being ridiculed can coexist with a feeling of belonging. These acts of online offense giving, then, are performances designed to appeal to like-minded others, thereby aligning bodies with antiracist counterpolitics and drawing boundaries between "us" and "them." Some of the comments below the online image highlight this uniting power of offensive humor: Expressions such as "Love it" or the powerful arm emoji illustrate that sharing "great signs" beyond the marches enables new forms of collectivity, temporary zones in which feminists are able to take a stance and make their anger visible while enjoying themselves in the process. The glee and pleasure that users experienced when engaging with these offensive online images can be seen as producing carnivalesque moments of countercontrol where activists no longer feel helpless in the face of patriarchy and racism, but where they feel powerful and impactful. Our premise, then, is that offensive humor, as communicated through these images, is affective and, as such, drives online exchanges and attaches people to particular platforms, threads, or groups. A direct, tangible and measurable "effect" of activism might not be easy to locate, yet it would be wrong to ignore results like the production of feeling, which, we argue, is necessary for social change. --- Filth, Cultural Transgression, and Immigrant Bodies We now turn to the second recurring theme in the circulated images from the 2018 Women's March, which was the degradation of Trump through the semiotic resource of filth. Combining the concept of the carnivalesque with Stallybrass and White's notion of cultural transgression, we examine a sign shared on Twitter and consider how it responded to Trump's offensive behavior by shifting the shame and otherness he inscribes on immigrant bodies onto the president himself. This sign illustrates the recurring association between Trump and feces. The image was posted by a private Twitter user, who photographed and shared his "favorite" signs from the Women's March in New York City. At the time of our data collection, the tweet had been shared 625 times, favorited 1.7K times and had received 70 comments. The sign depicts Trump's face as a bottom that emits a brown puddle. Across his yellow hair, the text reads "F*ING MORON," while text within the brown puddle reads "LIAR." The discursive association between Trump and feces work in two key ways here. First, the sign is a critique of Trump's use of the term "shithole countries" to refer to the nations of origin of immigrants he considered undesirable . Second, the sign uses comic inversion and grotesque imagery to construct Trump as abject. We explore how these strategies work together and reflect on how they invite onlookers to feel both offense and pleasure. Trump's "shithole" remarks were made in a meeting with the U.S. senators on the January 11, 2018, and received extensive international media coverage. The president of Senegal said he was "shocked," the government of Botswana said the remarks were "irresponsible, reprehensible and racist," while an African group of ambassadors at the United Nations described them as "outrageous, racist and xenophobic" . The remarks, then, were widely constructed as offensive. The protest sign shifts the "shithole" label from these nations onto Trump himself, repositioning the offender as the target of offense. Here, Trump becomes the "shithole," reduced to an abject body part and dismissed as a "moron." As in the grotesque imagery described by Bakhtin , we see a decentered body that is ruptured by bulges and orifices. The close-up image fills the entire sign. The buttocks are comically round and disproportionately large. The anus protrudes and leaks filth. The vulgarity of the picture is echoed by the crudeness of the written language: Trump's debased body contaminates the world with its "shit." Such rhetorical strategies work "to mock, destabilize, and publicize private parts and activities we are socialized to hide" . Situated within protest culture, the sign employed this carnivalesque language and imagery to contribute an affective critique of Trump to public sphere debates around his presidency. By shifting the "shithole" label from developing nations onto Trump himself, it reverted the cultural hierarchy of Trump's racist immigration policy, simultaneously articulating offense at his racism, and giving back offense by degrading and insulting the president. The reach of the sign was extended beyond the moment and geographical context of the march through the circulation of the image on social media. The Twitter user who shared it used the platform's comment function to share a number of other photographs of protest signs, while other users responded by expressing excitement, laughter, calling for Trump to be impeached, and sharing photographs of other signs that resonated with them: We can locate these photographed protest signs within a wider symbolic practice that degrades Trump by associating him with the lower stratum of the body . Trump's buttocks and feces were recurring themes in circulating images from the 2018 Women's March, while other social media users have also adopted the hashtags #PEEOTUS and #SCROTUS to avoid mentioning Trump's name and title, thereby denying the legitimacy of his presidency . This refusal indicates that he is seen as a transgressive figure. He is both president and other, both insider and outsider. Drawing on Stallybrass and White , we argue that the widespread use of grotesque representations signals a dual sense of disgust and fascination with Trump. His offensive behavior represents base impulses that should have been repressed from the rational public sphere, a notion that is also evident in Hillary Clinton's labeling of some Trump supporters as "deplorables" with "racist, sexist, homophobic, xenophobic, Islamaphobic" views . Trump-as-president is a hybrid of high and low discourse, a transgression of established cultural boundaries that creates a 'powerful symbolic dissonance' . The recurring use of filth as a semiotic resource in protest signs can be seen as an attempt to reaffirm the classification of him as other, suggesting the coexistence of a desire "to degrade the high and mighty" and "a paradoxical reverence for tradition and hierarchy" . Through this mingling of transgressive and conservative tendencies, the target of laughter and offense is not the presidency as institution, but Trump as an illegitimate president. Challenging the universalism that undermines Bakhtin's work on the carnival, Stallybrass and White were interested in examining the cultural transgressions of class binaries. Our study, in turn, underscores that class structures inter-sect with those of race, nationality, and gender. In each case, "discourses about the body have a privileged role, for transcoding between different levels and sectors of social and psychic reality are effected through the intensifying grid of the body" . Trump attempts to classify, legislate, and control the bodies of immigrants by associating them with excrement, while protesters use the same semiotic resource to degrade his body and delegitimize his presidency. As Stallybrass and White note, "somatic symbols...are ultimate elements of social classification itself" . Thorogood suggests that crude, ambiguous humor can help fight disengagement from formal politics. By "reducing politics to the excretions of the human body," protesters negotiate and challenge geopolitical discourse by connecting contemporary debates to our shared bodily vulnerability. As Grant-Smith writes, "we all shit," and so defecation can be mobilized to dem-onstrate our "common humanity and animality" . However, we would argue that this sign does not work to position Trump within a human collective but instead positions him as a repulsive other. Its strategy resembles that of the "DACA DACA" sign, as it employs the playground rhetoric of "no, you're the shithole." However, drawing on Bivens and Cole, we nonetheless argue that "grotesque protest" can still work as "an effective tool for opening space, transgressing boundaries, and demanding attention." The excess of protest signs like this one invites us to take gleeful pleasure in offending Trump, reminding him that he is out of place and illegitimate, and demonstrating resistance to his attempts "to control bodies." In this way, the grotesque is mobilized "to strategically reframe public and political discourse about the body" through street protests and on social media . Across the protest signs focusing on "orange" Trump and Trump-as-filth, his White, male, heterosexual, and wealthy body is subjected to degrading strategies that have long been used to oppress women and minorities. It here becomes violently appropriated as a site of resistance, used to articulate feelings of offense but also to cause offense. This strategy valorizes anger as a political emotion and invites us to take pleasure in voicing that anger without concern for the decorum imposed on women by patriarchal discourse. The last part of our analysis will focus in on this relationship between gender, anger, and offensive humor. --- The Retaliation of the Unruly In this final theme of circulated protest signs from the 2018 Women's March, we examine how female protestors used the characteristics of unruliness, such as offense , to appropriate Trump as the target of carnivalesque humor. To illustrate this trend, we conduct a close analysis of the sign saying "Little bitch, you can't fuck with me," which demonstrates the ambiguity that was evident in some of the offensive Trump placards: This sign mixes humor, politics, and popular culture to promote the embrace of unruly feminism and to protest an epidemic culture of sexual harassment and the policing of women's bodies. The sign features a quote from music artist Cardi B's song "Bodak Yellow"; a "diss" track that skewers those who have mocked B's rise to fame from Bronx stripper to music history maker. In this track, B uses the pejorative term "little bitch" to lambast those who criticize her achievements, but in the context of the Women's March, the term is used to insult Trump's behavior and body. According to The Urban Dictionary , "little bitch" is a whiny, petty person, willing to stab people in the back. Thus, the sign's reappropriation of the popular term connects these associations with Trump's actions and behavior, such as his claim that "no politician in history has been treated worse" than him, his childish exchanges with North Korean leader King Jong Un, and the Republican Party's cutting of Medicaid, which Trump's working-class supporters are reliant on . The gendered nature of the term "bitch" further attacks Trump's "inadequate" masculinity, emphasized by the word "little" and its connection to the long running joke about his small hands and penis. The sign combines this insult with a threat through the phrase "you can't fuck with me" and the image of Trump's head pinned down by a pair of Black, high-heeled feet. Together, they act as a warning that the Women's March activists are not to be "messed with." This threat is supported by Cardi B's own clarification of her lyrics: "I can be humble but...if you push me, I can really stamp on your head" . While the song incites violent behavior, its meaning within a humorous placard is much more ambiguous. As Lockyer and Pickering argue, the line between make-believe and reality is not clear in a joking context. This is because humor can be a form of exaggeration, but it can also be used to express real beliefs . Thus, while some may view the sign's message as threatening and offensive, others may defend it as "just a joke" or benign violation humor that appears immoral but is essentially harmless because the words are "just" borrowed from a song and are not the actual words of the protestors. This image garnered many affective responses because it was posted by Cardi B and circulated to her 17 million Instagram followers. B's cultural and symbolic capital afforded maximum exposure to the protest sign and this particular Women's March message. According to D. S. Meyer , one of the key benefits of celebrity-endorsed protest is "the increased mobilization of sup-port and publicity". Similarly, scholars have argued that music has the capacity to mobilize political action and collective identity . Such ideas are applicable to Cardi B's Instagram post, as it acquired over 1 million likes and nearly 13,000 comments. The comments were varied; some expressed support for Trump, arguing that he was "creating a fuck ton of jobs," while some expressed amusement through the crying laughing emoji. However, the dominant response expressed support for the sign's message and B's accompanying comment about the disrespectful treatment of women: "Yaaas" supported by the hands up in agreement emoji; "This is every-thing"; and "Yas girl, pussy power." The sign's divisive humor perhaps resonated with Instagram users because it reflects a shift in feminist tone since the 2017 Women's March. The first event launched as a reaction to Trump's misogynistic behavior and the GOP's attempts to cut female health care. But, since then, we have been confronted with the Weinstein scandal and a myriad of sexual harassment cases highlighted by the #Metoo movement and #Timesup initiative. The shocking extent of abuse, identified across different industries, has accelerated and widened the objectives of the current women's movement, accompanied by widespread expressions of anger and the adoption of a combative tone. That tone is evident in this sign, where it takes the form of "rebellious humor that simultaneously mocks the powerful" and creates connective laughter among the unruly women of the March. Unruly women do not conform to traditional norms of femininity that emphasize women's passivity, compliance, and agreeability . Instead, their unruliness is characterized by aggression, humor, and their will to offend and challenge the patriarchal status quo . Once again, these characteristics draw our attention to Bakhtin's work on the "carnivalesque" as a form of grotesque resistance. For Bakhtin, the female body signifies the grotesque body because "woman is related to the material bodily lower stratum" through menstruation, pregnancy, and childbirth. This is a problematic label that high-lights the limitations associated with Bakhtin's claims that the carnival was a liberating event that embraced all people. "The female association with the lower bodily stratum connotes shame and filth, which works in contrast to the cerebral upper body that is associated with higher functions of thought and emotion" . It suggests that the female body was a victim of the carnival's subversive comedy, rather than an instigator of it. This argument resonates with the work of other scholars , who argue that women were regularly subjected to physical and verbal abuse at the carnival. The characteristics of unruliness and the grotesque appear to be grounded in misogyny, as they have been used to attack women who do not conform to traditional standards of femininity. However, Rowe argues that "transgressing this line of acceptability can be a source of power for women, especially when the characteristics or unruliness are recoded and reframed to expose what they conceal": The oppression of women through the expectation that we stay silent, compliant, and do not make a spectacle of ourselves. The women of the March use the semiotic resources of unruliness as part of affective strategies to claim visibility, voice, and agency and to reposition Trump as the grotesque body. Consequently, the male body that stands accused of mocking, attacking, and attempting to police the female body becomes the protestors' symbolic target of collective, angry, and offensive humor. While Trump may be the sign's chief target, its humorous political message can also be read as an attack on all those who reinforce the patriarchal status quo. Its circulation on Instagram might be particularly valuable, then, because of the site's preoccupation with conventional beauty and body standards . However, while Instagram may bolster traditional notions of femininity, its audience of 800 million active users and its visual-led content make it an attractive platform to challenge these conventions via online activism. Deluca and Peeples stress the power of visual communication in their theorization of the public screen. Their work attempts to expand our understanding of political debate beyond the emphasis on face-to-face rational dialogue of the public sphere by arguing that the use of spectacle across image-led media platforms can expand dialogue and make political issues more accessible . Consequently, Cardi B's post indicates that the circulation of feminist content within the feminized sphere of Instagram might have the potential to attract new supporters to the Women's March movement and create a space for users to deliberate its messages, beyond the temporal and spatial confinements of the marches themselves. --- Conclusion This article has unpacked some of the ways in which protest signs in the 2018 Women's March used offensive humor to challenge Trump and reflected on how they were recontextualized and circulated on Twitter and Instagram. Through the close analysis of three social media posts, we have explored how Trump was dismissed as an improper White, masculine subject, how he was degraded through an association with feces and the lower bodily stratum, and, finally, how he was repositioned as the infantile, feminized victim of unruly women. Emphasizing the ambiguity of offensive humor, we identified a recurring tension not only between its uniting and dividing functions but also between conservative and radical tendencies: Protesters and social media users attacked Trump's patriarchal and racist policies and practices through the use of gendered and raced insults that simultaneously reinforced established notions of ideal White masculinity. This duality worked in two key ways: First, protesters identified and punished Trump as a transgressive other while redrawing the boundaries of appropriate White masculinity. This discourse articulated offense at his transgressions of established norms for public sphere debate and "presidential" behavior while simultaneously giving offense back through the spectacle of unruly, carnivalesque protest. Second, participants repeatedly appropriated aggressive, humorous strategies of offense giving that have been associated with masculine cultures and used to oppress female and non-White bodies . Here, they reversed that hierarchy by repositioning Trump as the abject body; malformed, leaking, and prostrate. We argue that the use of offensive humor in feminist protest in online and offline spaces can open up new opportunities for unruly dialogue and civic participation. Online networks are central to this practice, as humorous content grabs our attention and is shared through followers and hashtags . This is "spreadable" media content, which is privileged by popular platforms because it is entertaining and drives online traffic. As such, it "floats to the top" among representations of the 2018 Women's March and becomes part of the "popular memory" of the event. This pattern is evident in the frequent publication of news articles listing the "funniest" protest signs from the march, which promotes the value of humor and spectacle in protest communication. The signs, at the protests and in their new online contexts, offer the pleasures of creative, transgressive humor and offense giving . This invites us to see the world differently through the carnivalesque lens of affective intensities, reversed hierarchies, and a grotesque aesthetic. Thus, in addition to laughter, offensive humor provides an effective intervention in the dominant regime by allowing unheard voices to be heard and to respond to the issues they face. But offensive humor as spectacle is not a tactic solely used by liberal protestors, it has also been used by right-wing Tea Party activists to draw attention to America's economic issues. Interestingly, bar one study on the political content of Tea Party protest signs there is a deficit of research on conservative movements and humorous collective action strategies. Therefore, we believe that this would make a worthy topic of further research or comparative analysis of oppositional political movements. Returning to the subject of the Women's March, we argue that offensive humor is a worthy political tool that readdresses traditional understandings of protest strategy in an attempt to publicize neglected political issues. This is because its attention grabbing power might introduce citizens who do not see themselves as "political" to relevant issues, thereby "preparing them for civic participation and political engagement" . Furthermore, offensive humor appeals to like-minded others, thereby aligning bodies with feminist and antiracist counter politics in communities of resistance that include and transcend the geographically and temporally bound march events. This process can facilitate new insight and energize participants to continue their feminist and antiracist work in other spaces.

At the Women's March in January 2018, many protest posters featured offensive jokes at the expense of Trump's body and behavior. Such posters were shared widely online, much to the amusement of the movement's supporters. Through a close analysis of posts on Instagram and Twitter, we explore the role of "vulgar" and "offensive" humor in mediated social protest. By highlighting its radical and conservative tendencies, we demonstrate how we can understand these practices of offensive humor as a contemporary expression of "the carnivalesque" that is complexly intertwined with social change.

INTRODUCTION Over the last decade, the possibilities and pressures that cities are facing related to complex challenges like climate change have surged. The global migration of people into cities has continued, and organizations like the International Panel on Climate Change have begun emphasizing the vital role of sub-state actors in climate adaptation and mitigation in actions related to land-use planning, infrastructure, transportation, housing, community development, and others 1 . National and international networkserving and advocacy organizations working on climate change and other sustainability issues at the municipal level have emerged, for example, C40, the Global Covenant of Mayors, the Carbon Neutral Cities Alliance, and the Urban Sustainability Directors Network. Political and civil servant leadership from cities is a powerful force for change on the world stage, with growing influence on other levels of government, business, civil society, and the public imagination in far-reaching ways. While city leadership on climate-related work continues to strengthen, it is often treated as a technical, engineering, land use, and financial challenge and the related systemic challenge of ongoing and growing inequity tends to be un addressed in urban climate plans and policies [2][3][4] . While cities work to mitigate and adapt to a changing climate, the very same people, lands, and waters who have been systemically excluded, oppressed, and exploited by the dominant structures and processes of government continue to be left out [5][6][7] . This is a compelling impetus for transformative innovation, so that in their eagerness to act on climate, cities do not recreate and perpetuate problematic practices from the past/present that result in inequity and oppression 8,9 . There is an opportunity to reimagine the policy-making and planning processes of local government to ensure that climaterelated concerns are considered in an integrated way with equity, justice, reconciliation, and decolonization work 10,11 . Climate change issues have the same systemic root causes of colonialism, capitalism, white supremacy, patriarchy, ableism, and cisheteronormativy and can and should be worked on together [12][13][14] . The standard policy-making and planning processes used in local governments in many parts of the world, and embedded in the New Public Management, western, colonial governance paradigms, are being reinforced rather than being reimagined in many cases as governments work to innovate in response to complex challenges like climate change [15][16][17][18][19][20][21] . This remains the case even as the suitability of these paradigms to respond to complex, social and ecological justice challenges at the rate, scale, and depth required is being questioned, with calls for transformation and innovation becoming louder and more urgent [22][23][24] . The standard policy making cycle of agenda-setting, policy formulation, decision making, policy implementation, and policy evaluation is deeply embedded into the thinking, structures, and processes of governments around the world. This cycle "offers practitioners a means to conceptually and practically guide an issue methodically through the policy process" 25 and has been both described and critiqued for decades, with academics seeking further nuance and complexity to the model and practitioners requiring a variety of process interventions that work in the real world on increasingly volatile, uncertain, complex, and ambiguous challenges 26 . Similar to the standard policy making cycle, urban planning is characterized by evidence-based rationality and objectivity, established administrative routines, rule-bound bureaucratic procedures, and significant emphasis on a planner's ability to know what is good for people generally, or the public interest 17,27,28 . While planning has evolved over decades to consciously seek to redistribute power through equity planning 29 and to better align with the goals of communities through radical planning, planning processes often still involve "the analysis of problems, the setting of broad objectives, the survey of available resources, and the establishment of specific operating targets'' 28 . As planning processes continue to impede innovation and adaptation to dynamic pressures, planning must evolve to address the unpredictability of the complex challenges we face 27 . For instance, a pressing challenge for planning is reconciling its role in perpetuating the dispossession, oppression, and marginalization of Indigenous Peoples and communities and grappling with the reality that all planning in Canadian cities happens on stolen Indigenous lands [30][31][32] . In Decolonizing Methodologies, Linda Tuhiwai Smith provides a working definition of decolonization, which "does not mean and has not meant a total rejection of all theory or research or Western knowledge. Rather it is about centring [our] concerns and world views and then coming to know and understand theory and research from our own perspectives and purposes" 33 . Central to decolonizing methodologies within the context of working to transform complex urban climate, equity, and decolonization challenges is recognizing the coexistence of distinct knowledge systems, and how these distinct knowledge systems can work alongside each other in a good way 34 . Conventional systemsbased approaches have foundations in the scientific method that pursues knowledge in an analytical way, while Indigenous ways of knowing is about the pursuit of wisdom in action 35 and using the mind, body, emotion, and spirit to make meaning of the world around us 36 . Importantly, Indigenous ways of knowing are innately tied to the land 37,38 . Scholars and practitioners are applying decolonizing principles to their work, and central to this process is the idea that change happens from people going inward within themselves 34 . Planning practice is starting to acknowledge this turn inwards as well, and is recognizing that people are not disconnected from other living beings and that we must connect our present lives with nature 39 . These principles are a departure from the rational and objective mindsets that planners and policymakers have been trained to use for decades. Concepts like justice, equity, and decolonization are not static, singular, with clear and agreed upon meanings. They are contested and actively being worked out in institutions, processes, and everyday life 40,41 and are expressed differently according to place, context, history, and culture. Tuck and Yang point out that settler colonialism and its decolonization implicates and unsettles everyone and argue that decolonization is only about land and Indigenous life 42 . In other words, decolonization is not a metaphor for "other things we want to do to improve our societies" 42 . This article pushes us to think critically about the word decolonization and what it means in planning and policy making practice. Tuck and Yang's settler moves to innocence reveal attempts to reconcile settler guilt and complicity that do not require giving up land, power, or privilege. We agree the central work of decolonization should be the return of land. At the same time, as planners and policymakers doing research entangled with practice, we are motivated to find meaningful entry points to decolonization through practices that every planner and policymaker can enact. In this research we take the perspective that settler colonialism is an ongoing structure and not an event 43 . Our guiding principle, at least for now, is that decolonization work is for all of us to do and so we are enacting it as practice, and expressing these practices in multiple and embodied ways in the specific context of complex policy making and planning challenges. That said, decolonization work will look different for non-Indigenous peoples than it does for Indigenous Peoples. For Bob Joseph , acts of decolonization restore Indigenous worldviews, restore culture and traditional ways, and decolonization replaces Western interpretations of history with Indigenous perspectives of history 44 . Here we focus on policymaking and planning practice happening within settler colonial systems of local government, and explore what and how we might learn and practice equity, justice, and decolonization for/among each other in the context of urban climate work. The nature of planning and policy-making processes as they relate specifically to climate change mitigation and adaptation challenges are distinct and complex. Climate change research has rapidly expanded and evolved over the last decade, with studies being undertaken in every discipline, using diverse methodologies, providing insights for every sector, and in every part of the world. The collective of researchers that forms the Intergovernmental Panel on Climate Change is capturing this vast field and then informing, motivating, and compelling responses from global actors through their Working Group reports and summaries for policy makers 45 . With each report of the IPCC there is stronger consensus about the impacts, extent, and urgency of a changing climate. Promising pathways for action are becoming clearer and better evidenced, and enabling conditions and barriers to action are descriptive and explicit. The processes of activating and scaling robust, global responses by elected leaders, civil servants, financial institutions, businesses, nonprofits, in the short time frames to reduce the most significant impacts on our already changing world remains fraught and inconsistent. There is much to gain and much to lose, with unequal distribution of burdens and benefits. There is an unreliable patchwork of country-and sub-state level policies, plans, targets, and financing commitments with varying degrees of commitment and impact, and very few enforceable accountability mechanisms in place. This approach continues to leave those most vulnerable to a changing climate improperly resourced to act 46 . The roles of sub-state actors, including cities, is becoming more visible in this recent IPCC work, and this global climate context plays out in the daily lived experiences of city dwellers. There are many tensions and conflicts that exist in urban climate policy and planning, with civil servants needing to navigate the complexities of varying political support, limited jurisdiction, inadequate financing tools, challenges with data availability and staff expertise, bureaucratic silos, and inconsistent regulatory and accountability frameworks. What then is demanded/required of urban planners and policymakers if the dominant approaches taken in their work no longer serve complex, intersectional, and systemic challenges of climate, equity, and decolonization? Theories and practices from the fields of social innovation, systemic design, and decolonizing methodologies offer promising alternatives for working differently on these complex challenges. Social innovation theory integrates many fields of social science and systems practice in service of shifting deep structures -hearts, minds, and culture -and of scaling and movement building for transformative socio-ecological change. Systemic design brings a designerly mindset, lineage, frame of reference, training, approach, and set of experiences into this mix. The experimentation-and action-aslearning bias of systemic design ensures that innovations land in the real world, in testable experiences, with-and for the people and places most affected by a particular challenge. Decolonizing methodologies ensures that social innovations and systemic design do not work within dominant and problematic paradigms of colonization and oppression and inadvertently perpetuate these systems under the guise of 'innovation.' Instead wisdom and insight is sourced from culture, history, people, and possibility and is deeply grounded in place-based and relational practice. Together they can inform the evolution of policy making and planning processes in ways that may result in more significant shifts than are possible when working within dominant systems that have oppression, inequity, and exploitation of humans and nature baked in. Social innovations work to transform the behaviours, structures, mindsets, and beliefs of a social system with an intent to more skillfully and effectively respond to problems than is possible through existing or commonly used approaches that are less likely to address root causes of wicked challenges [47][48][49] . An innovation is 'social' in that it aims to shift social practices, ideas, beliefs, interests, power, and agency so that innovations are diffused, scaled, institutionalized, or otherwise integrated and made routine. Social innovation processes follow a series of stages, often facilitated as iterations or cycles of experimentation and learning as new/different solutions emerge through a collaborative process, including the following stages: problem/challenge framing; action or user research; generating ideas; developing and testing ideas through experimentation; evaluation, reframing, and iteration; implementation at increasing levels of fidelity; scaling successful solutions out/up/deep; and ultimately changing systems [50][51][52][53][54] . Some social innovation process archetypes stay focused on the specific challenge they are tackling, whereas others bring in a stronger personal transformation orientation, recognizing that inner work is required for outward-oriented change [55][56][57][58][59] . Systemic design integrates systems thinking into humancentered and service design approaches for working creatively on complex challenges. Jones & Kijima say that "systemic design advances an integrative interdiscipline with the potential to implement systems theory with creative methods and mindsets, by bringing deep technical knowledge, aesthetic skill, and creative implementation to the most abstract programmes of collective action" 60 . Systemic design processes tend to oscillate between convergent and divergent thinking and doing, hold a strong bias toward action and experimentation, and enable right-sized risk-taking and micro-failures early on as critical pathways for learning. Examples of systemic design process archetypes include: ask, try, do; strategy, discover, design, develop, deploy; and inquire, frame, formulate, generate, reflect, facilitate [60][61][62][63] . Systemic design has roots in human-centred or service design practice which center user experiences; however, it significantly differs from these modes of design in that it holds deeper attention to mindsets, values, context, power relationships, leadership, and systems, and then connects these activities in a strategic learning system [64][65][66] . Recent critique of the field and practice of design from an equity and justice point of view is resulting in theory, frameworks, and approaches to encourage/enable/require centering equity in design processes more skillfully and consistently [67][68][69][70] . Social innovation, systemic design, and decolonizing methodologies have strengths that were woven together into a different planning and policy making process in this applied research. In the next section we describe how we designed and implemented this weaving in our project, as well as into a framework that describes a transformative approach to policy making and planning processes. --- RESULTS --- Experimenting with transformative process design and facilitation The transformative policy making and planning process framework described in this section resulted from a year-long participatory action research study that was delivered in 2021. The learning journey was designed and facilitated by a diverse seven-person core research team , consisting of formal academic researchers as well as researcher-practitioners. Learning journeygoers, also known as our co-researchers, included municipal staff from multiple departments working in ten Canadian city-based teams , and staff from two Canadian network serving organizations . Each team began their learning journey by preparing a design brief that described their understanding of the complex climate, equity, and decolonization question that they were beginning with, the North and Near stars for their work, the theory of change informing their current approach to transformation, description of the people/places most impacted by their challenge, and the inner work and learning they committed to do. The learning journey was delivered through thirteen half-day virtual workshops over nine months and active 'invitations to practice' in between sessions. A visual representation of the process arc and high-level focus of each of the workshops and stages is shown in Fig. 1. The learning journey was designed and facilitated as an intentionally transformative alternative to more standard and familiar processes. Each of the dots represents a half day synchronous, virtual workshop led by the core research team. Each workshop had a specific focus topic and practice, and overall the sessions were designed with consistent core elements, including: slow and reflective experience to enable deeper systemic work; mindfulness/embodiment practice; reflective practices; more formal teaching and content-giving; stories from the field; connecting and relationship building; team working/ practice activities; and sharing our works in process. In between the workshops were specific activities/homework, described as 'invitations to practice,' for the teams to apply and experiment with. The core research team used interactive slide decks, online collaboration tools and templates, one-on-one and team coaching with the core research team, and dialogue as central andragogical approaches. The teams experienced a transformative approach to policy making and planning which required them to reimagine and reframe their work. The process was designed to open up thinking about potential planning and policy making process interventions outside of what they were practiced/expert in, and to inquire into and challenge the Fig. 1 Transforming Cities from Within learning journey process map. paradigms that hold the current dominant ways of working in place as standard practice. It was also designed to carefully and lovingly hold the discomfort, tensions, and ambiguity that this approach generated for people to help them to stay with the learning journey. Transformative policy-making and planning process framework This framework for transformative policy making and planning was generated through engagement with literature and through interacting with empirical data generated during the learning journey. This framework is presented here as two parts -a process map , followed by a description of the transformative approaches taken at each stage of the process . The process map captures key stages of this transformative approach, as well as the movement and connections between the stages. Note that these are not drawn as discrete, linear steps but rather as interrelated and entangled stages with iteration between. There is a general movement from broad/messy/open/exploratory that begins with framing, and toward a gradual focusing/clarifying process as we move through initiating, researching, experimenting, and eventual implementation at the scaling stage. Engaging and enabling, and evaluation and learning, happen throughout. As the connecting dotted lines indicate, there is iterative movement between stages when insights and learning require/indicate reframing, renewed experimentation, and/or return to learning from those most impacted by the challenge. Transformative policy-making and planning approaches Tables 1-6 detail the process map into descriptions of transformative approaches at each stage. Common policy-making and planning approaches are compared to transformative approaches at each stage, along with a description of how the transformative approach took shape in the learning journey. Some additional thinkers informed the ideas about the transformation that inspired these approaches [71][72][73][74][75][76][77][78][79][80][81] . This transformative policy-making and planning framework can be considered as a whole, and can also be considered as smaller/ independent interventions into unfreezing standard processes when an overall process redesign is not possible or supported. Careful creation/curation of transformative processes that centre learning are essential when working with practitioners who are used to being expert professionals in standard policy-making and planning processes, with many thoughtful, intentional, and reflective process design choices required to do this well. The next section describes seven of these significant choices and interventions from our research as key moves to provide texture about how policymakers and planners might think about designing and delivering a transformative process. --- DISCUSSION The seven key moves toward transformative policy-making and planning processes explained below were intentionally integrated into the learning journey as an attempt to experiment with meaningful entry points into decolonization practices for planners and policymakers. The hope was that participants might enact these moves in their everyday professional lives. We express these practices as moves because significant reflection and interaction took place during and between each session for both learning journey goers and the research/ facilitation team. For example, the research team/facilitators had to consistently design for/work with/be comfortable with emergence and ambiguity, as the learning journey often and intentionally challenged participants to turn inward, reflect on their identities and positionality, show up as their authenticity selves and encouraged vulnerability in small and large group settings. The details of these moves are discussed below. Move 1: beginning well Our core research team members each had previous experience with designing and delivering transformative processes in diverse public sector, community-based, and academic environments which gave our team a strong collective foundation from which to grow. We began by crafting a compelling invitation to join this process based on the real-world challenges that we knew city staff were struggling with. This clarity of invitation and sensing of the field meant that the right people found their way in, and we had very little attrition along the way. The learning journey started with an intensive virtual retreat in order to co-create excitement, build relationships, and establish clarity about what people Fig. 2 Transformative policy-making and planning process map. Complex, systemic, and adaptive challenges demand different approaches to working on them -they are unknowable, unpredictable, and emergent. These features require different approaches to learning and working with/on these types of challenges. Example practice: systems mapping is used to understand dynamics at work in the challenge, and to expand the view to include interrelated issues. Leverage point analysis is used to identify high impact places to intervene in the system through policy and planning work. --- Table 2. Initiating Stage --- Common approach Transformative approach Brief description of how this transformative approach shapes process design --- Begin with outcome Begin with question Transformative processes begin with an active, relevant, and critical question that is framed along the way as user research, experimentation, and learning reveals insights and possibilities. Example practice: designers often begin work with a creative question framed as a 'How might we…' or 'What if…' possibility. Good creative questions bound the challenge space and also open up creative possibilities and visionary potential. --- Project charter and project management plan --- Design brief and theory of change Captures the vision, goals, values, process, evaluation and learning, and potential outcomes of an innovation process in a concise and compelling way. Outlines a structure for a process in a way that also allows for non-linearity, not knowing the outcome, taking a user-centered orientation, having a long-term and shorterterm vision and goals, and enabling creativity, experimentation, and learning. Clearly describes how a team thinks that change happens, and their intended/ desired contributions to change. Example practice: a theory of change articulates vision, goals, a description of how a team thinks that change happens , and how their work will contribute to that change. It is a helpful practice to surface implicit assumptions and uncover differences or tensions, and then to design for these in a productive way. --- Table 3. Research Stage Common approach Transformative approach Brief description of how this transformative approach shapes process design --- Scenario planning Speculative and visionary fiction Imagines possible futures beyond the limited frames that planning and policy making typically take, by tapping into creative, speculative, protopic, future world-building through imaginative writing. Example practice: creative writing/drawing and worldbuilding using techniques from fiction can help a team to expand their sense of what is possible and necessary, to see potential pathways, and to be inspired by one another's visions. --- Policy research and best practices Action and user research Centres the lived experiences of people most affected by a complex challenge, the lived experiences of land, water, and more-than-human kin, and future generations in research. Example practice: ethnographic and/or action research insights are regularly returned to users to test how they are informing a planning and policy making process throughout in generative and reciprocal ways. wanted the journey to be for them. The retreat series included connecting to land, place, history, context, and Indigenous pasts/ presents of each place where cohort members were coming from. One cohort member reflected that "I think one thing that was probably challenging for us at the beginning was figuring out how these things relate to engineering and our very specific transportation engineering field." The retreat series also included cohort members reflecting on their personal and team calls to action and accountability based on their identities and experiences, reframing the complex challenges that they were coming in with based on systems practice, and imagining possible and desired future states resulting from their transformative work. These choices clearly signaled that this process would be different from other planning, policy-making, and/or professional development processes than they were used to. Move 2: co-creating an equitable, just, feminist, and decolonized space Concepts of equity were intentionally introduced throughout the learning journey to practice foundational concepts and develop a shared language. Grounding the process in place-based, relational, and decolonial approaches set the tone for how the cohort was invited to bring themselves into the space. Some features included the practice of abundant time, and the dance between this and the sense of urgency associated with the nature of the challenges everyone was working on. We did deep self-in-system work to explore how our individual positionality, intersecting identities, and life experiences impact how we each see, understand, and experience the world and the challenges we were working on. We worked to transform the construct of being 'expert', and of 'professional' looking and behaving in a particular way characterized by colonial and white dominant paradigms. One cohort member described it this way: "The reimagining of these core city processes has the potential to impact other systemic challenges as well, because at its foundation this is a shift toward more place-based, peoplecentered, experimental, and exploratory processes where we collectively need to lean into the emergent solutions that we are working toward because they are not yet known." Ongoing reflection on how there is no step-by-step guide to do decolonial practice was central in the process as, for most of us, this is a lifelong learning process filled with discomfort. Move 3: pro-love pedagogy Partway through the process our research team began using the phrase pro-love pedagogy as a short-form descriptor for a collection of choices we were making in design and facilitation. Pro-love meant that we prioritized being in a caring, loving, supportive, and collaborative community with/for each other as research team members, and consequently how the cohort experienced the holding space that we co-created with them throughout the process 75,81 . One member of our research team reflected that "being in a caring and loving research team changed my perspective about research, and what that can look and feel like." Evidence of this pro-love pedagogy looked like an ongoing trust and willingness of the cohort to come along on this journey, even when it got uncomfortable and they didn't know where it was taking them. It looked like the research team navigating and holding our own and each others' discomfort, disease, and uncertainty and modeling a skillful and honest navigation of this liminal space. It looked like an opening up to the diverse experiences and perspectives of the research team to co-create a rich learning-oriented process for everyone even when it wasn't clear how all of the pieces fit together. It looked like radically trusting ourselves and each other at every step along the way. --- Move 4: focus on practicing not problem solving The process emphasized building and practicing competencies and capacities for transformation rather than on solving problems. We thought about this as exercising different muscles, and rehearsing the new/resurgent. We did this by co-creating a learning, practice, experimentation, and reflection-focused space with/for each other. The dominant problem-solving orientation in planning and policy-making processes taps into well-honed and exercised skills, habits, approaches, and subsequent results. We needed the cohort to question, reframe for themselves, and look beyond this standard approach if they were to have a real chance at getting to a different outcome. Working on complex, ambiguous challenges demands a different set of skills and approaches, and practitioners can quickly find this overwhelming and daunting to navigate. One cohort member shared that the journey helped them to "think about how we can overlap our different programs to create multiple effects instead of just using one program to treat one problem." Move 5: drawing on different theories and practices Our research team carefully curated a set of theories, approaches, and practices that are not commonly used in city governments, including action and user research, systems mapping, prototyping, reflective practice, developmental evaluation, imagining possible futures/speculative fiction, prototyping, and others. The cohort worked the techniques, made mistakes and had breakthroughs, shared their experiences with one another, and built their literacy and confidence in approaches more suited to the complex challenges they were working on than their go-to approaches. An example of this was the regular returning to systems mapping throughout the journey to develop each person's understanding of the deeply connected and shared roots of climate, equity, and decolonization challenges. We worked and iterated these systems maps as people expanded and shifted their perspectives, and then used the maps to generate promising fractals and feedback loops where they could intervene. One participant captured their experience in this way: "It's so easy to just be in the normal ways of doing and being; it's so ingrained and easy to go along with it. I realized I can question this, and that it takes so much presence and intention to hold a different possibility as this isn't part of people's jobs or recognized as legitimate work." Move 6: working with fear We also noticed the implicit influence of fear as a motivator of in/ action and behaviour -fear of messing up, of saying the wrong thing, of offending, of failing to have impact. Surfacing and shining a light on these fears and supporting the cohort to develop different, more generative relationships with these fears and what they might contribute or catalyze was an important part of the process. A particular fear that many people acknowledged was that they lacked the skills, abilities, and connections necessary to work in highly relational, respectful, and mutually beneficial ways with the people and communities whom they serve, and who are most impacted by these intersecting challenges. One cohort member said that they "don't have the relationships with equity-denied people and groups. The fear that staff have is projected onto people -otherizes them, focuses on risks, and is based in fear. What if the foundations of our work were about good relationships instead?" Another member realized that we need to "prioritize the time and space needed to reframe the complex problem, asking 'who have we not talked to?' and then prioritize action and user research to understand their experiences and perspectives." Move 7: redefining impact and outcomes Standard planning and policy-making processes offer the comfort of a clear endpoint -the successful approval of a piece of work by decision-makers, elected officials, and community stakeholders. Transformative processes are not as clear and tidy. They are concerned with transformation at the personal, team, organizational, cultural, relational, and systems scales and in this case, on complex intersecting climate, equity, and decolonization challenges. There is no clear finish line in this work as it is long-term, ongoing, and generational. Cycles of action, learning, and reflection are necessary, putting planning and policy-making processes in a direct and ongoing relationship with implementation, rather than viewing them as separate processes. Generating other ways to measure, understand, reflect upon, evaluate, and tell stories of impact was an important part of grounding the experience of this process for people, and supporting them in articulating why it mattered and what resulted. We used ongoing developmental evaluation, speculative fiction, and persuasive storytelling to support individual and collective sensemaking about impacts. Imagining, enabling, and enacting transformative policy-making and planning processes will require equipping both new and established planners and policymakers with emerging/resurging approaches to complex urban climate, equity, and decolonization challenges. The transformative policy-making and planning process framework shared here intends to support research and practice in this important domain, and contribute to enhancing research and action more generally on complex, intersectional urban sustainability challenges. --- METHODS --- Participatory action research bricolage The core method used in this study was participatory action research , defined as "a participatory process concerned with developing practical knowing in the pursuit of worthwhile human purposes. It seeks to bring together action and reflection, theory and practice, in participation with others, in the pursuit of practical solutions to issues of pressing concern to people, and more generally the flourishing of individual persons and their communities" 82 . In PAR, research subjects are co-participants in the processes of inquiry which includes activities of generating questions and objectives, sharing knowledge, building research skills, interpreting findings, and implementing and measuring results. Recent turns in PAR call for the method to support transformation while providing rigour and quality in ethics, process, and outcomes and this provided additional guidance to our approach [82][83][84] significance which informed the method taken in this work 82 . PAR was embedded within an interdisciplinary, critical, qualitative, activist, and practice-oriented bricolage 85,86 "concerned with the depth and complexity of a question, what lies below the surface, and the form in which that complexity might best be understood and revealed" 87 . This methodological layer was added to ensure that PAR would critically engage with the dominant systems and structures of planning and policy-making that this research was aiming to question and reimagine. The two authors of this paper, along with the five additional members of the core research team, each brought diverse lived experiences, professional practices, distinct disciplinary lineages, and diverse positionalities and perspectives together to construct this PAR bricolage, and to design, facilitate, and generate insights about this learning journey together based on our collective of experiences and perspectives. Indigenous, feminist, anti-oppressive, and ecological methods were features of our bricolage, and necessary to better support working in/with processes of transformation, and to find our ways out of the dominant paradigms of planning and policy-making through our methodological approach [88][89][90][91][92][93][94][95][96][97] . Using a PAR bricolage meant that we took an orientation of gathering up and applying what was useful from these different methods, in a generative dialogue with these methods and with each other. This also helped us to remain focused on being in service to generating insights about our practice-oriented questions, and those of our 40 co-researchers in the learning journey. This methodological approach enabled our research team to balance theoretical rigour, transdisciplinarity, practice-based foundations, and emergence. Together we held an entangled and engaged researcher orientation where the opinions, perspectives, and experiences of our research team and action coresearchers were considered a strength. --- Research design The University of British Columbia Behavioural Research Ethics Board provided approval for this study . All research participants provided 'written informed consent' to take part in the study. The core research team crafted the learning journey and invited participation through an open call that was distributed through existing networks supporting city governments working on climate and equity-related work in Canada. This invitation was open to public sector staff working within City governments, and could include team members from community-based partner organizations if this would be relevant to the challenge that the team wished to work on during the journey. It was also open to network-serving organizations that worked to support City governments in their climate, equity, reconciliation, and/or decolonization work. All of the ten Canadian city-based teams and two network-serving organizations that applied to participate in the journey were invited to join. City staff were from the environment, sustainability, social policy, planning, engineering, equity and reconciliation, and public engagement departments/ roles. Teams completed the application requirements, and our research team held the approach of saying yes to as many teams that were ready as we could properly support. One team dropped out of the learning journey about halfway through due to their limited capacity to fully participate, although they continued to be invited into all learning journey activities through to the end of the project. Team's were informed that this was an applied research project before applying, and upon acceptance into the learning journey their formal consent to participate was confirmed. Several new people joined the journey part-way through, with their consent confirmed when they joined. Each multi-departmental team brought a different, locally relevant, complex planning and/or policy-making climate, equity, and decolonization challenge that they actively worked on throughout the nine-month-long learning journey held from April -December 2021. Social innovation, systemic design, and decolonization theories, processes, and practices were used to design the backbone of the learning journey. Other theories and processes informed the andragogical approach as well, including transformative adult learning; critical race theory; feminist and queer methodologies; sustainability transitions; complexity and emergence; and AfroFuturism. Our core research team shared the responsibility for the overall design of the learning journey, with individual team members and external guests invited in to lead and facilitate specific workshops based on their expertise and experience. The research team engaged in ongoing reflective practice throughout the whole journey, documenting insights and learning along the way, and shifting and pivoting the design of the journey in response to what was happening in-session with co-researchers and their emerging questions and needs. Qualitative data was collected throughout the journey, and included observation, interviews, reflection of the research team and cohort members, team coaching sessions, document review, active process design and facilitation interventions and their impacts, and shared sense-making with the research team to uncover themes and insights. Active and ongoing sensemaking of this data occurred through regular dialogue amongst research team members, tracking what was emergent, responding through the design of the learning journey, and noting themes throughout. Emerging insights and themes were regularly returned to the co-researchers during the learning journey workshops to ensure that the sensemaking being done by the core research team was reflective of the co-researchers' experiences. Active consent of cohort members to participate in the research continued throughout the journey, including reaffirmed consent at their last interview three months after the journey had finished. --- DATA AVAILABILITY Data from this research is not available for other users or purposes, as per our behavioural ethics board review and approval. --- Reporting summary Further information on research design is available in the Nature Research Reporting Summary linked to this article. --- --- COMPETING INTERESTS The authors declare no competing interests. --- ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s42949-023-00126-9. Correspondence and requests for materials should be addressed to Lindsay Cole or Maggie Low. Reprints and permission information is available at http://www.nature.com/ reprints Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Cities are facing increasing pressures to address complex challenges of climate change, equity, and reconciliation with Indigenous Peoples as intersecting issues, and innovation into planning and policy-making processes is urgently needed to achieve this. It is no longer good enough to work on these challenges discreetly, or solely within the dominant, western colonial paradigm and practices of governance. There are ongoing harms being caused by climate work that does not embed justice, and there are missed opportunities for synergies across these domains as they have the same systemic root causes. Cities must adapt and transform the processes and practices of planning and policy-making in order to work at these problematic roots. Drawing on an empirical study, this article describes how social innovation, systemic design, and decolonizing practices can shape a different approach to planning and policy-making processes when working at the intersections of climate, equity, and decolonization.

Introduction n examining the reasons for the failure of development programs in different countries in the world, the neglect of development governance has been emphasized. Considering that the implementation of development policies is challenging for activists, there is a need for innovative governance models of the public and private sectors to go beyond traditional thinking. In examining the new governance paradigms, network governance, as one of the emerging models with the premise of the complexity of issues and the inability of institutions to solve issues independently, emphasizes the cooperation and participation of key stakeholders in the decision-making process, including government and non-government actors and society. In this form of governance, government agencies directly engage non-government actors in decision-making by consultation and negotiation and take collective action. This governance model reflects the participation of the non-governmental sector, government sector, and society in a decentralized structure. Network governance describes the improvement in the ability to solve sustainable problems in social systems by empowering the capabilities of various stakeholders, and strengthens the response to the problems of social and environmental systems by supporting participatory and collective actions. Tehran city's fourth development plan has been compiled based on the opinions of experts in Tehran municipality and by paying attention to the shortcomings of the previous plans and the need to be problem-oriented and avoid inefficient expansion. This plan has determined five key domains: transportation/traffic, environment/air pollution, urban safety/worn-out fabric, cultural/social issues, and intelligentization/management improvement. In examining the current situation in the field of safety and disaster management in Tehran, in addition to the existence of worn-out urban fabrics, more than 33,000 high-risk buildings in terms of fire hazards, unsafe buildings of shopping, medical, educational, industrial, and warehouse centers, more than 100 high-risk unsafe pits, accident-prone passages, lack of standard fire-fighting machines, failure to complete the dredging of roads, river valleys, and city canals, and lack of attention to the city's safety measures against floods, earthquakes and land subsidence have been reported. Considering the emphasis on participatory governance in the fourth development plan of Tehran city, this study aims to analyze the network governance of safety and disaster management in Tehran by examining the laws of the fourth development plan and with emphasis on the participation of key actors, including government institutions, private sector, public institutions, and local communities using social network analysis. --- Methods In this study, 22 laws with 101 issues were identified, and 47 involved organizations, including 26 internal institutions of Tehran Municipality and 21 external institutions involved in the safety and disaster management program in the areas of transportation/traffic, urban safety/ worn-out fabric, cultural/social issues, and intelligentization/management improvement were evaluated. The organizations and institutions were considered as the rows and columns of the governance network matrix, and the number of times of cooperation between organizations and institutions was considered as the entries of cells of the matrix. The matrix of actors' cooperation network was analyzed based on the degree centrality, betweenness centrality, geodesic distance, density and cohesion. --- Results The results of the analysis of the degree centrality index for the safety and disaster management actors' cooperation network showed that among the internal organizations of Tehran Municipality, the Crisis Prevention and Management Organization, the Deputy of Urban and Environmental Services, and the Information and Communication Technology Organization had the highest scores, followed by the Fire Department and the municipalities of 22 districts of Tehran. The internal organizations with the lowest degree centrality were the General Department of Citizenship Education, University of Applied Sciences and Technology, Neighborhood Management Headquarters, Deputy of Finance and Urban Economics, and the Renovation and Administrative Transformation Center. Regarding the betweenness centrality index, the Crisis Prevention and Management Organization, the Information and Communication Technology Organization, the Deputy of Urban and Environmental Services, the Deputy of Urban Planning and Architecture, and the municipalities of 22 districts of Tehran had the highest scores, while the Urban Research and Planning Center, the General Department of Citizenship Education, University of Applied Sciences and Technology, and the Neighborhood Management Headquarters had the lowest scores. In general, the results of the analysis of the governance system of Tehran city showed the planners' belief in participatory governance for solving the safety and disaster management issues of the city. However, the participation of citizens, local communities, non-governmental organizations, and the private sector requires improvement and expansion of legal cooperation such that the lack of transparency in determining the roles, responsibilities and powers of institutional actors outside Tehran Municipality, such as executive, legislative and supervisory institutions, and absence of policies for facilitating and encouraging the participation of private sector, economic institutions, universities, scientific and research centers, local communities and citizens need to be addressed by legislators and planners. They should support the network governance in solving disaster management issues in Tehran. --- Conclusion The recommendations for improving the governance system of disaster management in Tehran for balancing the distribution of legal power among the responsible institutions and actors of Tehran's safety and disaster management program include: Strengthening the cooperation of municipal and inter-institutional actors to solve key issues of safety and disaster management; strengthening the communication network and continuously exploiting the capacity of academic elites and consultants; strengthening the mechanisms for the participation of citizens, nongovernmental organizations and the private sector; rising awareness and strengthening citizens' trust in safety and disaster management programs using social media, developing incentive policies to utilize the capacity and participation of knowledge-based organizations and new and creative firms in solving safety and disaster management issues; developing incentive policies for the participation of managers and employees; strengthening the legal communication of financial and economic institutions and financing projects; and transparency and accountability of responsible institutions. --- Ethical Considerations Compliance with ethical guidelines All ethical principles were observed in this study . --- --- Conflicts of interest The authors declare no conflict of interest.

Background and objective This study aims to analyze the network governance of safety and disaster management in Tehran by examining the laws of the fourth development plan and emphasizing the participation of key actors, including government institutions, the private sector, non-governmental organizations, and local communities using social network analysis. Method In this study, 22 laws with 101 issues were extracted from the fourth development plan, and 47 involved organizations, including 26 internal institutions of Tehran Municipality and 21 external institutions involved in the safety and disaster management program in the areas of transportation/traffic, urban safety/worn-out fabric, cultural/social issues, and intelligentization/management improvement were evaluated. The organizations and institutions were considered as the rows and columns of the governance network matrix, and the number of times of cooperation between organizations and institutions was considered as the entries of cells of the matrix. The matrix of actors' cooperation network was analyzed based on the degree centrality, betweenness centrality, geodesic distance, density and cohesion.The results of the analysis of the governance system of Tehran city showed the planners' belief in participatory governance for solving the safety and disaster management issues of the city. However, the participation of citizens, local communities, non-governmental organizations, and the private sector requires improvement and expansion of legal cooperation such that the lack of transparency in determining the roles, responsibilities, and powers of external institutions of Tehran Municipality, such as executive, legislative and supervisory institutions, and absence of policies for facilitating and encouraging the participation of private sector, economic institutions, universities, scientific and research centers, local communities and citizens need to be addressed by legislators and planners. They should support the network governance in solving disaster management issues in Tehran.The recommendations for improving the governance system of disaster management in Tehran for balancing the distribution of legal power among the responsible institutions and actors of Tehran's safety and disaster management program include: Strengthening the cooperation of municipal and inter-institutional actors to solve key issues of safety and disaster management; strengthening the communication network and continuously exploiting the capacity of academic elites and consultants; strengthening the mechanisms for the participation of citizens, non-governmental organizations and the private sector; rising awareness and strengthening citizens' trust in safety and disaster management programs using social media, developing incentive policies to utilize the capacity and participation of knowledge-based organizations and new and creative firms in solving safety and disaster management issues; developing incentive policies for the participation of managers and employees; strengthening the legal communication of financial and economic institutions and financing projects; and transparency and accountability of responsible institutions.

Introduction Public stigma related to mental illness involves the general public's negative misconceptions about people with mental health conditions [1]. These negative misconceptions motivate individuals to fear and discriminate against PMHC, resulting in PMHC being shunned by society and losing opportunities [1]. Public stigma related to mental illness is becoming a significant concern and is detrimental to PMHC as it also leads to unemployment [2]. Stigma is a major issue because it leads to bullying, and a recent study has focused on public stigma-based bullying [3]. The structure of public stigma related to mental illness can be explained using the social cognitive model [4]. Based on this model, stigma comprises three components: stereotypes, prejudice, and discrimination [4]. Stereotypes are public attitudes , prejudice is the emotional reaction to agreeing with the public attitude , discrimination is the behavior that results from stereotypes and prejudice [2,4]. In other words, stereotypes are associated with knowledge of the mental illness, prejudice is related to emotional reactions toward patients, and discrimination is associated with behavior change such as social distance [4]. A review in 2013 based on studies from 2001 to 2011 found that Japan has intense stigmatization of PMHC [5]. This strong stigma was suggested to be, in part, a result of the culture of Japan [5]; discussing mental illness in Japan has been taboo in the past [6], given the influence of the honne/tatemae culture, which promotes maintenance of harmony [7]. Japanese are hesitant to express their true feelings and provide their real opinions and are more inclined to give opinions that are less likely to cause disputes, to maintain harmony . In addition, insufficient education, anti-stigma campaigns, and the high institutionalization rate of PMHC decrease the public's opportunity to have contact with PMHC and have been sug-Mental Illness Stigma in Japanese Employees J Clin Med Res. 2023;15:139-147 gested to contribute to the stigma of mental illness in Japan [5]. Therefore, the Japanese government has tried reducing public stigma and improving knowledge about mental illness. In 2002, the Japanese Society of Psychiatry and Neurology changed the name of schizophrenia to remove the harmful impact of the old naming term on patients and their families [8]. As a result, patients reporting their diagnosis increased from 36.7% to 69.7% from 2002 to 2004 [8]. Since then, the mental health system in Japan has undergone continuous changes for reasons such as a lack of mental health literacy and a shift toward community-based care [9]. One of the more notable reforms was the 2004 Reform Vision for Health, Medical Care and Mental Health Welfare, which promoted the transition to communitybased care [10], and decreased the number of psychiatric care beds by approximately 25,000 from 2004 to 2018 [11]. In 2014, the Japanese government passed the Act on Promotion of Preventive Measures against Karoshi and Other Overwork-Related Health Disorders, which enhanced the public's knowledge of occupational mental disorders [12]. Researchers have also adopted educational interventions to reduce stigma since 2013, which have proven effective [13][14][15]. As a result, a new educational curriculum on mental illness will be implemented in schools in 2022 [16]. Various surveys on the level of public stigma of mental illness have been conducted in Japan. In 2003, a Japanese survey found that respondents below 30 years old have lower understanding and knowledge on stigma, implying that those below 30 years old have had less chance of interaction with patients or attended lectures [17]. Another study in 2005 found significant effect of age on stigma, with older participant tend to be less socially accepting towards patients with schizophrenia [18]. The association between sex and mental illness stigma is more complex as it varies from regions to regions [19], but prior studies found that male tended to have lower understanding of mental illness compared to female [17,19]. Employment position such as being a workplace manager, who is a decision maker in employment, is associated with knowledge of mental illness [20]. Consequently, those with experience in employing PMHC were more knowledgeable about mental illness [20]. Concerning employment status, regular employees in a Japanese company are given induction training, which may include mental health and work stress management [21], affecting their knowledge of mental illness. In addition to the demographic information, attendance in mental health lectures also affect knowledge on mental illness [17]. As for prior contact experience, a prior study [5] suggested that the lack of prior contact experience due to PMHC being admitted into the hospital resulted in a more substantial stigma among the Japanese. However, most of the surveys in Japan used different stigma scales [7,17,19] and it may be difficult to look at the trend of public stigma in the past 20 years based on the results of different stigma scale. Although it was an indirect comparison, a survey conducted in 2012 [19] found that respondents associating depression to personal weakness to likely have decreased compared to the previous study [7]. Another survey in 2018 [22] found that most Japanese believed that mental illness could be cured via treatment, but still have slightly strong stigma towards schizophrenia. As such, it is expected that the state of mental illness stigma in Japan at present might have improved much more compared to the past survey conducted on knowledge of mental illness [17] or social distance towards mental illness [7] among the Japanese. Thus, the purpose of this study was to clarify the characteristics of public stigma toward PMHC in Japan in recent years, focusing on knowledge about mental illness using the Mental Illness and Disorder Understanding Scale and social distance toward PMHC using Attitudinal Social Distance scale employed in previous survey [7,15,17]. --- Materials and Methods --- Participants and procedure This study was a cross-sectional, population-based design. Healthy participants without mental health conditions were recruited from employees of three companies , located in a standard regional city with a population of 270,000 in Japan. In this study, we requested the participation of occupational health physicians and public health nurses affiliated with the company by using the kinship method. The survey instrument was designed using Google Forms so participants could complete it online. The advantages of using online Google Form were that it was convenient and can be completed anonymously at any time but was less accessible to the older generations who were not familiar with recent technologies. In the second week of January 2022, 1,468 participants were informed of the online survey by the occupational physician at their workplace. They were provided a quick response code to access the survey online via mobile phone. Once a week, the occupational physician reminded participants to complete the survey online. Data collection lasted about 3 weeks and ended on January 31st, 2022. Of the 1,468 participants, 970 did, and 498 did not respond to the survey, respectively; thus, the response rate was 66.1%. The protocol received approval from the Ethics Committees of the University Hospital Medical Information Network Clinical Trials Registry and the Research Ethics Committee of the Faculty of Medicine and Graduate School of Medicine of the University of Tokyo . All procedures performed in this study followed the ethical standards of the Institutional Research Committee and the 1964 Helsinki Declaration and its later amendments. --- Survey questionnaire The stigma of mental illness is typically measured using a selfreported stigma scale [23] together with demographic variables such as age [17][18][19], sex [17][18][19], employment position [20], employment status [24] , attendance at mental health lectures or training [17], and prior contact experience of PMHC [5]. The knowledge component of the stigma of mental illness was measured using the MIDUS [15,17]. The social distance component of mental illness stigma was measured using the ASD scale [7]. Prior contact experience of PMHC was measured using the Reported Intended Behavior Scale-Japanese [25,26]. We used the MIDUS, ASD, and RIBS-J because their internal consistency and validity have been confirmed [7,17,25,27]. The survey had two parts. Part one consisted of the demographic characteristics of the participants in the form of multiple-choice questions, sex , and age group measured in years [17], employment position , employment status , and attendance at mental illness stigma training . Part two consisted of questions on knowledge of mental illness measured by MIDUS, social distance toward PMHC measured by the ASD scale, and prior contact experience of PMHC measured using the RIBS-J. --- Accurate knowledge of mental illness The MIDUS assessed the participants' accurate knowledge of mental illness [17]. The MIDUS consisted of 15 items with three subscales: treatability of illness, the efficacy of medication, and social recognition of disease. All items were rated on a 5-point Likert scale . The MIDUS scale was developed in Japan, confirming its internal consistency and validity [17]. The Cronbach's α was 0.81 in the present sample. We used the mean total MIDUS score, as in the prior survey [17]. A low total score indicates good understanding. --- Social distance The ASD measured the social distance of participants in relation to PMHC [7,28]. The five items were rated on a 5-point Likert scale . We used the mean total ASD score, with a low score indicating favorable attitudes. The five items of the ASD were: 1) move next door to the person; 2) spend an evening socializing with the person; 3) make friends with the person; 4) work closely on a job with the person; and 5) have the person marry into the family. Participants were given a vignette depicting a patient with depression and were instructed to rate their social distance in relation to the person in the vignette. The vignette using patient with depression was chosen in this study because depression has been closely associated with malingering, such as being absence from work using depression as an excuse [29], which may influence the stigmatized view of public towards patients with depression in a workplace environment. The ASD scale was developed [30] and has been validated [27] , and we used the Japanese version of the scale [7,28]. --- Prior contact experience with mental illness patients We measured the prior contact experience of participants with PMHC using the RIBS-J [25,26]. The scale consisted of four binary items for prior contact experience with patients of mental illness: 1) Are you currently living with, or have you ever lived with, someone with a mental health problem? 2) Are you currently working with, or have you ever worked with, someone with a mental health problem? 3) Do you currently have, or have you ever had, a neighbor with a mental health problem? and 4) Do you currently have, or have you ever had, a close friend with a mental health problem? . The participants' data were analyzed based on having prior contact experience to no previous contact experience . The Japanese version has good reliability and validity [25]. The Cronbach's α was 0.81 for present sample. --- Statistical analysis --- Demographic variables and public stigma A t-test was used to compare the MIDUS and ASD scores separately by the independent variables sex, employment position, employment status, attendance at mental illness stigma training, and prior contact experience with mental illness patients. --- Age and public stigma One-way analysis of variance was conducted on the MIDUS and ASD scores by the independent variable age group . Post hoc pairwise comparisons between the age groups were performed using Bonferroni correction. Independent variables that achieved significance were added as covariates to a one-way analysis of covariance by age group to test their effects on the MIDUS and ASD scores. A one-to-one case-controlled ANOVA controlling for covariates that achieved significance was conducted to confirm the effects of these covariates on the ANOVA results. --- Relationship of the independent variables on MIDUS and ASD score A multiple regression analysis was performed to assess the relationship of the independent variables on the MIDUS and ASD scores for the current sample. --- Comparisons with previous studies We calculated effect sizes using Welch's t-test by subtracting the mean MIDUS and ASD scores from the mean MIDUS [17] and ASD [28] scores in the previous studies and dividing the results by the pooled standard deviations of the Mental Illness Stigma in Japanese Employees J Clin Med Res. 2023;15:139-147 respective surveys. We refer to effect sizes as small , medium , and large based on benchmarks suggested by Cohen [31]. Comparisons between past contact experience and our results could not be made because no previous study used the same evaluation scale. Participants with no MIDUS or ASD scores, as well as those with missing data for the key demographic variables were not included in the analyses because these variables were required for weighing. Statistical analysis was performed using the Statistical Package for Social Sciences version 25 for Windows. The level of statistical significance was set at P < 0.05. --- Results --- Demographic characteristics Among the 970 participants, 474 were males, 481 were females, and 15 refused to provide infor-mation on sex. Thus, these 15 participants met the exclusion criteria. Of the 970 participants, 955 had no missing values among the MIDUS score, ASD score, and demographic variables and were eligible for analyses. Among the 955 participants, 40 -49 years old was the largest age group. Regarding employment positions, 217 participants were managers, and 738 were non-managers. For employment status, 511 participants were regular employees, and 444 were non-regular employees. Among the 955 participants, 20 had attended mental illness stigma training. Concerning prior contact experience with PMHC, 665 participants had previous contact experience, and 290 did not. The demographic characteristics of the participants are provided in Table 1. There was no significant difference in the mean MIDUS = 2.91, P = 0.06) and ASD = 0.60, P = 0.55) scores between the companies. --- Effects of demographic variables and age on accurate knowledge The MIDUS scores are listed in Table 1. There were statistically significant differences in MIDUS score by employment status and prior contact experience but ANCOVA comparing age groups was performed with employment status and prior contact experience as covariates. The results showed statistical significance in MIDUS scores by employment status = 5.40, P = 0.03) and prior contact experience = 25.75, P = 0.01). The participants were matched according to age group based on employment status and prior contact experience with PMHC . The resulting four categories were divided by age group as follows: regular employees without prior contact experience , regular employees with prior contact experience , non-regular employees without prior contact experience , and non-regular employees with prior contact experience . For the regular employee without prior contact experience group, the age group of ≥ 60 years was insufficient for comparison purposes. One-way ANO-VAs of these four matched categories showed no significant difference in MIDUS score by age group . --- Effects of demographic variables and age on social distance The ASD scores are listed in Table 1. ANOVA revealed significant differences in ASD score by age group, = 2.54, P = 0.04). The Bonferroni test for multiple comparisons showed that the mean score was significantly different between participants < 30 years old = 3.97) and those 50 -59 years old , suggesting that the younger group had the less social distance concerning PMHC. No significant differences in ASD scores were found by sex, employment position, employment status, attendance at mental illness stigma training, and prior contact experience. --- Relationship of the independent variables on MIDUS and ASD For the multiple linear regression analysis, the MIDUS and ASD scores were entered into the multiple regression model along with possible confounders, including sex , age group , employment position , employment status , attendance at mental illness stigma training , and prior contact experience with PMHC . Employment status and prior contact experience significantly affected the MIDUS score, whereas no variable had a significant effect on the ASD score. --- Comparisons with previous studies Although the current sample size is different from the prior surveys [7,17], the usage of the MIDUS and ASD scales which were employed in the prior surveys [7,17] allow for an indirect comparison in the level of public stigma. The difference in mean MIDUS score between this study and the 2003 survey [17] was statistically significant = -6.07, P < 0.01, d = 0.28). The difference in mean ASD score between this study and the 2007 survey [28] was statistically significant = 3.46, P < 0.01, d = 0.16). Albeit an indirect comparison, these results suggest that the level of accurate knowledge of mental illness is higher in the current sample compared to the 2003 survey. Yet, the social distance level was similar to the 2007 survey. --- Discussion Accurate knowledge of mental illness was significantly higher among regular employees and those with prior contact experience while social distance was significantly lower among participants < 30 years old for the current sample. --- Age and public stigma In our study, although the MIDUS score did not differ significantly by age group, accurate knowledge of mental illness 1), as reported previously [17]. The 2003 Japanese survey, which used the MIDUS score [17], showed that respondents < 30 years old have lower understanding and knowledge of stigma, implying a lower likelihood of interacting with patients or attending lectures. Our results suggest that participants in the older age groups were less socially accepting of PMHC, especially those of 50 -59 years old, compared to those of < 30 years old. This result agrees with a previous report that social distance towards such patients worsens throughout life [32]. Another study [33] also identified a significant effect of age on stigma: older participants tended to be less socially accepting of PMHC and unwilling to have them marry into the family. A study suggested that the increase in social distance with age may be a result of accumulated negative experiences or unpleasant contact experiences with patients throughout life [18]. This is not impossible because negative contact experience reportedly increases stigma [34]. --- Sex and public stigma A prior study [19] found that the association between sex and mental illness stigma varies regionally. In Japan, males tend to have a lesser understanding of mental illness than females [17,19]. Another study [35] found that female has stronger stigmatizing attitudes than male due to their beliefs, suggesting that it varies based on sociodemographic profile. However, we found no significant differences between sex, knowledge, and social distance scores , likely because the population examined may have been derived from a relatively homogeneous population. --- Education and public stigma A prior study found that individuals who attended mental health lectures tended to understand mental illness better than those who did not [17]. However, only 2% of the participants had training related to the stigma of mental illness. This result might be because such training and lectures are still uncommon in Japan. With the implementation of mental health education in high school, the younger generations will be exposed to an accurate knowledge of mental illness at an early age and can be reinforced with more training and lectures, as accurate knowledge of mental illness tends to increase with age. --- Prior contact experience and public stigma Prior contact experience influenced the MIDUS score , indicating that previous contact experience enhances understanding of mental illness. This suggests that prior contact experience may decrease stigma [36], and the lack of contact opportunities may increase stigma, as reported elsewhere [5]. Lack of prior contact experience due to hospitalized PMHC increases stigma among the Japanese [5]. As part of the 2004 Reform Vision for Health, Medical Care, and Men-tal Health Welfare, the Japanese government took measures to reduce long-term hospitalization, such as moving from institutionalized care toward community-based care and improving the quality of mental health care [10]. These ongoing reforms by the government in mental health welfare policy have resulted in a 50-day decrease in the average length of hospital stay [11]. Length of stay among newly admitted patients has also decreased, with 85.7% of patients discharged within 1 year in FY2014 and FY2015 [37]. A shorter average stay means more comprehensive care in the community, and the public may have more opportunities to interact with patients. Compared with the 2003 survey [17], these government initiatives aimed at increasing knowledge and contact opportunities may have worked as intended. --- Employment status and position and public stigma Regular employees had a significantly better understanding of mental illness than non-regular employees . Regular employees are more knowledgeable than non-regular employees because most Japanese companies provide on-the-job training [24], which may include mental health and work stress management, affecting their knowledge and understanding of mental illness. The Ministry of Health, Labor, and Welfare reported that more regular than non-regular employees receive training [24]. As such, there is a need for mental health and stress management for non-regular employees. There was no significant difference in accurate knowledge of mental illness and social distance between managerial and non-managerial employees. Therefore, being a managerial or non-managerial employee did not affect the individual's understanding of mental illness and social distance in this study. Being a workplace manager or a person responsible for personnel matters is reportedly associated with better knowledge of mental illness [20], but another study yielded disparate results [38]. A company's human and material resources may influence managers' understanding and behavior toward PMHC [39]. The results of this study differ from those of prior works, warranting further investigation. --- Relationship to intervention policies Over the past 20 years, the Japanese government has taken measures to improve knowledge of the mental illness. One of the most noticeable changes was the 2002 change of the Japanese term for schizophrenia [8] from Seishin-Bunretsu-Byo to Togo-Shitcho-Sho . In a survey of psychiatrists after the renaming, 82% regarded the new term as more suitable for obtaining consent from patients and effective in reducing stigma [8]. The results suggest that citizens' impressions of schizophrenia changed after the name change, leading to greater understanding and more willingness to disclose their illness. In response to overwork-related disorders and mental disorders due to overwork, the government passed the Act on Promotion of Preventive Measures against Karoshi and Other Overwork-Related Health Disorders in 2014 to improve mental health in the workplace and to promote a healthy work-life balance [12]. The Act promotes public awareness of overwork-related disorders by, for example, designating Labor Thanksgiving Day in Japan, which may have indirectly improved the public's knowledge of mental health [12]. Using the exact same MIDUS scale to the prior survey [17], we found that the level of accurate knowledge of mental illness in the current sample is significantly higher than the prior survey, mildly suggesting that the level of accurate knowledge of mental illness may have improved compared to the past due to the efforts by the Japanese government. However, the level of social distance toward PMHC was similar to the past survey [7]. Compared to other countries, a study in the United States found that social distance towards depression improved compared to the past [33]. A study in England also found improvement in behavior towards PMHC compared to the past, attributing success to the effectiveness of Time to Change campaign which lasted from 2009 to 2017 [40]. On the other hand, a study in Germany [41] found that social distance has not improved despite having national antistigma initiatives at reducing stigma on mental health conditions. Although the MIDUS score has a weak positive correlation with the ASD score in this study = 0.18, P = 0.01), our results confirmed the conclusions of stigma experts that behavior change in response to stigma is a complex process and that knowledge alone is unlikely to reduce social distance substantially [42,43]. In addition, other than the Japanese government's renaming schizophrenia in 2002 [8], none of the mental health initiatives have directly targeted mental illness stigma. This lack of direct efforts to address stigma at the national level [44] may explain the absence of improvement in the social distance concerning mental illness. Therefore, governments should adopt direct initiatives and interventions targeting the stigma of mental illness beyond office workers and high-school students. Interventions using sophisticated techniques, such as perspective taking and empathy induction using modern technology, may also enhance public understanding of the suffering of PMHC and promote pro-social behavior toward them . --- Limitations This study has a few limitations. The population differed from those in the prior surveys [7,17,28]; thus, a direct comparison was not possible and an indirect comparison using effect size estimation was conducted with Welch's t-test. Although respondents were informed that the survey was anonymous [45], the results were self-reported, and the possibility of social desirability bias cannot be excluded, such that the actual level of stigma might be more severe than the results of the survey. The participants were recruited from companies in the same city, resulting in a risk of homogeneous sampling. The proportion of respondents who attended mental illness stigma training was significantly skewed. The vignette used for ASD were based on depression and the level of social distance may not be a good representative of mental illness in general as different conditions may lead to different level of social distance [33]. Other possible factors which may have affected the stigma score such as per-sonal belief or cultures [46,47] were not surveyed in this study, which warrant for further investigation. --- Conclusions Accurate knowledge of mental illness was significantly higher among regular employees and those with prior contact experience. Social distance was significantly lower among participants < 30 years old. --- Data Availability The datasets generated and/or analyzed in this study are available from the corresponding author upon reasonable request. --- Conflict of Interest The authors declare that they have no competing interest. --- Informed Consent Informed consent was obtained from all participants. --- --- Abbreviations

The public stigma related to mental illness is the general public's negative misconceptions about people with mental health conditions (PMHCs). The public stigma of mental illness is detrimental to PMHC as it leads to loss of opportunities and unemployment. The aim of the study was to clarify the status of public stigma related to PMHC, focusing on knowledge about mental illness and social distance concerning PMHC.A survey was conducted among 970 Japanese office workers aged 20 to 60 years. Accurate knowledge of mental illness was assessed using the Mental Illness and Disorder Understanding Scale (MIDUS). The Attitudinal Social Distance (ASD) was used to determine social distance in relation to PMHC. The demographic characteristics of the participants evaluated were sex, age group measured in years, employment position, employment status, and attendance at mental illness stigma training. Results: Regular employees (P = 0.03) and those having prior contact experience (P = 0.01) had more accurate knowledge. Participants between 50 to 59 years old (M = 15.87, standard deviation (SD) = 3.35) had greater social distance than those under 30 years old (M = 14.78, SD = 3.97, P < 0.05). The results of multiple linear regression analysis found that employment status (partial r = -0.07, P < 0.05) and prior contact experience (partial r = -0.15, P < 0.01) significantly affected the MIDUS score, whereas no variable had a significant effect on the ASD score. Conclusions: Accurate knowledge of mental illness was significantly higher among regular employees and those with contact experience. Social distance was significantly lower among those under the age of 30 years.

introduction The Slovene Roma are usually classified as members of the lowest social classes. Even though their social status varies from region to region, it is still significantly lower than that of the rest of the population, which also affects, among other things, their low level of education . Roma pupils' academic success depends on many factors, i.e., their linguistic and cultural differences, teaching content that does not take into account the characteristics of the Roma community, poor pre-existing knowledge before entering school, differing behavioural and thought patterns, different habits , an undisputed environment, poor work habits , inadequate living conditions for homework assignments, uneducated parents, pupils skipping and leaving class , the absence of trained professionals who have experience in working with Roma children, low expectations regarding their school performance, lack of confidence in the school system, and Roma parents' poor involvement in the educational process and their lack of cooperation with the school . Pupils' performance and academic success are also significantly influenced by the school itself, especially its dominant culture, the school atmosphere, and interpersonal relationships. School is not only a place for obtaining knowledge, but also a space in which pupils can encounter other individuals with whom to shape important relationships. Numerous studies have confirmed that pupils' relationships with their peers and teachers are significantly linked to the well-being and acceptance of the pupils into the school environment, their learning performance, and their academic success. Vonta and Jager state that schools cannot resolve all the factors that affect the educational success of Roma pupils, but that they can change or have a significant impact on at least some of them, thereby increasing the pupils' chances for academic success. Based on this, we decided to explore how Roma pupils and their mothers from Roma settlements in Prekmurje and Dolenjska perceive the pedagogical work and attitude of teachers towards Roma pupils in Slovene elementary schools and whether their attitudes and relationships with the children contribute to the poor academic performance of Roma pupils. We wished to determine whether the teachers' attitudes towards Roma pupils were encouraging, supportive, understanding, and non-discriminatory. Through this study, we aimed to give a voice and an opportunity to the members of the Roma community and thereby raise awareness of the current challenges in the primary school education system. The article consists of a theoretical section and an empirical section. The first focuses on a fair and inclusive school that considers the varying needs of its pupils and thereby offers quality education for all. We considered social acceptance, respect, and a sense of belonging as important indicators of social inclusion and emphasized that with the realization of all cultures within the school community we could help to create a tolerant coexistence. At the same time, this type of school community could help the pupils that belong to vulnerable groups to have different and more pleasant interpersonal experiences than in wider society. Next, we focused on the importance of the teacher's willingness to establish good interpersonal relationships with their pupils, which significantly influences the involvement and activity of those pupils in the classroom. We also introduced the results of our study to confirm that positive relationships between teachers and pupils are not only related to the well-being of the pupils at school, but also to their learning success and behaviour. In the empirical section, however, we will introduce the results of our qualitative study that focused on Roma pupils' and mothers' perceptions of teachers' pedagogical work and their attitudes towards Roma pupils. We aimed to determine whether their relationships affected the inclusive orientation of the studied elementary schools and whether the teachers' attitudes towards Roma pupils were an important factor that contributed to the lesser academic success of Roma pupils. --- creating a Fair and inclusive School Pupils enter school with varying prior knowledge, experiences, skills, abilities, interests, as well as wishes and plans for the future. Thus, it is the school's job to take into consideration all of the children's unique traits and desires and offer them a fair, successful, satisfying, and quality education . In modern democratic societies, fair education is one of the most important characteristics of national education systems. When we talk about justice within the educational system, we also refer to social justice in general . A fair school, according to Peček and Lesar , is one in which the processes and activities are equally accessible to all pupils. In this case, it is allowed or even necessary for underprivileged and privileged pupils not to be treated equally, so that the underprivileged can also benefit from the learning process. Among the latter, Rawls mentions those with lower social status and the less naturally gifted. It is important to add that the privileged are not supposed to be punished because of their initial advantages, which are neither fair nor unjust. Only the ways in which the fundamental social institutions deal with the pupils' basic differences can be considered fair or unfair. In the past, Slovene primary schools had to deal with various cases of group exclusion, which confirm that the school system was unfair, as these groups did not have the opportunity to develop to their full potential. According to Lesar these groups included pupils with special needs and ethnic minorities . In contrast, it is necessary to advocate for an inclusive school, which tends to "reduce all exclusionary procedures and devaluation of pupils based on disabilities, race, gender, age, ethnicity, sexual orientation, religion and other factors that make school life even more difficult" . Inclusive schools are meant to introduce a series of adjustments that take into account the unique and varying needs of its pupils. Lesar highlights social acceptance, respect, and a sense of belonging as important indicators of inclusion. Namely, inclusive schools are supposed to create a culture that does not give priority to the values of certain dominant groups, but rather focuses on including all the cultures of its pupils and promotes tolerant coexistence. At the same time, Lesar highlights that teachers and schools, the latter embedded in the socio-economic and cultural relations within an individual society, can help their pupils have more positive experiences in society. With their knowledge and engagement, as well as the awareness that they can influence their pupils, teachers can prevent the migration of existing relationships from wider society to the school environment as well as stop enlarging the differences between individuals, especially between the nation's members of the majority and the members of marginalized groups. --- The importance of establishing Good Relationships between Teachers and Pupils Teacher-pupil relationships can significantly influence pupils' development and their success in various areas of life . According to the PISA survey , positive and constructive relationships between teachers and pupils are key to promoting the pupils' social and emotional wellbeing. The results of this survey showed that, on average, in OECD countries, pupils who reported having good relationships with their teachers also made friends more easily, were more satisfied with their school, and had a stronger sense of belonging. A study by Chiu et al. also confirmed that teacher-pupil relationships have a significant impact on pupils' well-being and their sense of belonging to their class or school. The same study demonstrated that pupils' sense of belonging to the school depends not only on relationships with their peers but also on those with their teachers. In another study, Uslu and Gizir concluded that relationships between teachers and pupils can strongly influence pupils' sense of belonging and feeling that they are a part of the school. Other authors have even stated that pupils' well-being at school depends on their relationships with their teachers and peers. Hamre and Pianta similarly determined that having a positive relationship with teachers is crucial, as they can help chil-dren feel more comfortable at school, build more meaningful relationships with peers, and have a greater sense of safety. Furrer and Skinner pointed out that a teacher's relationship with their pupils is not only related to the pupils' well-being in the classroom, but also to their perception of learning. In this study, pupils who felt valued by their teachers were more likely to report feeling happy in the classroom. They even found school activities more interesting and fun. On the other hand, pupils who felt unimportant or ignored by their teachers reported to be bored and angry when they had to participate in learning activities. Based on previous research, Juvonen pointed out that relationships between teachers and pupils that are based on support and characterized by a low level of dependence and conflict make it much easier for children to function at school. Such relationships are especially important for pupils who lack social support in their home environment. Moritz Rudasill et al. came to similar conclusions, confirming that high-quality relationships between teachers and pupils have a positive effect on children, especially on pupils who come from less stimulating environments. Hamre and Pianta determined that negative attitudes reflected in conflicts and dependence on teachers can be an important predictor of pupils' grades and achievements on standardized knowledge tests. They can also predict the work habits of pupils in elementary schools. Cornelius-White studied the relational characteristics of teachers in connection with pupils' achievements. He found that the associations of these variables with cognitive variables were above average, especially critical/creative thinking and academic achievements, while associations with IQ, grades, and perceived achievements were at a medium level. Associations with scientific achievement and performance in the field of social sciences, however, were very low. Another study also confirmed that teacher-pupil relationships significantly impact pupils' academic achievements. McCormick and O'Connor stated that conflicts between teachers and pupils are more frequent at schools associated with lower reading achievement of elementary school pupils. We can also draw conclusions about the connection between teacher support and pupils' academic achievements from another Slovene study conducted by Puklek Levpušček and Zupančič . The authors found that younger adolescents who perceive their teachers positively express more positive motivational beliefs and achieve better results in mathematics. Pupils who rated their mathematics teachers as responsive and helpful, and who evaluated their work positively, were more successful at mathematics compared to their classmates with negative teacher experiences . Research further showed a connection between teacher-pupil relationships and pupil behaviour. O'Connor et al. determined that quality relationships between teachers and pupils predict fewer behavioural problems and help prevent behavioural issues in the later stages of pupils' development. Obsuth et al. concluded that a positive relationship between teachers and pupils serves as a protective factor against behavioural problems in adolescents. Children aged 10 who reported having better relationships with their teachers were less likely to be involved in delinquent behaviour later on at the age of 12, 15 and 17. Similarly, Archambault et al. ) found that pupils who had close relationships with their teachers reported greater behavioural adjustments compared to those who had less positive relationships with their teachers. Conflicting relationships between teachers and pupils have proved to be especially harmful for boys. Research has also confirmed that relationships between teachers and pupils are strongly related to the occurrence of peer violence. Thus, Marengo et al. ) found that pupils who were victims of violence or bullies perceived their relationships with teachers as more conflictual compared to other pupils who were not involved in any form of violent behaviour. Pupils who were both bullies and victims reported having more conflictual relationships with teachers than those who were just bullies. On the other hand, pupils who were only victims of violence reported having a similarly conflictual relationship with their teachers as the pupils who were both victims and bullies. The latter turned out to be the group that had the worst relationship with their teachers. The research presented above shows a significant connection between teacher-pupil relationships and the pupils' academic success, behaviour, and well-being at school. These findings further point to the importance of good and healthy teacher-pupil relationships. Peček and Munda investigated the attitude of teachers towards Roma pupils on a sample of 77 Roma elementary school pupils from Maribor. The results of their study showed that the majority of pupils who were included in the study had positive experiences with their teachers, while 4 pupils gave negative descriptions of their teachers and mentioned cases of discrimination. Of the pupils,14 gave a mixed description of their teachers. However, we should not forget that the study also confirmed that teachers are not fully aware of the importance of their influence on the learning performance of Roma pupils . --- empirical Research In the following section, we present the results of our empirical research, in which we tried to find out how Roma pupils and their mothers perceive teachers' pedagogical work and their attitudes towards Roma pupils. The paper introduces some of the results of a more extensive empirical study, which was conducted as the basis of a master's thesis at the Department of Pedagogy at the Faculty of Arts of the University of Maribor . --- Research Method and Sample We used an empirical qualitative research method and included a non-random occasional sample of 20 Roma pupils in the third educational period of primary school in the 2019/2020 school year. We decided to conduct the interviews in Prekmurje and Dolenjska because these two regions are known to have the largest number of community members with the most noticeable social differences . The interviews in Prekmurje were arranged directly with the interviewed subjects, most of whom, just like the first author of the article, originate from the Roma settlement in Dolga vas and have children in the third educational period of primary school. Some interviews, however, were conducted with the residents of another Roma settlement in Prekmurje -the interviewees were not the author's acquaintances but were recommended by them. For the interviews in Dolenjska, we contacted the Roma activator employed at the Novo Mesto Developmental Education Centre, who made appointments with the interviewees and accompanied us to the Roma settlements during the interviews. These were conducted in the months after the distance education period, right after the first wave of the COVID-19 pandemic in Slovenia. The pupils had just returned to school and had only spent a month back in the classroom. We interviewed 10 Roma pupils from Prekmurje and 10 from Dolenjska, of which 12 were girls and 8 were boys. One of the interviewed pupils , had failed that year, nine had finished school with a 2, sufficient knowledge , seven pupils had finished with a 3, good knowledge , and two girls had finished with a 2, very good knowledge . One of the interviewed pupils from Prekmurje could not provide information on her grades. The interviewed Roma pupils from Prekmurje lived in two Roma settlements and attended four different primary schools, while the pupils in Dolenjska, who also came from two separate Roma settlements, attended three different primary schools. Besides Roma pupils, the study also included the mothers of Roma children. Of these, ten were from Dolenjska and ten from Prekmurje. To be precise, our sample included 16 mothers of the interviewed Roma children and four other mothers, whose children we did not interview. On average, each of the interviewed mothers had 3.2 children. Ten of them had not completed primary school, seven had a primary school education, and three a secondary vocational education. Of the mothers, eleven had work experience, the remaining nine had never been employed, nor had they taken on any occasional jobs. --- Research instruments and Data collection Procedures We collected our data from a semi-structured interview with predefined basic openended questions, which, in some cases, were followed by sub-questions that arose naturally during the interviews. The data was collected between July 2020 and August 2020, which is when the interviews in the Roma settlements in Prekmurje and in Dolenjska were conducted. In Prekmurje, the majority of the participants were interviewed on the premises of the Multipurpose Roma Centre Dolinsko-Dolga vas, while the rest of the interviews were conducted in the subjects' homes. All interviews in Prekmurje were conducted in a calm environment without any disruptions. In Dolenjska, the subjects were interviewed at home, mostly in their backyards, some of which were located next to the road or in family yards and louder environments where it was not possible to ensure enough peace and quiet. Some of the interviews were disrupted by small children. The interviews were conducted by the first author of the article, who explained the purpose of the study to all the interviewed subjects and asked them to answer the questions honestly. In Dolenjska, where the author did not know the interviewees, she also had to introduce herself and inform them of her origin. During the relaxed informal gatherings in Dolenjska, we had the opportunity to get to know the subjects and explain the purpose of our visit. Nevertheless, some of the mothers were slightly tense and less relaxed during the conversations, since they were not actively involved or included in their children's educational process. We also encountered quite a few organizational obstacles, as we had to wait for or even go in search of the interviewees who forgot about our appointment. We did not encounter such problems in Prekmurje. All interviews were conducted individually and in private. The conversations during the interviews were held in the Slovene language. Before each interview, the parents of the interviewed children had to sign a written consent form in which they were informed about the purpose and course of the study. By signing this form, they also confirmed that their children would participate in the interviews in accordance with the ethical principles of voluntariness and that their answers would remain anonymous. This consent also gave them the right to change their mind and withdraw from the study at any time without any consequences. All interviewed subjects consented to being audio recorded. The average interview lasted about 20 minutes, the shortest 12, and the longest 32 minutes. --- Data Processing Procedures The audio recordings were transcribed into written form and changed into Word format after repeated listening. In order to analyse the results for the purpose of this article, we separated the interviewees into two groups and marked them with consecutive interview numbers . After the transcription was done, we listened to the tracks again to make sure that everything had been transcribed correctly. To process the collected data, we first carried out a qualitative content analysis and then marked the parts of the transcribed text which we thought were relevant to and important for our research. Next, the text was broken down into its smallest meaningful components, thus obtaining coding units with information relevant to the purpose of this study. Following this, we gave code names to the latter coding units and divided them into categories based on similarity. During the last step, we sorted the categories according to their meaning and determined the central themes of the study . --- Results and interpretation Following this, we introduced the way in which pupils and their mothers perceive the attitude of teachers towards Roma pupils. In the context of the previous chapter, we were initially interested in how Roma pupils perceive their teachers, and what they think about their pedagogical work and knowledge assessment. In the second part of the introduced results, we focused on how Roma mothers perceive the teachers' pedagogical work and attitudes towards Roma children. --- The Pupils' Perception of the Teachers' Attitudes towards Roma Pupils First, we presented the teachers' attitudes towards Roma pupils from the perspective of Roma pupils, which, as seen in Table 1, were divided into two categories. --- Perception of the Teachers' Attitudes towards Roma Pupils The Roma pupils mostly described their teachers in a positive way. They stated that they are kind, helpful, and really make an effort. One of the pupils said: "They were friendly and liked to help. If you did not know things, they tried to help, so you could finish the class with a positive grade" . The pupils also stated that the teachers explain things well, are very understanding, and are open to negotiation and common agreements. Some of the pupils pointed out that the positive attitude of the teachers is the main reason why they like coming to school. Some of the pupils stated that not all teachers are the same -some are friendly, while others are not. This was also confirmed by the following statement make by one of the interviewed pupils: "The best teacher is the math teacher, but the class teacher and the physical education teacher are also very good. They are very nice, they listen to what we want to say, while the others just shout at us" . The same pupil used the Slovene teacher as a negative example. In his opinion, the Slovene teacher is a bad teacher, because she screams all the time. However, he did not say that she only screams at Roma pupils. Similarly, other Roma pupils, who gave a negative assessment of their teachers, stated that teachers often yell at pupils, but none of them said that this only applies to Roma pupils. Most of the Roma pupils said that their teachers treat them the same as they treat their non-Roma schoolmates and that sometimes they even defend Roma children in front of other pupils. The following is the statement by one of the pupils that confirms this: "At this school the teachers stood up for us Roma pupils and told the other children that we are the same as anybody else. They even told them not to behave like this [note: be discriminatory]" . The interviewed pupils stated that the teachers' attitudes towards Roma pupils depends on the Roma pupils' attitude towards the teachers and other children in the class. When asked if the teachers were equally kind to all students, one of the pupils replied: "They are equally nice to everyone, however, if the Roma pupils are not nice to the teachers or their peers the teachers react in the same way" . The interviewed pupils also believe that the attitude of the teachers towards Roma pupils depends on the actual effort made by the Roma children. One of the interviewed pupils even said: "I consider the teachers nice. They are nice to me [note: teachers], because I am trying harder than the others" . The opinion that the teachers' attitudes towards Roma pupils depends mainly on the Roma pupils' attitude towards their teachers was more common among the interviewed subjects from Dolenjska. They admitted that Roma pupils often disrupt class, which then affects the teachers' attitude towards them. The interviewed pupils had negative experiences and admitted to having discriminatory teachers as well. This can be confirmed by the following answer given by one of the pupils: We were doing something, I don't remember what exactly, and she [note: teacher] told me that I don't have the same rights as the other children. She kept telling me I will never be like my mother because I am lazy. She was also rude to others, but she told me that I don't have the same rights as others. She said that, because I am a Roma. Another pupil described a discriminatory experience with a teacher as follows: "There was this one time when me and him [note: a classmate] quarrelled. I was yelled at, he was not, because he has problems with his nerves" . At the same time, the pupil pointed out that this does not happen very often. --- Perception of the Roma Pupils' Knowledge Assessment Most pupils believed the teachers were fair in assessing their knowledge, as they always got the grade they deserved, sometimes even a higher one. One of pupils stated: "I always got the grade I deserved, sometimes an even higher one" . Some of the interviewed pupils pointed out that teachers are unfair and strict in assessing their knowledge and that they give unjustified negative grades. One of them said: "Yes, the history teacher is like that. Some pupils don't do what they are asked during class, and she starts screaming and then gives them a negative grade. She should first warn them before giving out negative grades" . Some of the interviewed pupils also thought that teachers, when assessing knowledge, are more considerate of the academically more successful students, which was confirmed by the following answer: "Yes, for example the […] teacher […]. When writing a test, I was only missing one single point for a positive grade -just one point -and he immediately gave me a 1 . To others, who were one point short, he gave a 3 , not a 2 " . When asked why she thought she didn't get that extra point like the others, she said: "Because I was apparently stupider" . When the interviewed pupils talked about why their teachers were unjust, none of them, with the exception of one student from Dolenjska, stated that the reason for it was their ethnicity. Some of the answers regarding the teachers' assessment of knowledge insinuated that teachers treat Roma pupils differently. The latter is clearly highlighted in the following answer: "If we know everything, they still do not give us an 5 . If I know everything, they only give me a 2 , while the civilians [note: students from the majority population], even if they know a bit, they get an 5 or a 4 " . During the interviews, Roma pupils repeatedly pointed out that teachers adapt the teaching process to their capabilities. They explained that during exams and oral assessments of knowledge, teachers adjust the tasks or questions, so that they can get a positive grade, or that they only ask them questions or give tasks for a positive grade. One of the pupils summed up her thoughts as follows: "In some subjects, not everyone is treated the same way" and emphasized that the teachers ask her easier questions or give her less challenging assignments. The pupils from Dolenjska talked about the adaptation of tasks and assessment, while the pupils from Prekmurje did not mention any of that. --- The Teachers' Attitudes Towards Roma Pupils from the Perspective of Roma Mothers This topic was divided into two categories, both of which are presented in Table 2. Source: Own Data --- Perception of the Teachers' Attitudes towards Roma Pupils The interviewed mothers perceived the attitudes of the primary school teachers towards their children as friendly and equal. They evaluated that the teachers' attitudes towards Roma pupils depends mainly on the children's attitude towards the teachers and their peers, as well as on the work and effort the Roma pupils are willing to put into their school obligations. One of the interviewed mothers stood out with her statement, as she said that teachers go easy on Roma pupils, continuing with: "[…] as far as I can tell, it is harder and stricter for those other kids, the Slovene ones" . The Roma mothers referred to their kids' teachers in a rather positive way. They believed that the teachers are friendly, helpful, and always available if the children or parents need them. We detected a slightly more positive description of the teachers in Dolenjska, some of which even conducted home visits, especially during school closures due to the COVID-19 pandemic. One of the interviewed Roma mothers, who does not attend school meetings because she communicates with the teachers over the phone or through a Roma assistant, described the teachers as follows: "As far as I can see -they have come here a couple of times during this corona period -they are very friendly, also the class teacher. I have nothing bad to say about them. They are normal, they help if we need them" . When Roma mothers described the teachers' attitudes negatively, they mostly referred to unequal treatment of the pupils, but the reason behind it was not always ethnicity. Some of the interviewed mothers, however, warned that the teachers' attitudes towards Roma pupils is discriminatory. They believed that teachers are not fair to Roma children, especially in conflicts with non-Roma children. One of the mothers shed light on the problem, saying: "Well, maybe sometimes there are conflicts. If they fight, it's almost certain that the Roma child will be to blame" . Another mother shared examples of verbal abuse -a teacher insulted her daughter by saying she is a "Gypsy" . Another interviewed subject even mentioned a case of physical violence: Yes, the class teacher. He threw chalk at him. […]. They were loud in the class, and he threw a piece of chalk at him. Then he and I [note: husband] went, but he did not tell us [note: son]. The other children told us that he threw chalk at him, so we went to school and had a meeting. He [note: husband] warned that this can never happen again. We had peace for a while, but after two months the teacher hit him on the head again. From the statements of the interviewed mother, it is hard to tell whether the teacher's actions were committed because of the child's ethnicity or for other reasons. The same mother also said that the teacher probably hit the other childher son's classmate -who was also involved in the same conflict. --- Roma Pupils' Knowledge Assessment from the Perspective of Roma Mothers The Roma mothers pointed out that their children's teachers are fair when it comes to assessing knowledge and added that teachers used to be more unfair when they themselves attended elementary school. One of the interviewed mothers explained the situation as follows: Because when I was at school, we were learning the same stuff, well actually we weren't. They gave us something else to learn. But even if we knew a lot, they would only give us a D. Now, it is different. When school started, and the teachers came to talk to us, Four Roma mothers from Prekmurje perceived their children's knowledge assessment as unfair, two of them even pointed out that the reason behind it is the children's ethnicity. One of the mothers stated: "I think they are treating them differently, the Roma children and those others. Some deserve it [note: a bad grade], but the Slovene children get fewer bad grades than the Roma children" . The other two mothers said that it is the teachers from smaller schools who are problematic. They explained that the teachers and parents from smaller schools know each other and as a result the teachers prefer the children of certain parents. One of the interviewed mothers added: Acquaintances are a problem. Some come from the same village, live nearby and know each other. He will deserve a B, make mistakes, and still get an A. Some [note: children who's parents do not know the teachers privately, outside of school] try really hard and have it [note: knowledge], yet the teachers will not acknowledge it -they will even give them a lower one [note: grade]. There are huge differences, but that's acquaintances. Another critical description came from a mother who stated that the teachers are not experts in their field, do not take into consideration the differences between pupils or know how to adjust their work to kids with special needs: Ugh, the teachers at this school are really terrible. Well, not to be rude to all of them; there is one young teacher who is new, and she is great, but the others, especially the older ones -they don't behave like teachers at all. You know that my older one [note: son] got a written confirmation, he is dyslexic. Still, they always bothered him with handwriting and were telling him he should write bigger or smaller. When he received the official diagnosis, they still didn't take this into account, even though everything was official and written on paper. He could have taken the tests orally, that was his right, but they did not care. And it is not just my son […]. They don't understand that pupils are different, mine can't be like the others, and the others aren't all good either. They don't know how to adapt to anyone. On the contrary, some of the interviewed children and mothers from Dolenjska pointed out that the teachers adapt their teaching methods and content to Roma pupils, which means that they prepare easier questions and tasks for them. All of these mothers, with one exception, perceived this as a positive thing. The mother that was critical on this matter pointed out that this is not correct and that Roma pupils should be treated the same way as any other pupil in the class: --- Discussion The purpose of this article was to investigate how Roma pupils and their mothers perceive teachers' pedagogical work and attitudes towards Roma pupils. Namely, the teacher-pupil relationship can have a positive impact on the pupils' well-being, sense of acceptance, and academic performance. In the case of Roma pupils, this relationship is even more important, since Roma are considered one of the most marginalized groups in Slovenia. This group faces many challenges both in the educational system and in other social realms. Although school is embedded in the socio-economic-cultural side of society, explains Lesar , it can give pupils a different, perhaps more positive experience than what they are experiencing outside of school in wider society. We use the term positive experiences to encompass interpersonal relationships, social acceptance, respect, and a sense of belonging to the school, which are all important indicators of an inclusive school. Based on the obtained results, we could not connect the poor performance of Roma pupils with bad teacher-pupil relationships. At least in our study, the latter did not contribute to poorer academic performance in Roma pupilsnamely, the interviewed pupils rated their relationships with their teachers as quite good. Therefore, we believe that in the future, our research should focus on other potential factors that may have an impact on lower academic performance of Roma pupils. These other factors were discussed in the introductory part of the article. The results of our study, collected from the answers of Roma pupils and mothers, showed that Roma pupils from Prekmurje and Dolenjska generally perceive their teachers as kind and fair. They even pointed out that the teachers are helpful and good at their job -explaining and passing on knowledge. Good social relations with teachers proved to be important to Roma pupils. Their teachers, according to the pupils' statements, are the reason why they like attending school. Both the pupils and mothers assessed that the attitudes of teachers towards Roma pupils often reflects the attitude of Roma pupils towards teachers. Another study that refers to Roma elementary school pupils in Grosuplje, which was conducted by Koščak , showed similar results and confirmed that Roma pupils look for shelter and attention in their teachers, which is primarily a result of non-acceptance by their non-Roma peers. Based on this, we can conclude that the Roma pupils included in the study feel accepted and respected by their teachers. Both the interviewed pupils and their mothers stated that teachers in primary schools treat Roma pupils the same as other pupils, which is definitely encouraging and confirms the inclusive attitude of primary schools. According to one of the interviewed subjects, teachers are warning non-Roma pupils not to discriminate against their Roma peers. However, the personal attitudes of the teachers towards Roma pupils sets an example for other non-Roma pupils and has a significant influence on the extent of the Roma pupils' being accepted by their non-Roma peers. The relationship between teachers and pupils is therefore essential for the social inclusion of pupils in the classroom . Even though Roma pupils and their mothers perceive teachers mainly in a positive way, the study also revealed a few cases which are less encouraging. The teachers' negative assessments pointed towards an unfair assessment of knowledge and an unfair attitude towards the children, however, there was no confirmation of this attitude deriving from ethnic discrimination towards Roma pupils. The described examples and issues with the teachers could also apply to non-Roma pupils. Only one interviewed pupil from Dolenjska referred to teachers as being ethnically discriminatory when assessing knowledge and grading Roma pupils. In comparison to the interviewed pupils, more Roma mothers gave negative assessments of the teachers. The mothers from Prekmurje also gave more negative teacher assessments than the mothers from Dolenjska. At this stage, we should point out that in most cases the negative teacher assessments did not stem from discrimination against Roma pupils. The Roma mothers from Dolenjska stated that teachers adjust schoolwork for their children by giving them simpler tasks or evaluating them on the basis of pre-prepared questions and answers, which enable them to achieve the lowest positive grade. Some mothers perceived this as a discriminatory act -this way, their children, unlike non-Roma pupils, are not given the opportunity to attain a higher grade. Such adjustments for Roma pupils show that teachers demand less from them compared to other pupils. The latter was also confirmed in other, previously conducted studies . This kind of attitude towards Roma pupils is problematic, not only because it is discriminatory, but also because the teachers' low expectations lead to the Roma pupils' low academic results . At the same time, it shows disrespect towards Roma pupils, who are being deprived of the same opportunities, knowledge, and progress other pupils are granted. The unfair teaching methods and knowledge assessment could be related to the ingrained stereotypes about the Roma, which e.g., say that they are not interested in education. This stereotype probably derives from a lack of knowledge regarding the home environment of Roma children. Bešter and Medvešek found that most teachers in Slovene primary schools try to encourage Roma children in their educational process and help them achieve the best possible results, however, they still lack knowledge and understanding of inter-cultural differences. At the same time, the teachers do not go beyond an ethnocentric attitude and worldview, and they are also not too engaged in overcoming the existing social relationships of inequality between the majority population and minority groups. The authors pointed out that there are a few exceptions among teachers, who have well-developed intercultural competences. These result from training, openness, and the teachers' individual efforts, rather than from systemic incentives and support, which teachers could really use. We believe that all the aforementioned difficulties could be at least partially omitted if the educational programs of future pedagogical workers as well as the programs for continuous professional development included more inclusive and Roma-related topics. Therefore, it would be wise to assess the curricula of university programs that train future teachers and if necessary enrich them with additional content on the typical characteristics of Roma culture and Roma communities in Slovenia, as well as national and other documents related to the Roma community and the education of Roma pupils. The universities that educate future teachers could connect their students with representatives of the Roma community who could directly introduce them to Roma culture and the Roma community itself or give the students an opportunity to visit Roma settlements and explain to them the specifics of living in such communities. These are just a few ideas with which students and future teachers could become familiar with the specifics of Roma culture in the course of their studies, get a better understanding of Roma pupils, gather knowledge on how to approach Roma pupils properly, and learn how to introduce Roma pupils' nature and characteristics to other non-Roma pupils in the class and at school. The latter is extremely important in light of the reasoning pointed out by Vonta and Jager who said that professionals in Slovene elementary schools seem to be aware of the importance of including Roma culture in the educational process, but that they are currently pursuing this step only by encouraging Roma pupils to perform and introduce their culture at school events. To fully understand and properly interpret the results, we have taken into consideration that the interviews were conducted during the first wave of the COVID-19 pandemic in Slovenia, in the months after home schooling, when the pupils had only spent a month back in the classroom. After the long period of home schooling, most pupils could not wait to go back to school, which of course could have affected their perception regarding the teachers' attitude towards them. The long-term distance education is also likely to have been reflected in the children's schoolwork. Bešter and Pirc investigated this topic further and wanted to determine how distance education affected Roma pupils in Slovenia during the quarantine period. Based on the data they collected from Roma assistants and teachers, the authors of the study concluded that approximately half of Roma primary school pupils were in regular contact with their teachers during quarantine, while the proportion of non-Roma students who were in regular contact with teachers was much higher. Almost half of the Roma pupils in primary schools maintained contact with the school through Roma assistants. The same survey showed that in schools with Roma assistants, a quarter of all interviewed teachers estimated that during the quarantine period most Roma parents were not in contact with the school at all, while another quarter of teachers pointed out that the parents were in regular contact with the Roma assistants, but not directly with the teachers. Only a fifth of the interviewed teachers stated that Roma parents were in contact with them directly. Most of the teachers in schools with Roma assistants also reported that cooperation between Roma parents and the school remained the same as before the pandemic, slightly more than a tenth noticed an improvement, and about a fifth a deterioration in this area. In schools without Roma assistants, the situation was pretty much the same. The authors also determined that some teachers made home visits and brought Roma pupils class materials as well as gave instructions for schoolwork. In our study, two interviewed mothers from Dolenjska also reported having home visits from teachers. One of these two mothers had no contact with the teachers prior to the pandemic, but during the home-schooling period, thanks to home visits, she came to regard the teachers in a very positive way. We must also point out some of the limitations we faced during the study. One of the major limitations were the interviews in Dolenjska, where it was not always possible to ensure a calm and relaxed environment. The interviews were often interrupted by the interviewed mothers' young children or by the brothers and sisters of the interviewed children, as well as by the noise coming from outside. We could have avoided this issue if the interviews had been conducted in a common space within the Roma settlements , but the interviewed subjects wanted to talk to us in their homes, so we did as they asked. Some of the interviewed mothers in Dolenjska were very nervous -we even detected slight mistrust, which may have influenced their answers. Even though the results do not apply to the entire population of Roma pupils and their mothers in Slovenia, they still present a valuable insight into the experiences of Roma pupils with teachers and their pedagogical work. The extra value of this research lies in giving voice to the ignored Roma community. By including Roma pupils and their mothers in this study and giving them the opportunity to speak up and share their perception of teacher-pupil relationships and the pedagogical work of the teachers, we can better understand the communities' experiences and perspectives. The findings are also important for raising awareness among teachers and school management not only about the impact of teacher-pupil relationships on the pupils' well-being and academic performance, but also about the fact that all pupils, regardless of nationality or special needs, deserve equal opportunities. The results can potentially contribute to a more subtle perception and understanding of the well-being and the situation of Roma children in Slovene schools and, as a result, help raise the quality of pedagogical practice in the field of Roma education in Slovenia. The study also opened up a number of new research questions. Namely, the interviews were conducted during the COVID-19 pandemic, which could have affected the results, so it would be wise to repeat the entire procedure in the upcoming school year. In the study, we focused on two locations -Dolenjska and Prekmurje -however, for an even better understanding of the relationship between teachers and Roma pupils, it would be wise to include Roma pupils and parents from other regions as well. Another interesting thing to investigate would be how teachers perceive their relationships and work with Roma pupils. Our study confirmed that the relationships between teachers and Roma pupils are mostly encouraging, but it also revealed some discriminatory behaviour in teachers that prevents Roma pupils from achieving the same learning results as the rest of the population. We should focus on these practices during future research and study the remaining factors that could have an impact on the learning success of Roma pupils in Slovene schools. We should also focus on exploring examples of good teaching practice in the field of Roma education. --- conclusion In the empirical qualitative study based on the answers of Roma pupils and their mothers, we focused on the experience and perception of teaching methods as well as teachers' attitudes towards Roma pupils. Based on the interviews conducted with the members of the Roma community in Prekmurje and Dolenjska, we can conclude that the elementary school teachers in the two mentioned regions are successful in providing a safe and inclusive environment for Roma pupils. Even though the results are mostly encouraging, we should not ignore those few examples that indicate a discriminatory attitude in teachers towards Roma children. Such cases should not be swept under the rug. They call for further efforts to eliminate discrimination, reduce differences, and promote the inclusion of minority groups in Slovene schools. Considering both the positive and the negative experiences of Roma pupils and their mothers, the results indicate that we need to continue with measures that will help not only improve relationships between teachers and pupils, but also provide adequate support and resources for equal opportunities, learning, and progress of Roma pupils. However, we should also explore other factors that affect Roma pupils' wellbeing in school, their learning performance, and advancement in the educational system.

In this article, we will discuss the case of fair and inclusive schools that are considerate of the specific and unique needs of their pupils and offer quality education for all. We will highlight the most important indicators of inclusive education and focus on research that confirms the importance of the teachers' willingness to establish good interpersonal relationships with their pupils. In the empirical section, we will introduce the results of our qualitative study. Based on interviews conducted with 20 mothers and 20 primary school pupils in two Slovene Roma settlements, we explored how pupils and their mothers perceive teachers' attitudes towards Roma children. Our findings were very promising, for they showed that the interviewed pupils and their mothers had a positive experience with the teachers and their pedagogical work with Roma children. However, we should not ignore the few cases that pointed to ethnic discrimination towards Roma pupils.

INTRODUCTION The family is a unit formed by interconnected and interacting parts that have a special structure or in other words have ties in the form of marriage and blood relations that influence religion, emotional ties, and customary law The family environment is the most important environment for children to learn about their cultural norms, values, attitudes, and beliefs so that children can grow into children who have a positive attitude to act in a wider environment afterward . In instilling these traits, the family has a very important role, namely the nuclear family , and extended family both of whom influence each other on the spiritual development of children, for example in worship . The family nucleus consists of a father, a mother, and children. The existence of the elderly who live together with the nuclear family, is referred to as a large family or extended family. Extended family is one of the characteristics of families in Indonesia. The extended family consists of family members not only the father, mother, and children, but also other family members. The extended family means that the family relationship between parents, children, and grandparents can be said to be very close. So that the duties and responsibilities in caring for children are gradually given to grandparents who have entered their final age or are elderly because the child's parents are working or so on. Cases of divorce and death are other factors that reasons for children to be cared for by grandparents besides being busy working to fulfill the family's economy . In addition to caring for grandparents, grandparents also instill spiritual qualities in their grandchildren which are useful for embedding themselves in perfect religion, such as: reciting the Koran, praying at the mosque, or something else. Mlaten Village is a village where grandparenting occurs due to many things such as infidelity, divorce, and being left to work abroad so that parents are not directly involved in raising children, the task of caring for children is transferred to grandparents who mostly work as farmers. Therefore, the description of the problems above is the reason for researchers to study more deeply about grandparenting in the spiritual development of children. in Mlaten Village, Mijen District, Demak Regency. --- RESEARCH METHOD The research in the article above uses field research. Field research is research in the form of descriptive analysis, namely research that is focused on certain events to be observed and analyzed, and researched. Data analysis in qualitative research begins with preparing and organizing data, where data is taken from interviews, observations, and others . According to Moleong, the qualitative data analysis process begins with reviewing all available data from various sources, such as interviews, and observations that have been recorded in field notes, personal documents, official documents, photographs, and drawings. After reviewing, the next step is data reduction, unit arrangement, categorization, and finally data interpretation . --- FINDINGS AND DISCUSSION --- Background of the Occurrence of Grandparenting in Mlaten Village, --- Mijen District, Demak Regency Forms of parenting grandparents can influence the formation of a child's personality after he becomes an adult. This matter is due to the characteristics and elements of the character of an individual maturing long before the seeds are already implanted into the soul of an individual from the very beginning, that is when he was a child. That is, the treatment of grandparents of their grandchildren since childhood will have an impact on their social and moral development as they growing age . It is this social-moral development that will shape character the nature and attitude of the child later, although several other factors influence the formation of reflected child attitudes in his character. Such as the main factor in the occurrence of grandparenting which was stated by Mrs. Nurul Wahidah in Mlaten Village, Mijen District, Demak Regency. "Because his parents had an affair with another woman, but his father still provided for his child. After this incident, his father divorced his wife and then married Siri to his mistress. Therefore, to make a living for his child, his mother was willing to work as a migrant worker in Arab Saudi". Some researchers have grouped Affairs into two types, among others sexual infidelity and emotional infidelity in which individuals share an emotional bond deeply with someone who is not a partner According to Guitar when someone is in a relationship or marriage and engage in sexual activity with other individuals who are not partner, it can be said that the individual is having a sexual affair. Some of the things included in This affair are having the intention to have a relationship sexual intercourse with other people, engaging in pseudo-sexual behavior , engaging in activities of sexual intercourse and intimacy with other people, either directly or through electronic devices by sharing things of an erotic nature. Then, someone is said to do emotional infidelity if the person has created emotional distance from the main partner and spending too much time with or thinking about other people outside of the relationship, in other words, the main partner becomes neglected or emotionally rejected. According to Guitar that was included in the affair above type is deceiving a partner about his feelings towards someone else, dedicating yourself emotionally to others, not emotionally satisfying your partner compared to others, falling in love with others, lying to a partner, discussing relationships with other people, and having romantic feelings for other people outside partner It is not only infidelity that causes grandparenting, but also divorce, as interviewed by Sumarti's mother as follows: "Because the father and mother are divorced, then the father is not responsible for providing a living for his wife and children. Therefore, to make ends meet for her children, her mother is willing to work as a female worker abroad ". According to Agoes Dariyo, a divorce is an event that is not planned and desired by the two individuals who are both bound by marriage. Divorce is the breaking up of a family because one or both partners decide to leave each other so they stop carrying out their obligations as husband and wife . Divorce doesn't only cause emotional disturbances for couples divorced but also the children will be affected. The impact of divorce on children will be heavier than on their parents. Sometimes the child will be feeling caught in the middle when the parents' divorce. Anger, fear, Separation anxiety, sadness, and shame are reactions for most children from the effects of divorce. Divorce that occurs in a family has an impact that affects the soul and condition of the child. Children experience obstacles in their fulfillment related to love and belonging parents must face the fact that their parents are divorced. The child gets a bad picture of family life. In a child's feelings, divorce is a shameful deficiency. Divorce almost always makes children sad, angry, and weak in spirit, the point is that children are in deep dilemmas and feel various problems psychologically On the other hand, the factors that affect grandparenting are not only infidelity and divorce but also being left to work by their parents, as said by Mrs. Suwarti and Mr. Muji as follows: --- "Mother and father work and I live with them and my grandchildren". "Because the father and mother work every day" Working every day or traveling is a normal and very common phenomenon in Indonesia, especially in Mlaten Village, Mijen District, Demak Regency. Community factors for working, especially in Mlaten Village, Mijen District, Demak Regency, are as follows: cultural factors and economic factors. 1). This cultural factor occurs because most of these people choose to migrate or work every day as the main alternative to earning a living. So that wandering becomes rooted and the culture of wandering becomes hereditary. 2). Economic Factors Another explanation is population growth that is not accompanied by an increase in natural resources that can be managed. If previously agricultural and plantation products were their main source of residence to support their family, now the results from natural resources which are their main source of income are no longer sufficient to provide results to meet their common needs, because they must be divided among them. several families. Someone who develops a child's spiritual intelligence must pay attention to opportunities to develop a child's self-awareness through activities related to art, such as drawing, playing drama, or playing music. These activities can develop children's self-awareness, e). Invite children to often ask "why?" or "what if?" and look for basic answers when solving problems, f). Provide opportunities and rewards for children to show good behavior and empathy, g). Teach children to try to overcome pain and suffering, h). Provide opportunities for children to think about the meaning and values of life in calm and solitude. --- Grandparenting in Strengthening Children Not only does the above understanding strengthen children's spiritual intelligence, but also instills an attitude of honesty that can be carried out through simple daily activities and as a habit for young children, namely behavior that can distinguish personal property and that of others. The basic ability to differentiate is the basis for being honest . Honesty is stated as a positive value because this behavior is beneficial both for those who commit it and for other people who are affected by it. Honesty is one of the principles that must be upheld by everyone as early as possible, not only important for students, students, and students. Honesty is very valuable for oneself, society, people, or a nation. Honesty will bring peace, inner peace, and even happiness to someone in society . So that people judge as someone good, not only in the family but to the wider community, as applied by Suwarti's mother in strengthening the spiritual intelligence of her grandson. "Invited to socialize with new people in terms of goodness, an example of joining jam'iyyah, and little by little introduced to the obligation of prayer, and often advised not to lie, try to tell the truth wherever he is" According to Baumrind, the above parenting style is included in the demandingness dimension. Dimensions demandingness is a dimension related to caregiver demands regarding the desire to make children part of the family, hope for mature behavior, discipline, provision of supervision, and coping with problems in child behavior. These demands are the hopes and efforts of caregivers so that children can fulfill their standards of behavior, attitudes and social responsibility are high or have set. Claims vary depending on the degree to which caregivers maintain, supervise, or try to make the child meet the demands . Spiritual strengthening does not only introduce worship but motivates people to do good because humans have spirits that sometimes go up and down, so when humans are in a state of low spirits they need to be motivated. Humans have the potential that when motivated they will show even better performance because motivation has a very good and positive impact on the development of the human soul, especially the development of children's education. Childhood is a period of ups and downs in terms of learning, so we must always provide motivational methods, as expressed by Mrs. Nurul Wahidah and Mrs. Suwarti in motivating their grandchildren in spiritual strengthening as follows: "I often motivate my grandchildren about this incident, in terms of patience in dealing with all problems" "Give motivation, and respect to others. Don't like lying, be honest and from your parents also give awards when you can memorize prayers or short letters" Motivation can be defined by everything that drives demanding behavior or encourages someone to meet his needs. At this point, motivation becomes the driving force of behavior as well as a determinant of spiritual Motivation and enthusiasm are the same units in strengthening the spirituality of their children because enthusiasm is a condition to take an action to be motivated, so enthusiasm has a function as a driving force for the human mind to act with strong determination. Self-motivation is also very necessary in grandparenting to raise grandchildren, where is the parenting style? in grandparenting, this also has an impact on grandchildren . --- The Impact of Grandparenting in Strengthening Children's Spiritual --- Intelligence in Mlaten Village, Mijen District, Demak Regency The form of parenting grandparents to their grandchildren is carried out according to the experiences and insights of grandparents. Grandfathers and grandmothers provide opportunities for their grandchildren to carry out their activities, make light work schedules and give verbal rules to their grandchildren with the hope that children will obey them. However, in applying the oral rules, grandparents include explanations that use words that are good and easy for children to understand, so that harmonious interactions are created between grandparents and children and with people around their environment Like the parenting applied by Suwarti's mother in educating her grandson as she has expressed: "There is a significant impact, quicker to understand, responsive, and quicker to accept new things than his age " From this expression it can be interpreted as a comprehensive understanding of early childhood, needed to achieve the goals of each child that will be carried out Becoming a capable person by socializing requires three processes one of which is learning to behave socially acceptable, play acceptable social roles received, and develop social traits . Using spiritual intelligence means functioning with the ability to have an impact and meaning of worship on every behavior or activity through steps and thoughts that are natural, towards a complete human being, and have an intergalactic and principled mindset only because of Allah . all of that has been done by Sumarti's mother in spiritual development for her grandchildren as follows. "Grandchildren can study the Koran and perform prayers at the mosque". From the expression above, there are several steps to grow and develop a typology of children's spiritual intelligence, namely independence . Instilling independence in children from an early age may be so that the child is detached from the character-dependent personality of others and providing motivation in children is the most important thing in growing children's courage in knowing knowledge through a caregiver The actions carried out by these children can be categorized as character-building values which have several components including: --- Religious Religion is an aspect related to the level of connection which includes the frequency and intensity of several behaviors, where the behavior has been determined by religion such as procedures for carrying out worship and religious rules. So it can be said as behavior in worship is determined by religion such as procedures for carrying out worship and rules in religion. In this case, the religious character of a child who is raised by his grandmother is expected to be used to diligently carry out worship daily and can develop his religious values . --- Advice This advice can open children's inner eyes to the essence of something, push it toward the situation sublime, decorate it with noble character, and equip it with Islamic principles. In cultivating prayer worship or studying the Koran in children and always giving advice so that children understand the importance of carrying out worship such as praying and studying the Qur'an . --- Discipline According to Hurlock said that Discipline is behavior and discipline following the rules and objectives to train and provide direction for children to be orderly, cooperative, and have noble character. So it is disciplinary behavior following rules that exist around or in the life child every day. The discipline itself can be seen through diligence and adherence to disciplinary conduct applied such as praying in congregation and learning to study the Koran . Discipline education is very important in a child's life. Disciplined children will be successful in In their life and life in society, discipline will reflect calm and serenity. Conversely, children who are not disciplined will lose their life and harm others. Method good at giving understanding to children in the discipline of prayer is to provide an understanding of fiqh about good prayer and true, performing prayers is guided by good discipline time, and obedience Discipline is control over the behavior of a person to always obey the rules of other people or yourself. Instilling principles so that children have a firm stance is a very important part of the strategy for upholding discipline and not just choosing friends, because friends can eliminate discipline if children are influenced by a bad culture. As for the method of enforcing discipline carried out by Mrs. Nurul Wahidah and Mrs. --- Sukarti to their grandchildren as explained by her: "There is a positive impact from my grandson, with my firmness and supervision, my grandson does not dare to hang out with naughty friends" "Because I educated him firmly and hard, the impact was that he was more independent, and diligent at school and studying the Koran" According to Baumrid, this form of parenting suggests that family care for their grandchildren is an authoritative form, namely encouraging children to be independent but still applying limits and controls to their actions . Authoritative grandmother show fun and support in response to her granddaughter's constructive behavior. they also expect the behavior of their grandchildren to be mature, independent, and ageappropriate. Grandchildren who have an authoritative grandmother are often cheerful, self-controlled and independent, and achievement-oriented. They tend to maintain relationships friendly with peers, work together with adults, and can cope with stress well In the pattern of care above , this parenting has rigid rules for raising her grandchildren. Each violation will be subject to punishment, is pushy, and tend to be uncompromising and deep one-way communication. Caregivers apply this parenting style when interacting with grandchildren, caregivers provide directions to grandchildren firmly without any resistance from the child himself, but when directed given is positive it will have a good impact on grandchildren, and if the direction given is negative, it will have a bad impact on grandchildren in their daily interactions. --- CONCLUSION Grandfather is the activity of caring for, raising, and educating grandchildren in survival. Some of the factors for the occurrence of grandparents are cases of divorce, the occurrence of affairs, and doing work every day. Then the identification of grandfathers in strengthening children's spiritual intelligence above is religious education such as: reciting the Koran, diligently carrying out worship at the mosque, and being able to instill honesty in all matters. Then the impact of grandparents in Mlaten Village, Mijen Subdistrict, Demak Regency, is always carrying out their obligations as Muslims, namely worshiping at the mosque, getting used to reciting the Koran, and being able to choose friends who can be invited to join. they. good and not bad, grandchildren who are cared for by grandparents can also live independently and be disciplined, and have the same abilities as children who live with both parents.

Grandparenting is childcare carried out by surrogate parents (grandparents) where grandparents are part of a large family that is trusted by parents to care for their children either temporarily or permanently, in their care grandparents educate their grandchildren in the ability to strengthen spiritual intelligence which is covering a wide variety. This study uses qualitative field research, where data is obtained through interviews. The results of the study stated that the background of the occurrence of grandparents was due to various things: infidelity, divorce, or being left by their parents to work. Then grandparents in strengthening children's spiritual intelligence by familiarizing themselves with religious activities such as: learning the Koran, praying in congregation, and training themselves to be honest in all things. Then the impact of grandparents is being able to strengthen religious traits in everyday life such as worship activities, socializing well, and being able to distinguish between good and bad traits.

Summary The once-in-a-100-year Coronavirus Disease 2019 pandemic has resulted in over 44 million confirmed cases and 1.2 million deaths by the end of October 2020 [1]. Due to the airborne human-to-human transmission of the infectious disease, global governments issued restriction orders of social distancing, self-isolation and travel bans to reduce transmission risk, and consequently caused crucial impacts. In particular, the labor requirements decreased across all economic sectors and unemployment numbers increased [2,3]. To record, measure, and reduce the negative influence on the society and economy caused by the pandemic, a comprehensive socioeconomic factor collection for the pandemic period is urgently needed for comparing, modeling, and predicting the socioeconomic impact of COVID-19. There are numerous scientific papers on modelling the mechanisms of disease spreading in a spatiotemporal perspective [4][5][6]. In such studies, geography matters in that place, both absolute location and relative spatial relationships, have been widely recognized as an essential dimension of epidemiologic research [7]. The region-based characteristics and the interconnection among spaces have attracted more attention from both public health experts and policy makers for understanding the mechanisms and controlling the spread of disease. The interdependent processes of health, political, economic, and social issues work together and construct the complex social system. Many scholars have attempted to discover the correlation and determination factors of the novel disease from the socioeconomic aspects [8,9]. Under the background of regional or global epidemic events, the affected countries always have social and economic issues. For example, during the Spanish flu, the SARS pneumonia, and the Ebola virus, areas including Europe, various parts of China and West Africa experienced different degrees of recession, including dropped GDP, population, and even sparked diplomatic conflicts and local wars [10,11]. During the severe acute respiratory syndrome coronavirus in 2002, there were serious economic consequences, including a collapse of stock markets in Asia, Europe and the United States, disruption of trade and tourism, stagnation and recession in manufacturing, and reduced supplies of goods, food and medicine [12]. Similarly, a large number of studies have shown that socioeconomic conditions have a strong relationship with the COVID-19 pandemic, and strong interrelations exist among the internal social and economic factors of geographic regions [5,[13][14][15][16]. For example, individuals with low socioeconomic status are more likely to suffer from harsher conditions with less accessibility to medical services and financial hardships. Moreover, regions with lower GDPs have limited abilities of public healthcare. Additionally, the demographic issue would also make the region more vulnerable to the pandemic. Furthermore, the mitigation-policy plays a pivotal role in the current process of pandemic control that restrict transportation, lock down cities, and shut down business, which all have great impacts for helping society combat the virus. Financial markets recorded continuous plummeting with four trading curbs in the USA stock market in March 2020 with COVID-19. International trade slowed down after international transportation limitation, businesses went bankrupt and people were left unemployed. A variety of government departments and agencies offer datasets with raw measurements and variables, but the criteria for data publishing are very mixed. As all of the different information rushes from different sources; a precise, focused, timely, and standardized data collection distributed by a stable platform is needed urgently to provide integrated and convenient data sources for users and decision makers in different domains. Due to the diversity of socioeconomic data sources, the access pipeline and data format are not uniform over all sites. Thus, users must access raw data files through time-consuming manual manipulation or customized operational tools to integrate the multi-source information. For example, the personal income tables [17] from the Bureau of Economic Analysis are accessed by URL links while interactive actions of selecting target options are required to obtain the unemployment insurance data from the U.S. Department of Labor [18]. Moreover, not all the indicators from the original datasets are readily meaningful for pandemic research and it is necessary to effectively screen out attributes to highlight the desired characteristics of socioeconomic measurements. For example, the Census Bureau [19] provides thousands of fields for multiple geographic scales each year, but less than 20 factors are widely recognized and used in COVID-19-related analysis and studies [11,20]. Socioeconomic factors such as the policy stringency index [21,22] are broadly accessed and utilized at country level, but not at the state-level in the USA. It is necessary to leverage the Data 2020, 5, 118 3 of 18 state policy as a quantitative constrain with other dynamic spatiotemporal attributes in COVID-19 data collection. To fill the aforementioned gaps, we propose a preprocessed, filtered and standardized COVID-19 socioeconomic data collection of the USA. The data collection is validated and contains intensive quality control based on integrity, consistency and effectiveness requirements. It offers a crucial data basis for the decision makers to assess loss due to the pandemic and make further mitigation and reopening plans. Researchers can easily implement studies between COVID-19 and socioeconomic factors with the data and the public can be better informed about the crisis impacts. The paper is organized as follows: Section 2 introduces the raw data, selected attribute values, and metadata of the derived data product; Section 3 describes the methodology concerning how derived attributes are organized and produced, how data are processed and stored, and how data quality is controlled; and finally, Section 4 illustrates the data publishing methods and provides access methods. --- Data Sources --- Raw Measurement of Socioeconomic Factors --- Qualitative Restriction Policy Orders To date, every state of the USA has made great efforts through policies and orders to mitigate the spread of the virus and support testing and treatment of affected communities. The official emergency declaration, policy orders and law documents that allow state governors to execute emergency forces and are a valuable and straightforward data source for researchers to evaluate the policy context. Governmental policy data were originally collected by the Oxford COVID-19 Government Response Tracker program during the pandemic to serve as an indicator of governments' responsiveness to the crisis among countries globally [21]. We also proposed and generated US state-level restriction policy orders including the social distance restriction and financial/economic supporting policies. To measure and evaluate the policy stringency index, the raw context documents and quantitative data are searched and identified from each state's official COVID-19 website, alongside executive orders that announce related policies and authoritative news websites by a team of over 10 students and volunteers in the National Science Foundation Spatiotemporal Innovation Center . The collection of these policies mainly follows the categories proposed by Hale et al. [21], and only policies that either fall under these categories , or have a great impact on people's lives under COVID-19 are collected. Since the documents are available online, the original policy materials are recorded as Uniform Resource Locator in "notes" for coding and retrospect. Collecting governmental policy data not only informs us about the actions that each government takes, but also enables us to conduct studies and analyses on the effectiveness and influence of the policies. The government's responsive policy is directly related to the spread of COVID-19. The combination of constructive policies and practical implementation can effectively stop the spread of virus. For example, closure schools and businesses and cancellation of public events helped China control the situation successfully. Moreover, a public information campaign at an early stage can raise the public's awareness of the pandemic and inform them of the actions that can be taken to protect them from being infected. Additionally, policies restricting national and international travel can help countries that have not had COVID-19 to stop its spread from the start. Overall, policies and executive orders issued by governments are necessary to inform and unite the whole public together to combat the current COVID-19 crisis. Data 2020, 5, 118 4 of 18 --- Macroeconomic Indicators Several macroeconomic variables are utilized to reflect overall economic conditions and play an important role in the analysis and reference of macroeconomic regulation and control during the COVID-19 pandemic from global to country scales [2,23]. Factors such as gross domestic product , personal income and international trade amount per capita could be used in analysis and to predict the mortality trend, finances and economic performance. Furthermore, the economic indices are expected to decrease with the second or third outbreak waves of COVID-19 in the USA. GDP by state is defined as "the sum of value added from all industries in the state" and it includes all private and public consumption, investment, government outlays and net exports. The Bureau of Economic Analysis provides the state-level GDP quarterly from the first quarter of year 2016 and will release the third quarter of year 2020 on 23 December 2020. Real GDP and current-dollar GDP and Chain-type quantity indices from raw products are collected in this dataset [24]. The business shutdown and job loss affected the personal income in many industries. According to Pew Research Center, around half of the American lower income class lost jobs because of COVID-19, and only 23% of them have enough money to last three months [25]. Since the lower income class is more likely to find labor jobs or jobs that do not require an educational background, it makes more financial sense for business owners to find a replacement after the pandemic ends [26]. The personal income data are another state-level product provided by the Bureau of Economic Analysis, which has been reliably collecting U.S. personal income quarterly data since 1948. The agency has been releasing reports on a quarterly basis and also at the national, state, and local level [17]. We selected four attributes from this dataset: Overall Compensation, which refers to the compensation including all categories; Farm Compensation, which is compensation for farm related employees; Nonfarm Compensation, which is compensation for non-farm related employees; and lastly the Per capita personal income, which is a ratio-based index calculated as the total personal income of the residents in a state divided by the total population of the state. The attributes were updated to the 2nd quarter of 2020, and the 3rd quarter of 2020 data will be released on 17 December 2020. International trade of goods and services information for the state are collected from the U.S. Census Bureau, in which the data have been tracked since 1960, every month. By examining and comparing the attribute values of the USA between 2019 and 2020, there are noticeable decreases for both exports and imports ratios starting from March to June of 2020, due to the limitation in import/export transportation policy and resource demand changing during the pandemic [27,28]. Four attributes were picked as the overall indices of international trade business, including the export and import values in the manufactured and non/manufactured commodities. The Manufactured Commodities refer to the trading amounts of manufactured commodities that export/import in millions of dollars. The Non-Manufactured Commodities cover the products from agricultural, forestry, fishery industries in million-dollar units. --- Employment The COVID-19 pandemic has created a major impact in the U.S. Labor Market [29]. Unlike the recession that happened in 2008 in which the labor market faced a slow hit, this pandemic impacted the labor market with an intense speed [29]. By late March 2020, the seasonally adjusted initial claims were 3,283,000, which was the highest number of initial claims filed in the history of the U.S. labor force [18]. To illustrate the impact of COVID-19 crisis in employment, we collected unemployment insurance claim and non-farm payroll datasets published by the United States Department of Labor every week. The dataset includes initial claims, insured unemployment rate, continued claims, and covered employment. An unemployed person separated from an employer files an initial claim [4], and if the unemployed person experiences a week of unemployment, he/she will file a continued claim to claim benefits for that week of unemployment [30]. The nonfarm payroll represents the total number of paid employees in the U.S. workforce excluding farm employees, government employees, private household employees, and employees of non-profit organizations [30]. It is one of the most significant economic indicators because it shows the number of new jobs created during the reference period. Thus, it has the potential to indicate that businesses are actively hiring. Additionally, it also indicates that employees can positively contribute to the economy by consumer spending. The Bureau of Labor Statistics publishes non-farm payroll data based on two comprehensive surveys, the Household Survey, and the Establishment Survey [18]. The Household Survey pertains to individuals and provides the unemployment rate report as well as details on employment demographics. The Establishment Survey pertains to jobs and provides the number of new nonfarm payroll jobs added within the national economy. Our nonfarm payroll dataset includes the total number of employees from major sectors which are obtained from the Establishment Survey and labor force participation rate, employed persons, and the unemployment rate obtained from the Household Survey. --- Housing Market The COVID-19 pandemic created an unprecedented disruption in the U.S housing market. During Spring 2020, home sales and the median housing price dropped significantly due to stay-at-home orders in comparison to a normal busiest season. This has been attributed to buyers not willing to invest huge amounts in housing and sellers not willing to list or sell their home due to the reluctance of strangers entering their home during the pandemic. However, the housing market immediately surged across metropolitan and rural areas and across house sizes in the summer and fall because of low loans rates and economy rebound triggered by the reopen and stimulus package [31]. After careful consideration, the data collection on the housing market includes the Building Permits Survey , Federal Housing and Finance Agency House Price Index , and Housing Inventory . These three housing factors are captured before, during, and after the pandemic period for policymakers, economists, and researchers to understand the housing market under the influence of the COVID-19 crisis. The U.S. Census Bureau and the U.S. Department of Housing and Urban Development have jointly announced building permits data since 1959 monthly and annually [32]. It provides statistical information on new privately-owned residential construction on the national, state, and local levels. The attributes of building permits include total housing units, single housing units, and multi-units . The FHFA HPI measures average price changes in repeat sales or refinancing on the same properties [32]. This index information on house prices is based on reviewing repeated mortgages that have been purchased or securitized by Fannie Mae or Freddie Mac on single-family properties [32]. The FHFA has provided house price index data since the mid-1970s. The attributes of HPI include All-Transactions, Expanded-Data, and Purchase-Only. The housing inventory focuses on the market trends and monthly statistics on active homes for-sale listings in a specified geography. The realtor.com website publishes Housing Inventory data based on accurate mapping of housing units in listing statuses at national, state, and local levels [33]. The attributes of housing inventory include Median Listing Price, Days on Market, and Price Increase Count. --- Medical Resources Medical resources show the capabilities of the health care system to support disease testing, diagnosing, and treatment. Three fundamental elements of ventilators, hospital beds, and medical staff were identified and widely reported as medical resources to plan for or deploy under COVID-19 [34]. A county-level dynamic medical resources deficiency index based on medical staff and intensive care unit beds data are generated to measure the local medical burden from the accumulated active confirmed cases [35]. In this collection, hospital number, licensed and intensive care units beds, registered nurses, and medical doctors for populations over 1000 are collected and calculated for each state. The number of hospitals and licensed/ICU beds information is provided by Definitive Healthcare consulting services via ArcGIS online repository [36] in a point-based shapefile format, which records the attributes of each unique hospital. The medical staff is separately accessed from the National Council of State Boards of Nursing [37] and the National Provider Identifier Registry database [38]. The number of registered nurses was counted by NCSBN at the end of 2019 based on Data 2020, 5, 118 6 of 18 the records of active nursing licensure of the electronic information system. The number of medical doctors was extracted and summarized from the NPI database based on taxonomy codes of health-care providers, which indicated the areas of specialization. Although these collected medical recourses are not a dynamic dataset that reflects the real-time medical capacity during COVID-19, they can be used as a baseline to estimate the geographic disparity of health resources in each state. --- Census-Based Socioeconomic Data With the changes in the economy, data from the decennial censuses are far from enough for governments and businesses to rely on for planning, as they are published too late and could be out of date. So, the Census Bureau began the American Community Survey program, a nationwide survey that collects demographic and socioeconomic, housing, and other characteristics about the nations' population each year. More than 3.5 million households across the country would be contacted every year by ACS. The data are published on several geographical levels, including state, metropolitan areas, as well as counties, cities, and smaller areas. For smaller areas, like census tract or census block groups, it would take 2-5 years to acquire enough samples for estimation. So, the Census publishes three types of data for users: 1-year, 3-year, and 5-year estimates. To choose the data product, the reliability, precision, currency, and geography unit need to be considered. The 5-year estimate data are the most reliable but least current data. The ACS program was proposed in 1996 and scheduled to be fully implicated in 2003. The oldest data available for ACS 1-year estimates are in 2005 with an annual release, and 2009 for the 5-year estimates. The 3-year estimates data have been discontinued and only 2007-2013 will remain available, and no new data have been produced later [19]. Census-based socioeconomic data are widely used in COVID-19 research as determinant factors per capita to show the heterogeneity of locations. The state-level attributes for topics of population, age, sex, ethnicity, education, poverty and income, housing, and employment status are selected from the American Community Survey in this dataset. The 2019 ACS was released recently on 17 September 2020, and the 1-year estimate data were extracted and integrated manually into the proposed data collection. --- Data Description Based on different utilization purposes and characteristics of temporal frequencies, three separately derived and aggregated datasets are provided to the public, including daily-based policy stringency index, dynamic economic indicators through the COVID-19 pandemic, and the socioeconomic determinants for U.S. states. The spatial resolution of these three datasets is state-level and the spatial coverage is the 50 states of the U.S. with the District of Columbia . A USA base map with the standard state name, state abbreviation code, and Federal Information Processing Standards code is used in each derived dataset to connect the geographic regions. Figure 1 illustrates the conceptual model and its derived attributes for the state level of the USA. The FIPS column is used as the foreign key for a joint table of indicators within the same location and each attribute table is stored separately based on time frequency. The timestamp column for each table stands for the first day point of the updated time range, e.g., 1 January 2020 in the quarterly table for the first quarter of 2020. All time-series based datasets are presented in two data formats, time-series summarized product, which holds historical data for every single attribute in a specific data table; and periodic temporal report, which records all attributes with the same updating frequency in a different timestamp-based data table. --- Daily Policy Stringency Index The Government Response Stringency Index was first introduced by the Oxford COVID-19 Government Response Tracker [21]. It is an indicator of the governments' responsiveness to the crisis of COVID-19 and the extent of lockdown in the specific region on a scale from 0 to 100. This index is used to graph the temporal dynamic policy change for both the global and the U.S. in our research. To quantitively measuring of the restriction policy inside the U.S. started from early March 2020; the Response Stringency Index for each state is recorded, coded, and calculated based on raw policy documents from the local state government. The raw data collection with original recording notes, coded scores in the specific coding division, and links to the specific policy used for coding is provided in a daily manner. Based on the global country-level index standard, the seven geographical-based categories of policy are used to indicate the scope of the policy. Based on the literature and data accessibility, we selected 29 features from the public dataset which have statistical values at the state level. Those data are from a variety of government and nongovernment agencies with relatively high reliabilities and have been widely used in different research projects. Compared with the annual report data, the aforementioned indices have higher temporal resolutions ranging from weekly to monthly and quarterly. Those data mainly focus on economic-related statistics, such as GDP, employment, import and export in international trade, and retail. Many core indexes are included under each category, and also some expanded indices are listed with more detail. For example, the unemployment rate is a core indicator in all kinds of employment discussions, also we considered some related features such as --- Daily Policy Stringency Index The Government Response Stringency Index was first introduced by the Oxford COVID-19 Government Response Tracker [21]. It is an indicator of the governments' responsiveness to the crisis of COVID-19 and the extent of lockdown in the specific region on a scale from 0 to 100. This index is used to graph the temporal dynamic policy change for both the global and the U.S. in our research. To quantitively measuring of the restriction policy inside the U.S. started from early March 2020; the Response Stringency Index for each state is recorded, coded, and calculated based on raw policy documents from the local state government. The raw data collection with original recording notes, coded scores in the specific coding division, and links to the specific policy used for coding is provided in a daily manner. Based on the global country-level index standard, the seven geographical-based categories of policy are used to indicate the scope of the policy. Based on the literature and data accessibility, we selected 29 features from the public dataset which have statistical values at the state level. Those data are from a variety of government and non-government agencies with relatively high reliabilities and have been widely used in different research projects. Compared with the annual report data, the aforementioned indices have higher temporal resolutions ranging from weekly to monthly and quarterly. Those data mainly focus on economic-related statistics, such as GDP, employment, import and export in international trade, and retail. Many core indexes are included under each category, and also some expanded indices are listed with more detail. For example, the unemployment rate is a core indicator in all kinds of employment discussions, also we considered some related features such as Insured Unemployment rate, Total Non-farm Employees, Manufacturer Employees, and so on. For unemployment insurance, we also list initial claim numbers and continuous claim numbers. These features could give users a more in-depth view of the current situation when combined. More detailed descriptions of attributes are list in Table 2. --- Methods --- Cloud-Based Data Warehouse and Spatiotemporal Aggregation A data warehouse is a large, centralized data repository of integrated data from one or more disparate sources [39]. In this study, we adopt the DW workflow to aggregate open-source socioeconomic datasets described in Section 2.1. As the data pipeline shows in Figure 2, data files from raw format are first obtained and stored in the staging area, then loaded into the DW as raw data manually coded or automatically converted. The processing scripts and formatted rules are developed based on a standardized spatiotemporal collection framework for COVID-19 to collocate in time and space among multiple factors [40,41]. Next, the formatted raw data are converted into summarized data products with corresponding metadata for sharing under open-source data policies. Last but not least, users will be able to visit, access, download, and leverage the proposed products for their specific analysis, mining, and reporting applications. --- Methods --- Cloud-Based Data Warehouse and Spatiotemporal Aggregation A data warehouse is a large, centralized data repository of integrated data from one or more disparate sources [39]. In this study, we adopt the DW workflow to aggregate open-source socioeconomic datasets described in Section 2.1. As the data pipeline shows in Figure 2, data files from raw format are first obtained and stored in the staging area, then loaded into the DW as raw data manually coded or automatically converted. The processing scripts and formatted rules are developed based on a standardized spatiotemporal collection framework for COVID-19 to collocate in time and space among multiple factors [40,41]. Next, the formatted raw data are converted into summarized data products with corresponding metadata for sharing under open-source data policies. Last but not least, users will be able to visit, access, download, and leverage the proposed products for their specific analysis, mining, and reporting applications. From a computational infrastructure perspective, cloud computing techniques are used to establish the operation system by George Mason University NSF Spatiotemporal Innovation Center . Based on the virtualization technique, cloud computing enables users to utilize computing resources on demand [39]. The operational environment of the staging area, data warehouse, and data sharing services are deployed in a private cloud. In real practice, data retrieving, and processing tasks are scheduled based on the updated frequency of original data sources in the cloud, and the script-scheduling approach could reduce the consumption of computing, storage resources. Since the frequency of the socioeconomic factors varies from daily, weekly, monthly, to quarterly, it is necessary to establish a data framework to build the relationships among different From a computational infrastructure perspective, cloud computing techniques are used to establish the operation system by George Mason University NSF Spatiotemporal Innovation Center . Based on the virtualization technique, cloud computing enables users to utilize computing resources on demand [39]. The operational environment of the staging area, data warehouse, and data sharing services are deployed in a private cloud. In real practice, data retrieving, and processing tasks are scheduled based on the updated frequency of original data sources in the cloud, and the script-scheduling approach could reduce the consumption of computing, storage resources. Since the frequency of the socioeconomic factors varies from daily, weekly, monthly, to quarterly, it is necessary to establish a data framework to build the relationships among different datasets in time and space. This study takes advantage of the proposed spatiotemporal data collection cube [40] of COVID-19 to collocate spatial and temporal scales with virus cases and environmental factors. The data cube is a three-dimensional structure with locations, timestamps, and attributes to record spatiotemporal measurements and corresponding factors. All attributes with multiple topics are organized to show the spatiotemporal variation before, during, and after the pandemic period. The time update report of each factor is provided based on timestamp and summary report according to the attribute variation across time series. Sections 3.2 and 3.3 detail manual policy index extraction, the coding process, and the automatic web crawler in the data Extract, Transform, and Load process. --- Policy Index Extraction and Coding Standard This stringency index by state in the USA is used to graph both the global and U.S. temporal dynamic policy changes in our research. The Response Stringency Index includes seven categories of policy as shown in Table 4, such as school closure, business closure, public event cancellation, etc., as well as a generalization code to indicate the scope of the policy. The calculation of the index includes three steps. 1. First, we code the qualitative policy into quantitative numbers according to the policy's stringency. For instance, a government's official announcement that recommends business closure will be coded as 1, and a strict requirement of business closure will be coded as 2. Another example is when there is no policy restricting internal movement, the code would be 0, and if governments recommended movement restrictions , the code would be 1, and if the government adopted and implemented the "Stay at Home Order" that requires people to stay at home, then the code for this policy's stringency would be 2. Additionally, there is another code for the national/regional coverage of the specific policy . For example, the code will be 0 if the policy is only targeted at some counties/cities of the state/country, and the code would be changed to 1 if the policy targets the whole state/country. 2. Then, the code numbers of each policy category are summed up and rescaled to create a score between 0 and 100. --- 3. In the end, the seven scores are averaged to obtain the overall Stringency Index for that region shows). The policy index extraction is conducted manually by the NSF STC group. --- PolicyStringencyInex = 1 7 [rescaled +rescaled +rescaled +rescaled +rescaled +rescaled +rescaled] In order to ensure the quality and reliability of the raw policy data, it is constantly evaluated through dimensions including data selection, integrity, consistency, and validity. To ensure that the policies included in the current dataset are reliable and up to date, the initial set of state-level policies of closure are collected only through each state's official website, and only state Governors' executive orders or official announcements are taken into consideration. Then, the relevant USA state-level policy collection site from the well-known and reliable news platforms , third party Non-Governmental Organization Kaiser Family Foundation , and academic institutes are tracked and considered in our collection procedure. By comparing and evaluating the relevant data products in an operational mode, the integrity and validity are guaranteed. To ensure the consistency of the state-level policy data, the policies are encoded by the same research using the same standard through the whole data-collecting process. --- On-Demand Web Crawler There are several ways to obtain useful information from web pages. While some websites provide an Application Program Interface to extract data in the structured format, others do not contain this kind of API. In this case, web scraping can be an ideal technique to obtain web content. Web scraping includes extracting useful information from a web page by understanding the web page structure [42]. A web page is composed of Hypertext Markup Language tags and is translated by a web browser in a human-readable format. With python packages like selenium, beautifulSoup4, urllib2, etc. we can directly extract the needed information from web pages. Understanding the Document Object Model is essential for web content extraction because DOM defines the logical structure of a web page and provides an insight on how or what page element to access [43]. It defines the attribute of a page element such as a class, and name which is used to identify a page element for page actions. The first step is to identify an officially authoritative and reliable website for the research demands. The website should satisfy both the spatial and temporal requirements of the dynamic socioeconomic datasets. The next step is to analyze the format and structure of useful data/content on the website. Some websites provide data in Comma Separated Values or spreadsheet format that can be readily downloaded, processed, and converted. Additionally, few websites provide data in a semi-structured or unstructured format, such as JavaScript Object Notation and Portable Document Format files. Then, we need to develop and perform a customized web crawler to extract useful information. This procedure is shown in Figure 3. The web crawler section comprises a sequence of steps to acquire information from a website. --- 1. First, we interpret and understand the DOM structure. --- 2. The next step is to provide the users with inputs in the required fields and submit the form. This is achieved by selenium API, which automatically interacts with a web page by calling browser drivers such as chrome, gecko , and Internet Explorer . Subsequently, by inspecting the input field web elements we provide input values and submit the form using a button click. Data 2020, 5, 118 13 of 18 3. The next step is to obtain the page source. For example, to extract unemployment insurance data from the United States Department of Labor website [18], we used selenium to automatically: a. Select the "State" radio button. b. Select "2019" from the drop-down options for the start year. c. Select "2020" from the drop-down options for end year. d. and select all states from the textbox. e. Finally, we submit the form by clicking the Submit button and obtain the page source information. --- 4. Once we have the page source with the required information, we use the beautifulSoup4 python package to identify the table that holds the unemployment insurance data and extract the text content from the table. The final step is to format the dataset in a standard format. --- On-Demand Web Crawler There are several ways to obtain useful information from web pages. While some websites provide an Application Program Interface to extract data in the structured format, others do not contain this kind of API. In this case, web scraping can be an ideal technique to obtain web content. Web scraping includes extracting useful information from a web page by understanding the web page structure [42]. A web page is composed of Hypertext Markup Language tags and is translated by a web browser in a human-readable format. With python packages like selenium, beautifulSoup4, urllib2, etc. we can directly extract the needed information from web pages. Understanding the Document Object Model is essential for web content extraction because DOM defines the logical structure of a web page and provides an insight on how or what page element to access [43]. It defines the attribute of a page element such as a class, and name which is used to identify a page element for page actions. The first step is to identify an officially authoritative and reliable website for the research demands. The website should satisfy both the spatial and temporal requirements of the dynamic socioeconomic datasets. The next step is to analyze the format and structure of useful data/content on the website. Some websites provide data in Comma Separated Values or spreadsheet format that can be readily downloaded, processed, and converted. Additionally, few websites provide data in a semi-structured or unstructured format, such as JavaScript Object Notation and Portable Document Format files. Then, we need to develop and perform a customized web crawler to extract useful information. This procedure is shown in Figure 3. The web crawler section comprises a sequence of steps to acquire information from a website. 1. First, we interpret and understand the DOM structure. --- Data Quality Control We evaluated the socioeconomic data in three dimensions including data integrity, data consistency, and validity to ensure quality datasets are delivered to geospatial researchers and policymakers. Raw data selection, cleaning, qualification: The first step for socioeconomic data collection is to select appropriate raw input data. We implement an extensive literature review related to the impact of COVID-19 on socioeconomic factors, thereby making a factor justification of datasets we have selected. We also ensure that the selected data are of the most significance to show the economic hit in the U.S. during the pandemic. The next step is to identify reliable official websites to download the datasets that satisfy both the spatial and temporal needs of the data collection process. During the data cleaning step, we filtered or discarded the invalid attributes and values that are not included in the research domain. For example, we filtered out the U.S. territories such as the Virgin Islands and Puerto Rico as we were focused on the U.S. states and the District of Columbia. In certain cases, we faced issues in downloading datasets directly in a structured format such as in spreadsheets or CSV. We overcome the issues by first obtaining data in a text format and then converting it to a structured format. Data integrity: Ensuring data integrity means that the collected datasets are complete, comprehensive, and accurate. This includes verifying datasets for manual errors, logical errors, and data type consistency. We stored the socioeconomic datasets in a relational database. The table constraints such as primary key ensure duplicate records are not inserted, not nullable fields prevents inserting null values, data type ensures consistent data are inserted into each field. For example, raw data for personal income farm compensation for the District of Columbia is . This is converted to 0 to satisfy the data integrity. Data consistency: The socioeconomic dataset in the repository is required to be consistent with the other sources. This means that the collected data should be consistent with the values of data sources and should also be consistent with the dynamic change of economic factors during the pandemic. For example, the COVID-19 saw the worst hit on the U.S. economy. During the start of March in which the COVID-19 cases were at a peak, the economic factors were comparatively lower than the pre-and post-pandemic periods. Data validity: To ensure the data reliability of socioeconomic factors, we provide data sources to the consumers along with the collected data. Consumers can investigate the collected data against the data sources, thereby ensuring the validity of the data. --- Data Sharing The derived data are all in CSV format which is automatically published in the GitHub data repository promptly based on the time-frequency and publishing timestamp. As one of the most popular open-source communities, GitHub is a solution for data consumers and application developers to share data and codes. Publishing the data cube on GitHub assures data release and sharing and facilitates the following: allows data consumers to report issues/problems through the GitHub repository; allows data consumers to folk or mark the repository to trace a data update for timely analysis; and allows data analysis code to be published with data to facilitate the data cube's utility in research and decision-making. Further notes on the usage of the dataset that will help other researchers to quickly access the dataset and work with it. To enable public users to explore, search and quickly identify the COVID-19 relevant data features and attributes, a Comprehensive Knowledge Archive Network based the open resource portal is created to provide metadata for the data collection and analysis models for the spatiotemporal COVID-19 rapid response studies . Both collected and processed datasets are prepared for querying, browsing, and sharing. The COVID-19 data discovery portal also enables data owners/editors from longtail sections to register user accounts, organization pages, and create resource pages. Multiple data licenses are used for data reusing, copying, publishing, distributing, transmitting, and adapting. All datasets can be accessed and cited for non-commercial purposes. More importantly, a well-designed tagging and grouping system is constructed based on research communities, topics, and interests, and it can be used to filter out the most relevant dataset for researchers. All socioeconomic datasets in this collection can be filtered out by the "social economics" group under the "Spatiotemporal Innovation Center" organization. --- Time Trend Analysis of Typical Attributes Several socioeconomic indicators are selected for visual analysis for time change patterns before and after COVID-19 pandemic, including daily new confirmed cases, policy stringency index, percentage change of real GDP and employees' compensation, weekly unemployment rate, median listing price in housing market, as well as export and import amounts of manufactured commodities. In each plot of Figure 4, a line represents a state's dynamic change from the period of before emergence , to outbreak , and to when COVID-19 was under control . The four typical states chosen in these time series plots are New York , Georgia , Texas , and Illinois . The visualization provides a straightforward way to understand the trend of the social economic situation, and corresponding significant impacts with exact time periods can also be identified. --- Conclusions To combat the COVID-19 pandemic, the reported socioeconomic data collection provides valuable spatiotemporal factors for research and decision support by disease control experts, decision-makers, government officials, sociologists, economists, and humanists. The spatiotemporal data collection framework could be a baseline for integrating other spatial and temporal based factors. The collection includes datasets of the daily updated policy stringency index, economic attributes in multiple time-frequencies, and a socioeconomic determinant for all the states in the US. GMU's NSF STC is maintaining data processing, quality control, storing, and sharing in an operational mode. The raw data tables are automatically accessed from multiple authorized departments of the U.S. by customized python scripts, and the processed attributes are extracted and converted into the GitHub repository according to the quality control and framework in near real time. In the future, more metadata of data sources will be provided by a crowdsourcing approach through a CKANbased portal and its inclusive socioeconomic factors and attributes under COVID-19 topic will be integrated into this spatiotemporal standard collection. From the new case numbers and policy index, we can find that New York has implemented and kept stricter policies which helps to decrease new infected cases of COVID-19. In addition, relaxation of restrictions causes a rebound of the new cases number in Texas and Georgia. Due to the lock down nationwide, the GDP and compensation of employees fall sharply for the 1st and 2nd quarter. At the same time the unemployment rate increases dramatically. Even though loose policy after June and July decrease the unemployment rate, it is still at a high level around 10% which was usually 3-4%. The housing market did not show a big change, and a slightly increase for the real estate listing price after June was recorded. Texas has a large portion of international trade of manufactured commodities, and the import and export rates for all states have risen back to normal amounts after the policy index trend to decrease occurred. Spatiotemporal analytics [44] can be easily conducted with the collected and published datasets. --- Conclusions To combat the COVID-19 pandemic, the reported socioeconomic data collection provides valuable spatiotemporal factors for research and decision support (policy making, academic research, and public --- https://github.com/stccenter/COVID-19-Data/tree/master/Socioeconomic%20Data is the data repository link to access the latest data collection with well-formatted documents. --- Data 2020, 5, 118 7 of 18 --- Socioeconomic Driven Factors/Determinants/Indicators The census data are a traditional source used to extract demographic, education, income, poverty, housing, and health resources information with geographical units. Socioeconomic determinants for the state level are mostly from Census ACS 2019 data and partially from other sources. For demographics, we extract the attributes for the total size, density, and by groups of age, gender, and race and ethnicity. For education information, the attributes contain percentages of the population with high school education, a bachelor's degree, or a higher degree. The two types of regional information, internet accessibility, and medical resources are recorded in separate columns. All mentioned attributes are considered as the factor strongly related to regional health in response to the pandemic COVID-19. Funding: This research was funded by the NSF . ---

The outbreak of COVID-19 from late 2019 not only threatens the health and lives of humankind but impacts public policies, economic activities, and human behavior patterns significantly. To understand the impact and better prepare for future outbreaks, socioeconomic factors play significant roles in (1) determinant analysis with health care, environmental exposure and health behavior; (2) human mobility analyses driven by policies; (3) economic pressure and recovery analyses for decision making; and (4) short to long term social impact analysis for equity, justice and diversity. To support these analyses for rapid impact responses, state level socioeconomic factors for the United States of America (USA) are collected and integrated into topic-based indicators, including (1) the daily quantitative policy stringency index; (2) dynamic economic indices with multiple time frequency of GDP, international trade, personal income, employment, the housing market, and others; (3) the socioeconomic determinant baseline of the demographic, housing financial situation and medical resources. This paper introduces the measurements and metadata of relevant socioeconomic data collection, along with the sharing platform, data warehouse framework and quality control strategies. Different from existing COVID-19 related data products, this collection recognized the geospatial and dynamic factor as essential dimensions of epidemiologic research and scaled down the spatial resolution of socioeconomic data collection from country level to state level of the USA with a standard data format and high quality.

of understanding cultural issues and linking social factors to better understand the barriers and facilitators of obesity risk in immigrants. Recommendations for developing culturally tailored weight-control programs matched to the ethnicity of immigrant populations are provided. The study investigating the role of acculturation, as measured by generational status, on BMI in Canadian youth reported that first-generation immigration was significantly related to higher BMI z-scores in adolescents as compared to second and third generations. Specifically, first-generation immigrants had more weight gain compared to other generation groups in adolescence. These findings expand on a growing literature demonstrating that generational timing of immigration may be a risk factor for obesity, and that maintaining traditional cultural orientation is associated with lower rates of obesity . Although the relationship between first-generation immigration and higher prevalence of overweight has shown mixed results in some previous studies, findings in the article in this issue showed that race and ethnicity were also important factors to consider in understanding these complex relationships. Specifically, analyses stratified by race showed that this relationship tended to be stronger in Black, South Asian, and other first-generation immigrant groups. These findings suggest that first-generation youth may be at greater risk for developing unhealthy lifestyle patterns, such as poor dietary and physical activity habits. However, further research should explore how acculturation and race may be related to barriers for engaging in healthy lifestyle habits. For example, a recent qualitative study by Shelton and colleagues showed that both African American and Latina women reported having many stressors, including economic hardships, demanding caretaking relationships, inflexible work policies, and mistrust of health care providers. Those who were immigrants also reported that they experienced social isolation and were experiencing high levels of stress in general. Thus, health promotion efforts should take into account the additional stressors experienced by ethnic minority women and their families, and how immigration challenges may further impact these difficulties. The study on acculturation and obesity reported in this issue is also one of the first to examine longitudinal patterns of overweight status across generational status in children and adolescents. Overtime, BMI change for first-generation immigrants was associated with greater weight gain, but no direct measures of acculturation were integrated into the study measures, making it difficult to know exactly what the mechanisms are by which generational status is related to the development of obesity in young immigrants. As noted in a growing body of literature, more qualitative studies that address longitudinal and developmental issues related to engaging in healthy lifestyles are needed in underserved and ethnic minority populations . In particular, research on intergenerational effects of culture may provide important insights into family and generational issues that are protective or which may increase the risk for developing obesity. For example, in a qualitative study by Renzaho, Green, Mellor, and Swinburn , Somali and Ethiopian migrant families indicated that intergeneration themes related to immigration included challenges with maintaining traditional parenting, family functioning, and family relations. In addition, these families reported that maintaining a healthy lifestyle of physical activity and diet was challenging, and that parents were more restrictive in controlling their youth's behaviors and discouraged autonomy. Thus, future efforts to develop policies and health promotion programs among immigrant populations should integrate a parenting focus to better assist parents with how to communicate with their youth about the importance of engaging in healthy lifestyles, given the transition to a new cultural environment. Recently, investigators have argued for an ecological approach to understanding lifestyle factors that relate to obesity in underserved and ethnic minority populations that incorporates a stronger focus on social factors, including social context . Recent studies demonstrate that social factors are important for improving health behaviors in underserved low income, ethnic minority populations . In particular, research is needed to better understand social interactions and parent-related supports specific to acculturation of young immigrants who may be at increased risk for developing obesity. Understanding the socialization and cultural integration of immigrants may increase our understanding of how to address the barriers and create more positive social environments that encourage increased activity, healthy eating, and supportive climates for engaging in healthy lifestyles. There is an increasing body of literature that indicates that parent support and positive neighborhood conditions are both important for buffering the negative effects of poverty and poor environmental conditions on the development of obesity in youth . In a recent study by Sussner, Lindsay, Greaney, and Peterson , immigrant Latina mothers reported that beliefs about feeding practices, lack of physical activity, lack of social support, and increased social isolation all contributed to a lifestyle that was prone to developing obesity. Taken together, these studies reviewed highlight the need for health promotion efforts that address the specific barriers related to parental beliefs and understanding of how to build social relationships that could reinforce health behaviors for engaging in physical activity and healthy eating. Health promotion efforts among immigrants need to address multiple systems within the family's environment, including the neighborhood, schools, and religious and cultural affiliations. Health promotion programs for obesity prevention in culturally diverse populations have focused on incorporating cultural tailoring approaches that address the specific needs of the target population with respect to promoting healthy lifestyles. The most successful health promotion interventions for underserved and ethnic minorities have incorporated culturally targeted and culturally tailored intervention components using multisystemic approaches, and these approaches could be applied to young immigrants . Furthermore, in a recent qualitative study by Lindberg and Stevens , Mexican-American women indicated that they wanted more opportunities to engage in behavioral interventions that provided culturally centered behavioral strategies for improving nutrition and food choices within the family context. Kreuter and colleagues have provided an overview of cultural tailoring for improving health behaviors. They propose that interventions can incorporate peripheral strategies that give program materials the appearance of cultural appropriateness by using certain images and pictures of group members. In addition, linguistic strategies have also been used to develop program materials that are dominant to the native culture of the ethnic minorities. Interventions that have ''socio-cultural'' or ''deep structures,'' typically integrate cultural values and norms into the intervention programming and have been demonstrated to be effective tailoring approaches for health behavior change . Deep structures or socio-cultural strategies have also been shown to be successful, when integrated within the health-related context of interventions that target culturally diverse ethnic populations . All of these tailoring strategies could be implemented within immigrant families for promoting healthy lifestyles and obesity prevention. In summary, the article on acculturation and risk of obesity in this issue raises important questions about why and how immigration may lead to greater prevalence of risk for developing obesity in youth. Obtaining contextual information for understanding the social and environmental processes will continue to inform future research for developing effective culturally tailored health promotion approaches in these high-risk populations of immigrants. In addition, developing culturally tailored approaches for improving parent support and social integration, as well as broader environmental supports for healthy lifestyles will be critical to reversing and preventing these trends in childhood obesity among immigrants. Future research should replicate these findings presented in this issue and explore the mechanisms across different ethnic and cultural groups to better understand how to best tailor obesity prevention efforts for immigrant youth. --- Conflicts of interest: None declared.

Understanding risk factors related to acculturation and the risk of developing obesity is important, given the increasing number of immigrants in the United States. A summary of key findings from the ''Generation of Immigration and Body Mass Index (BMI) in Canadian Youth'' article in this issue are presented, and the implications for future research and health promotion efforts are discussed in this commentary. An emphasis is placed on the importance

INTRODUCTION As early as the 17th century there were attempts by international lawyers protect indigenous peoples from the massacres of the Spanish conquistadors in the Americas through legal and moral theories deriving from natural law. 1 Yet these pre-positivist conceptions of natural law could not withstand the legal positivism of the 19th century since positive international law, in its classical form, was only applicable to States. Hence, indigenous peoples were not considered subjects of international law and were not subject to international legal protection at that time. 2 The few remaining guardians of indigenous peoples in the Americas were Christian missionaries such as the Jesuits, whose efforts are viewed as a mixed blessing. They tried to protect the Indians from killing and slavery, but introduced a policy of assimilation, attempting to 'civilise' and 'discipline' them. Since vibrant indigenous cultures were in fact antithetical to the missionaries' aims, 3 not only was indigenous peoples' physical existence threatened, but also their cultural survival. 4 Indigenous peoples have had a tragic history of forced assimilation and deprivation of their lands in many parts of the world. While descendants of European settlers achieved political independence from the colonial powers, indigenous peoples within the former colonies remained the target of discrimination and subject to invasion of their lands . As indigenous peoples did not hold title to land in a way that European legal systems of registration recognised, it was easy for colonial rulers to dispossess them from their ancestral lands. Indigenous tribes were assigned to reservations on marginal land, yet these are rarely sacrosanct even today in light of competing interests of economic operators and national governments. Regrettably, indigenous cultures are still viewed today by some governments as an anachronistic stage in human development. However, the growth of international human rights law has challenged this view, demanding modernisation of classical international law, so as to place individuals and groups of individuals as beneficiaries of international human rights. 6 Although neither the Universal Declaration of Human Rights , the two 1966 Covenants 7 , the UN Covenant on Racial Discrimination 8 , the European, 9 nor Inter-American human rights instruments 10 refer specifically to indigenous peoples, indigenous peoples' human rights have been developed, in particular, in instruments adopted by the International Labour Organisation and the UN General Assembly. 11 This paper focuses on one aspect of post-colonialism: population policies . It assesses the extent to which these policies can be invalidated light of the right to self-determination and to health under international law. This paper argues that the right to 'reproductive self-determination' of indigenous peoples should be recognised as a sub-category of their rights to self-determination and to health under international law. Moreover, this paper argues that government-sponsored population policies must be understood in light of the broader context of State failure to effectively protect indigenous peoples from harm and their land from invasion, which pose a serious threat to the health and long-term survival of indigenous peoples. This paper thus contributes to the literature on indigenous peoples' rights under international human rights law by exploring the limits of their rights to reproductive health and autonomy. In this context, this paper challenges the legality of government-sponsored population policies impacting on indigenous peoples' rights, a field which has been understudied from the perspective of international human rights law. Given that the long-term survival and generational continuity of distinctive indigenous groups depend on the effective protection of their reproductive rights, this is a crucial area for study from the perspective of international human rights law. The second section of the paper discusses estimates of indigenous populations worldwide, and the related problem of defining an 'indigenous identity' that could guide population policies, including government-sponsored family planning programmes. The third section presents an overview of international developments that have led to forceful birth control programmes being imposed upon indigenous peoples, with particular focus on Australia and Peru. It is argued that these past and contemporary policies suggest that indigenous peoples are particularly vulnerable to the impacts of national population control programmes. The fourth section assesses the extent to which substantive norms recognising a human right to self-determination, to health and to reproductive health have been elaborated in international human rights law, which provide the legal basis for the right to 'reproductive autonomy'. The fifth section analyses the procedural guarantees and remedies available for indigenous peoples to assert their reproductive autonomy, in particular the rights to prior and informed consent and to participation in decisionmaking. Further, it discusses the international, regional and national judicial and extrajudicial avenues for the condemnation of and for reparations for forceful interventions in 9 European Convention on Human Rights ETS 5; 213 UNTS 221. 10 See for example, Organisation of American States, American Convention on Human Rights OAS Treaty Series No 36; 1144 UNTS 123; 9 ILM 99. 11 indigenous reproductive affairs. This paper finally concludes that a rights-based approach, as established by the right to self-determination, to heath and to prior and informed consent under international law, provide the essential legal basis for the protection of indigenous peoples' reproductive autonomy. --- POPULATION POLITICS AND THE INDIGENOUS ETHNIC AND CULTURAL IDENTITY Recent estimates have suggested that there are around 370 million indigenous people worldwide, equating to just under 6 per cent of the world's total population. 12 This includes at least 5,000 distinct indigenous peoples in over 72 States. 13 A considerable majority of indigenous peoples are from Asia . 14 There are approximately 40 million indigenous people in Latin America and the Caribbean . 15 It has been further estimated that there are around 50 million indigenous peoples in Africa. 16 In the US it was estimated in 2010 that there are approximately 5.2 million indigenous people . 17 A 2011 National Household Survey estimated that the indigenous population of Canada is just over 1.4 million . 18 There are approximately 670,000 indigenous peoples in Australia, or 3.0 per cent of the total population. 19 In New Zealand, the Māori represent 15 per cent of the total population of 4.3 million. 20 Yet it should be noted that indigenous peoples do not always form part of minority groups within States. While, for example, indigenous peoples account for only 0.45 per cent of the population of Brazil, Latin America's largest country, they account for the majority of the total population of Bolivia. 21 With a total of over 15 million indigenous peoples, Mexico has the highest indigenous population in absolute numbers in the Americas. 22 The demographic survival of indigenous peoples is continually threatened by repression, dispossession of their lands and, the so-called 'McNeill effect', 23 that is the transmission of diseases arising from the contact of indigenous peoples with white colonial rulers. Despite these challenges, it has been suggested that the indigenous population in some parts of the world is recovering. The Programme for Action of the International Conference on Population and Development 1994 stated that: 24 In some regions of the world, indigenous people, after long periods of population loss, are experiencing steady and in some places rapid population growth resulting from declining mortality, although morbidity and mortality are generally still much higher than for other sections of the national population. In other regions, however, they are still experiencing a steady population decline as a result of contact with external diseases, loss of land and resources, ecological destruction, displacement, resettlement and disruption of their families, communities and social systems. Latin America and the Caribbean, for example, present a picture of population recovery. Citing several authoritative studies, Montenegro and Stephens provide the following figures: The estimated total population of indigenous peoples before European invasion ranged from 52.9 to 150 million … Within 100 years, the estimated total Indigenous populations dropped… to 11 million. [Henceforth] in the 18 th Century Indigenous populations represented merely 1.6% of the total population. However, some demographic recovery seems to have taken place: in 1960, the total Indigenous population of Latin America and the Caribbean was estimated as 14.1 million; by 2003 it was more than 48.4 million. 25 Nevertheless, they point out that even though the total indigenous population may have recovered, there are fewer distinct groups. 26 For example, in Brazil the total number of indigenous groups is estimated to have fallen from around 1,000 to 222 in 2006. 27 It has been suggested that the development of self-definition of 'indigeneity' may help explain the apparent recovery in population numbers in some parts of the world. 28 Indeed, estimates of indigenous populations vary according to the way 'indigeneity' is defined and measured, and depends on the extent to which indigenous peoples distinguish themselves from the rest of the population. In this vein, it is their distinctive language that has been used to define 'indigeneity' in most census counts. 29 Language is a paramount means of self-identification and group-identification, and in Latin America alone there are roughly 400 different indigenous languages. 30 On the other hand, it has been suggested that official census estimates of indigenous populations in some countries is likely to be distorted given that discrimination can lead to underreporting of indigenous identity. 31 Hence, the way in which 'indigeneity' is defined and measured has a major impact on estimates of indigenous populations, and will ultimately influence national policies of fertility control. 32 The subject of population politics has become a taboo for much of the past century, not least because of practices such as sterilisation and other controversial birth control programmes as discussed in this paper. Malthus claimed in 1798 that population growth would outstrip food supplies, 33 leading to a continued debate as to the extent to which government intervention is necessary to control population growth. While Cordorcet claimed in 1795 that targeted and voluntary change in reproductive behaviour could lead to a reduction in population growth, 34 post-Malthusians have claimed that State intervention was required to control population growth. 35 The food crisis in 2008 led to a degree of scepticism in the post-Malthusian theory that technological development would invariably match any deficiencies in food supplies, and thus challenges the idea that population growth would not inevitably outstrip food supplies due to changes in technological development. 36 Moreover, concerns over climate change and migration have led to a recurrent debate on global population policies. In this regard, although it was rejected on evidentiary grounds, the petition submitted before the Inter-American Commission of Human Rights by the Inuit Peoples against the US Government 37 for its green house gas emissions provides an indication of how States may be held accountable for the dislocation of indigenous populations due to the impacts of climate change. 38 There is also litigation before US courts in which indigenous peoples rights are implicated in the context of climate change. 39 Population policies have tended to be top-down and based on aggregated national targets and goals, disregarding the particular situation of indigenous peoples. The UN Millennium Development Goals are based on comprehensive targets and numbers at national and international levels. 40 So in the 1994 ICPD there was consensus that States needed to 'shift emphasis from a top-down imposition of population control 33 See Thomas Malthus, An Essay on the Principle of Population . 34 See Marquis de Condorcet, 'Esquisse d'un tableu historique des Progres de L'Esprit Humain' In Ceuvres de Condorcerm vi 256. This view was generally accepted by Thomas Malthus . 35 See further Shalini Randeria, 'Malthus versus Condorcet: Population Policy, Gender and Culture from an Ethnological Perspective' in S Berking and M Zolkos , Between Life and Death: Governing Populations in the Era of Human Rights 28-29. 36 See Malthus . Indeed, the global food crisis in 2008 which has arisen from the food shortages and the hiking of prices of a great number of agricultural commodities has been largely blamed on the increasing global demand for biofuels and consequent exhaustion of agricultural lands to meet this global demand. measures to community-based programs that better respond to the needs of individuals and families.' 41 Most States monitor progress towards meeting the MDGs targets through aggregate data, therefore the data uses averages from the whole population. 42 Such nationwide population targets fail to take into account that indigenous peoples are in some instances represented by small numbers. A few deaths within indigenous populations are irrelevant for the attainment of a 'national target' of mortality control, but have a significant impact upon groups that are demographically fragile. Therefore, while health indicators can be declining among marginalised communities, the country as a whole may appear to be making steady progress towards meeting a national target. 43 Although there has been resistance to disaggregation of data in population and health indicators, in recent years there has been increasing recognition of importance of data disaggregation. In particular, the 2010 MDGs Summit Declaration highlights the importance of taking account of inequalities within countries and the need for countries to have 'adequate, timely, reliable and disaggregate data' that can be used to improve policies and programmes. 44 This position is reiterated in the more recent Framework of Actions of the ICDP 2014 conference , which had as one its main objectives to assess the implementation of the Programme for Action adopted under the first 1994 international conference on population and development . 45 --- INDIGENOUS PEOPLES AND FORCEFUL BIRTH CONTROL PROGRAMMES It is widely recognised that there is need for lawful government intervention and programmes of birth control, including the distribution of contraceptives and family planning. The State has a duty to guarantee the health of the population and to ensure that resources are available to future generations, which could be undermined by uncontrolled population growth, especially as the rates of fertility tend to be higher in the poorest parts of the world. 46 In fact, it has been suggested that reproductive health and access to family planning is a fundamental human right. 47 Even though reproductive health is not specifically mentioned in the UDHR nor the two 1966 Covenants, it could be seen as 41 See the Programme of Action of the ICPD . 42 forming part of the broader right to health as enshrined in those international agreements.48 Moreover, other civil and political rights could be invoked to uphold reproductive health and freedom, in particular the right to life, 49 the prohibition against inhuman and degrading treatment 50 and the right to private and family life. 51 These rights could be raised to secure safe abortions and childbirth, confidentiality of patients seeking reproductive health services, respect for those living with HIV/AIDS and measures to reduce violence against and trafficking of women. 52 However, the principle of non-coercion is open to abuse, for example, where a State or racist practitioners within the State may seek to control an ethnic or social class within its borders . This could lead to forceful interventions aimed at controlling the population growth of one or more specific communities. This would offend the principle of 'reproductive freedom' defined as 'the individual's choice to reproduce or not to reproduce.' 53 Eugenic practices leading to medical and scientific experimentation amount to acts of torture and inhuman and degrading treatment, as recognised under the 1966 International Covenant on Civil and Political Rights. 54 Indigenous peoples have been targeted by forceful state-sponsored sterilisation and other coercive birth control programmes that violate their reproductive rights. A notorious state-sponsored programme led to the mass sterilisation of nearly 250,000 poor women by the Fujimori Government in Peru between 1996 and 2000. 55 The Peruvian Congress legalised sterilisation for family planning shortly after Fujimori announced in his second inaugural speech in July 1995 that family planning would be a priority for his Government. 56 The programme was undertaken in the remote, rural areas of the Andean Sierra and Peruvian Amazon region, which are heavily indigenous. 57 Although Peru's family planning public policy was officially considered to be part of a voluntary programme, the women underwent sterilisation without a proper consent process and monthly quotas led doctors to forcibly sterilise women. 58 Moreover, some health workers did not provide women with information regarding other available birth control methods and many times deliberately gave inaccurate information about the risks and consequences of surgical sterilisation. 59 As put by Jaichand and O'Connel, 'the Peruvian State ultimately viewed family planning as a means to reduce poverty rather than to promote women's rights.' 60 The Peruvian Human Rights Ombudsman Office played a vital role in documenting gross human rights violations during the implementation of the programme. Yet it was not until President Alejandro Toledo assumed power in 2001 that formal investigations were launched into the reproductive health programme. The findings of the two investigations were that gross violations of women's reproductive rights had taken place and were thus denounced. Further, the Ombudsman Office issued another Report in September 2002, 61 which found that, in the case of sterilisations, local health-care workers were required to meet quotas, creating incentives for coercive and negligent behaviour. 62 In another example, the Australian Government's forceful removal of Aboriginal children from their families between 1900 and 1969 breached the fundamental rights of the indigenous populations. 63 The 1997 'Bringing Them Home' report found that at least 100,000 Aborigines, in particular mixed race children from white and Aboriginal parents, were taken away from their parents and placed in the care of institutions, religious missions or white foster families. 64 The policies were supported by a legal framework particularly disadvantageous to Aboriginal peoples. Until the 1960s, the 'White Australia' policy encouraged white settlement, while a 'protectorate' system was applied to indigenous peoples. This 'protectorate' system prohibited indigenous peoples from marrying and forming families without the consent the State. 65 Responsibility for Aboriginal affairs, in effect, remained with the former colonies, which became State governments after federation. 66 Since British colonisation happened with almost no recognition of the existing legal order, Aboriginal peoples were not even counted in the census. 67 The notion of terra nullius was not rejected until 1992 when the Australian 59 High Court delivered the Mabo v Queensland 68 judgment recognising the aboriginal land rights of Torres Strait Islanders. 69 Other well-documented examples of forceful interference with indigenous peoples' reproductive rights are the US Government-sponsored sterilisation programmes of indigenous women in the 20th century. 70 In each of the above cases, the dominant society argued that it was acting in the national interest and for the communities' own good, though the indigenous people themselves are dissenters from this view. Other States have tried to control childbirth through coercive programmes, notoriously China's one-child policy , 71 and the sterilisation abuses in India in the 1975 'emergency' and the subsequent re-emergence in India of provider targets and disincentive schemes. 72 These practices are in direct violation of the 1994 International Conference on Population Development 73 and the 1995 Beijing World Conference on Women Declaration, 74 which held non-coercion to be fundamental to population and reproductive health programmes and that 'coercion in any form is unacceptable.' 75 Forceful child removals are expressly condemned in the 2007 UN Declaration on the Rights of Indigenous Peoples. 76 Yet according to the 2014 Framework of Actions of the ICDP, a considerably high number of States have not implemented and enforced national laws against coercion, including forced sterilisation. 77 Reproductive policy should 'address both the socio-economic forces which contribute to high rates of fertility as well as offer comprehensive reproductive health 68 Mabo v Queensland 166 CLR 186 Australian Law Reports 107; and Mabo v Queensland HCA 23; 175 CLR 1. 69 See Sanders . 70 There is an extensive literature on the subject, see for example Mary Dog Crow, Lakota Woman 75 Babor 118. 76 See Article 7 para 2: '[i]ndigenous peoples have the collective right to live in freedom, peace and security as distinct peoples and shall not be subjected to any act of genocide or any other act of violence, including forcibly removing children of the group to another group' . Resolution adopted by the General Assembly on 13 Sept 2007, United Nations Declaration on the Rights of Indigenous Peoples 61/295. 77 'Only 60 per cent of countries have promulgated and enforced a national law protecting against coercion, including forced sterilization and forced marriage; this proportion is lowest in the Americas .' See Framework of Actions for the follow-up to the Programme of Action of the International Conference on Population and Development Beyond para 317. care and informed choice for every inhabitant,' 78 but this is antithetical to the use of force. Policies focused on birth quotas and sterilisation inevitably conflict with the basic right of women to decide how many children they will bear. There may be need for limited incentives in order to guarantee the overall effectiveness of family planning programmes -this is particularly so in light to prevent the spread of sexual infectious diseases among indigenous peoples. Yet a state-sponsored practice of forceful sterilisation would most clearly contravene a State's obligations under international human rights law. 79 While this paper focuses on state-sponsored programmes that violate indigenous peoples' reproductive rights, this is not to say that the problem is limited to indigenous peoples. Women in general and women belonging to minorities groups, in particular, have suffered violations of their reproductive rights in many parts of the globe, including recently the Roma minority in Slovakia and Hungary. 80 Therefore, the analysis of violations of indigenous peoples' reproductive rights developed in this paper will include a discussion of the international legal mechanisms for protection of women's and minorities' rights in general, including decisions of international and regional courts and treaty-monitoring bodies relating to the protection of reproductive rights. It is pertinent that this paper focuses on the question of indigenous peoples' reproductive rights given the distinctive way in which their human rights have developed under international law. This paper argues that it is fundamental that international law recognises the illegality of practices that violate indigenous peoples' reproductive rights; and establishes effective mechanisms for the prevention of and reparations for past population policies that have breached indigenous peoples' fundamental rights. --- SUBSTANTIVE RIGHTS IN SUPPORT OF INDIGENOUS PEOPLES' RIGHT TO REPRODUCTIVE AUTONOMY This section of the paper examines the extent to which certain substantive rights under international law could define the basis for indigenous peoples' reproductive autonomy. It discusses in turn the development of the rights to self-determination, to health, as well as the right to reproductive health, as a sub-category of the right to health. --- INDIGENOUS PEOPLES' RIGHT TO SELF-DETERMINATION UNDER INTERNATIONAL LAW The right of self-determination could be a powerful basis for securing an element of autonomy and self-government for indigenous and minorities groups . Self-determination allows peoples 81 'the right, in full freedom, to determine, when as they wish, their internal and external political status, 78 Babor 118. 79 ibid at 114. 80 without external political interference, and to pursue as they wish their political, economic, social and cultural development.' 82 Self-determination only emerged as a principle of international law at the San Francisco Conference preceding the draft UN Charter and has been included in both articles 1 and 55 of that instrument. 83 In addition, an express reference to the right of self-determination can be found in article 1 of both UN 1966 International Covenants on Civil and Political Rights 84 and Economic, Social and Cultural Rights. 85 Yet neither the UN Charter nor any other human rights instrument define who the 'peoples' entitled to be beneficiaries of self-determination are. 86 It could be argued that if self-determination in international law is a right accruing to 'all peoples' then indigenous people should be recognised as beneficiaries of that right. Given the legal and political constraints for the recognition of the right to secession , 87 it is the right to internal self-determination that is more realistically achievable and thus could provide the legal basis against undue interference to the physical well-being and reproductive autonomy of indigenous peoples. James Anaya, the UN Special Rapporteur on Indigenous People between 2008-2014, 88 signalled that governments have increasingly moved away from the tendency to equate the word self-determination with an absolute right to form an independent state. 89 Therefore, if external self-determination is not, realistically, available nor politically feasible, the question arises as to whether another form of self-determination can be substituted for its external application. 90 The UN Human Rights Committee used the term 'internal self-determination' for the first time 91 and it is now the terminology followed by most authors referring to forms of self-government, autonomy, territorial integrity or exclusive enjoyment of the indigenous lands and resources. Significant progress towards the recognition of international rights for indigenous peoples happened with the adoption of UNGA Declaration on the Rights of Indigenous Peoples on 13 September 2007, following 20 years of difficult negotiations. The right to internal self-determination of indigenous peoples is at the heart of the Declaration, which defines it as 'the right to autonomy or self-government in matters relating to their internal and local affairs, as well as ways and means for financing their autonomous functions.' 93 Although resolutions of the UNGA are not legally binding, given that the UNDRIP was adopted by so many States with so few objections and abstentions, it may very well attain the status of customary international law. 94 Although Australia, Canada, New Zealand and the US had originally withheld their support for it, those four countries subsequently endorsed the UNDRIP. 95 Even if not accepted that the UNDRIP could be said to represent customary international law, a number of rights enunciated in it already form part of customary international law. 96 Indeed, a number of indigenous rights enshrined in the UNDRIP have been recognised by international courts and tribunals as having evolved into customary international law. 97 Self-government is 'the overarching political dimension of ongoing selfdetermination.' 98 The emergence of values such as democracy and the major notions of cultural and political pluralism have reinforced indigenous claims for governmental and administrative autonomy for their communities, which include claims for reproductive health and generational continuity. The idea in many societies that decisions should be made at the most local level possible reinforces the view that indigenous communities should be able to maintain their traditional decentralised system of governance and to 93 make decisions relating to fertility in their communities. In this respect, indigenous peoples should maintain their own institutions of autonomous governance, including their customary and written law and dispute resolution and adjudication mechanisms, which have existed not only de facto, but as the indigenous peoples correctly claim, also de jure, as recognised in the ILO Convention on Indigenous Peoples no. 169. 99 In order to give effect to the right to internal self-determination, effective procedural rights must be established under international and national law, including the indigenous right to prior and informed consent and to participation in decision-making. 100 International and regional human rights bodies have played a significant role in recognising indigenous peoples' right to internal selfdetermination, in particular by recognising the obligation of States to protect indigenous peoples' land and natural resource rights and to respect and protect the values, customs and traditions of indigenous communities. 101 Some of the most prominent regional developments are the decisions by Inter-American human rights bodies, 102 which have taken a progressive stance on indigenous peoples' land rights. 103 For example, in Sawhoyamaxa Indigenous Community v Paraguay 104 the Inter-American Court of Human Rights ruled that Paraguayan legislation failed to provide an effective judicial remedy that protected legitimate land claims by indigenous communities. 105 This constituted a violation per se of the American Convention on Human Rights , 106 and the displacement and expropriation of indigenous' lands amounted to a violation of the right to life. 107 Importantly, in Saramaka People v Suriname, 108 the IACtHR referred for the first time to the right of self-determination in its interpretation of indigenous land and resource rights under article 21 of the ACHR. 109 Moreover, in the more recent decision of Kichwa Indigenous People of Sarayaku v Ecuador, 110 the Court found that Ecuador had violated article 21 of the Convention by awarding a private company an oil exploration and exploitation concession that partially covered the ancestral lands of the Kichwa people of Sarayaku, without a consultation process or their free, prior and informed consent. 111 The right to internal self-determination is applicable to the context of policies for fertility control of indigenous peoples. Given the specific vulnerability of indigenous peoples in most societies and what makes them distinct as a people, the right to internal self-determination must also include the indigenous right to reject government policies that coercively subjects them to birth control programmes. From this perspective, autonomous governance is not only instrumental but also necessary for indigenous peoples to control the development of their distinctive culture. This includes the use of land, natural resources and reproductive health, against undue interference from disingenuous powerful economic interests or governments, as recognised in the 1994 Programme for Action 112 and the Framework of Actions adopted by the 2014 ICDP. 113 --- THE RIGHT TO HEALTH AS A HUMAN RIGHT Human rights are an increasingly important tool for the advancement of social justice. They provide the guiding principles to government agencies promoting social justice and provide individuals and groups with a powerful tool to ensure that government agencies commit themselves to human rights principles. 114 The right to health is also recognised in some countries as a constitutional right. 115 The right to reproductive health is a sub-category of the human right to health. The right to health was first articulated at the international level under the Constitution of the World Health Organisation 1946, and subsequently established in several binding international human rights treaties. 116 The right to health contains a number of elements that can be legally enforceable under national law, for example the right to non-discrimination in relation to health facilities, goods and services. 117 Yet at the international level the right to health goes far beyond discrimination within health care provision. 118 A rights-based approach to health indicators has been proposed in order to enable progressive realisation of the right to health to be monitored and measured. 119 It has been argued that a human rights approach to health could have several advantages for health care practitioners, for example as the basis for equitable policies and programmes that benefit the most disadvantaged, raising more funds and improving terms and conditions of those who work in the health sector. 120 The UN Special Rapporteur on the right to health pointed out in UNGA Resolution A/59/422 121 that there were profound disparities between the health of indigenous peoples and that of the non-indigenous population in many countries, the corollary of which is that 'indigenous people tend to die younger and generally live in poorer health than other population groups'. 122 It was further suggested that: In some jurisdictions, they are more likely to have chronic disorders such as diabetes, high blood pressure or arthritis, and are more prone to substance abuse, depression and other mental disorders than are nonindigenous people. Suicide rates among indigenous women in certain developed countries are as high as eight times the national average. HIV/AIDS and other sexually transmitted diseases are spreading rapidly in indigenous communities, a trend fuelled by factors including social and economic exploitation of indigenous women, as well as a lack of access to health related information. 123 Further evidence of health inequalities includes: discrimination by some health professionals ; the lack of health services available in indigenous languages; the lack of clean drinking water and adequate sanitation; the impact of environmental contamination on the health of indigenous communities; and violence, including sexual violence, against indigenous women and children. 124 The Rapporteur's 2004 report to the UNGA also noted the systematic inequality in access to medical services and in the quality of these services; the marginalisation of traditional medicine of indigenous peoples; high rates of diseases such as diabetes; and high suicide rates, particularly among young indigenous 117 men. Maternal and child mortality rates can be twice as high as the national average in indigenous communities . 125 Concerns over the state of indigenous peoples' health were reiterated in the third session of the Permanent Forum on Indigenous Issues, 126 which also highlighted the need to address the right to health of indigenous women, including sexual and reproductive rights. Furthermore, the World Health Organisation has published a number of reports specifically addressing health issues facing indigenous peoples, 127 urging States to 'protect the right of Indigenous people to the enjoyment of the highest standard of health.' 128 The WHO Commission has stated that 'every government should be assessed for its impacts on health and health equity; to make health and health equity a marker for government and economic performance.' 129 Importantly, the WHO has emphasised the importance of data disaggregation in order for policy-makers to effectively monitor the health of indigenous peoples. 130 It is also significant that the ILO no. 169 Convention 131the only international treaty specifically aimed to protect indigenous peoples' rights -recognises the right to health for indigenous peoples and that they 'may enjoy the highest attainable standard of physical and mental health.' 132 Yet it should be noted that, with 22 ratifications at present, the ILO n.169 Convention is not a widely ratified international treaty. 133 The adoption of the 2007 UN Declaration on the Rights of Indigenous Peoples , in particular the health-related Articles 22, 24 and 31 thereof, could further add support to indigenous peoples' claim for self-determination over their health, well-being and development. Even though, as discussed above, the Declaration is not legally binding, it could reflect States' opinio juris and practice, and thus provide the basis for of its status as customary international law. 134 133 The ILO 169 Convention was adopted on 17 June 1989 and entered into force on 5 September 1991.. It is notable that many Latin American states with large indigenous populations have ratified the convention. 134 See Claire Charters, 'Indigenous Peoples and International Law and Policy' 18 Public Law Review 1. 135 See for example the decision of the New Zealand Court of Appeal in Ngai Tahu Maori Trust Board v Director-General of Conservation [1995] 3 NZLR 553. See also ibid. Given that most estimates suggest that the MDGs 2015 global health targets are likely to be missed, 136 governments are now focusing on post-2015 health targets. This paper contends that the right to reproductive autonomy and health need to be further developed in the post-2015 agenda. Calls have been made for the indicators set to measure progress in the post-2015 framework to be broken down by ethnic and cultural groups. 137 Moreover, the protection of the natural environment and land belonging to indigenous peoples should be seen as an integral aspect of protecting their right to health, as indigenous peoples' well-being is traditionally linked to the conservation of the ecosystem and community well-being. The indigenous right to land and to natural resources are firmly recognised in both the ILO n.169 Convention and the UNDRIP, 138 and have been upheld in decisions of international treaty monitoring bodies and regional courts. 139 The need to improve indigenous peoples' access to health services is often constrained by financial, geographic and cultural barriers. 140 Access to 'culturally appropriate' health services is necessary for indigenous peoples to gain full health rights. 141 142 In some countries including Australia, New Zealand, Canada and Colombia, a further step has been made towards appropriate services where indigenous-controlled services exist with indigenous medicine practiced alongside mainstream medicine. 143 On the other hand, the recognition of indigenous traditional customs, traditions and healing practices under national health care programmes could pose a number of challenges. This can be illustrated in the case of South Africa, which has recently taken steps to formalise the role of traditional healers. In birth attendants and herbalists. 144 While one benefit of traditional healers is that they are often more likely to be contacted than doctors applying mainstream therapies , there are disadvantages. For example, 42 deaths and hundreds of injuries to Xhosa boys from botched circumcisions were reported in the first six months of 2012. 145 This has led to calls for stronger regulations applicable to traditional healers in South Africa. The regulation of traditional healing practices, on the other hand, could be in direct conflict with the need for universal access of health services that recognises the needs and priorities of indigenous peoples. It is thus paramount that those two often-conflicting interests are carefully balanced under national health care laws and policies. --- THE RIGHT TO REPRODUCTIVE HEALTH Family planning programmes are imperative public policy mechanisms for tackling reproductive and sexual health problems, given that they are preventive in nature. Even though not all sexual and reproductive ill-health represents a violation of the right to health or other human rights, it could constitute a violation of those rights when it arises, in whole or in part, from the failure of a State to respect, protect or fulfil a human rights obligation. 146 Under international law, the concept of 'family planning' has been recognised as a basic human right in relatively recent times. Specifically, the term was used in the 1965 UN Commission Resolution on the Status of Women, which states that 'married couples should have access to all relevant educational information concerning family planning.' 147 Yet as Ericsson points out, the term was not applied in the context of human rights until the 1968 UN Conference on Human Rights. It was held then that 'parents have a basic human right to determine freely and responsibly the number and spacing of their children.' 148 Although neither the ILO no. 169 Convention nor the UNDRIP contain legal provisions recognising a specific right to reproductive health to indigenous peoples, the right is recognised in other international human rights instruments. In this regard, the Commission on Human Rights confirmed in Resolution 2003/28 that 'sexual and reproductive health are integral elements of the right of everyone to the enjoyment of the highest attainable standard of physical and mental health.' 149 In a similar fashion, principle 8 of the 1994 ICPD Programme of Action states that 'everyone has the right to the enjoyment of the highest attainable standard of health', which includes the State's obligation to provide access to health care services 'related to reproductive health care, which includes family planning and sexual health.' 150 The 1994 ICPD Programme of Action further affirmed the 'basic right of all couples and individuals to decide freely and responsibly the number, spacing and timing of their children and to have the information and means to do so, and the right to attain the highest standard of sexual and reproductive health.' 151 In 1995 the Platform for Action of the Beijing World Conference on Women adopted similar provisions on reproductive health. 152 The main difference between the two instruments is that the Beijing Declaration confirms the right of women to control their sexuality. 153 This obligation of States to protect women's reproductive rights was generally restated in the 2014 Framework of Actions of the second Cairo conference. It declares that: States should guarantee indigenous peoples' right to health, including their sexual and reproductive health and rights, as well as their rights to both the highest standard of care and the respectful accommodation of their own traditional medicines and health practices, especially as regards reducing maternal and child mortality, considering their socio-territorial and cultural specificities as well as the structural factors that hinder the exercise of these rights. 154 Yet the report has criticised States' implementation of this obligation, noting that a considerable number have not addressed the issue of providing 'culturally appropriate' sexual and reproductive health care, including HIV prevention services for indigenous peoples. 155 It is also a cause for concern that some States in the Americas and Asia, despite having emphasised the right to health as a priority for indigenous peoples, have simultaneously excluded 'sexual and reproductive health' when listing areas that they prioritise concerning indigenous peoples. 156 The most important reference to family planning as a human right is however in the Convention on Elimination of All Forms of Discrimination against Women . CEDAW proclaims that women and men have equal right to decide 'freely and responsibly on the number and spacing of their children and to have means to enable them to exercise this right.' 157 The reference in CEDAW to family planning is particularly significant since, unlike the aforementioned UNGA resolutions and the ICDP Action Programmes, CEDAW is a legally binding international treaty. CEDAW is widely 150 1994 ICPD Programme of Action, principle 8. 151 ratified , including Australia and Peru, although notably the US is not a party. 158 The CEDAW Committee 159 has raised the prospects of 'family planning' being elevated to the status of an international human right. In 1993, the Committee's Recommendation stated that women have 'the right to the number and spacing of their children.' 160 In general, however, the Committee avoids the term 'family planning', which suggests that it may have become outmoded and the need for it to be replaced by the more comprehensive term 'reproductive health.' 161 In support of the right of reproductive autonomy for women, the Committee held in General Recommendation no. 19 that '[c]ompulsory sterilization … adversely affects women's physical and mental health, and infringes the right of women to decide on the number and spacing of their children.' 162 The Programme of Action of the ICPD conference defines under Chapter VII, par. 7.2 reproductive health as 'as state of complete physical, mental and social well-being and not merely the absence of disease or infirmity, in all matters relating to the reproductive system and to its functions and processes', further affirming that the definition implies that people must 'have the capability to reproduce and the freedom to decide if, when and how often to do so.' 163 It calls on governments and other important institutions to recognise the distinct perspective of indigenous people in relation to population and development. All human rights violations and discrimination, especially all forms of coercion, must be eliminated. 164 Moreover, the Programme of Action of the 1994 ICPD 165 declares non-coercion to be fundamental to population and reproductive health programmes and that 'coercion in any form is unacceptable.' 166 Yet the more recent 2014 ICDP Framework of Actions has noted the slow progress of some States in adopting implementing and enforcement measures to ensure full enjoyment of reproductive health rights. 167 The right to reproductive health, which includes the right to family planning, thus includes both a positive and negative State obligation. The positive obligation involves the requirement for States to provide all the essential health care facilities and services, while the negative obligation revolves around the State's need to abstain from interfering 158 The Committee on the Elimination of All Forms of Discrimination Against Women has been transferred on 1 January 2008 to the Office of the High Commissioner on Human Rights in Geneva. See Eriksson 183. 159 On the powers of the CEDAW Committee, see further, section 5.2.1. below. 160 Recommendation No 21. See also, Ericsson who argues that the Resolution is 'the only authoritative interpretation of the right to family planning.' 184. 161 ibid at 185. In its Fifty-second Session on 25 February -7 March 2008 the CSW has adopted resolutions on 'Ending Female Genital Mutilation' E/CN.6/2008/L.2/Rev.1 and on 'Women, the Girl Child and HIV/AIDS' E/CN.6/2008/L.5/REV.1 162 CEDAW Committee, General Recommendation No 19. See further section 5.2.1., which discusses communications brought before the Committee. 163 See 2014 ICPD Framework of Action . 164 Programme of Action of the 1994 ICPD, para 6.25 . 165 A/CONE 171/13. 166 Babor 118. 167 'In the area of sexual and reproductive health and reproductive rights, less than two thirds of countries have promulgated and enforced a law protecting the right to the highest attainable standard of physical and mental health, including sexual and reproductive health. See 2014 ICPD Framework of Action . with one's individual reproductive choices through coercive population control programmes. The 1994 and 2014 ICPD Programmes of Action and Beijing Declaration and Platform for Action are, alongside the CEDAW, the most influential and authoritative international documents on the right to reproductive and sexual health. 168 Furthermore, even though the subsequent MDGs do not expressly refer to sexual and reproductive health, at least three of the eight Goalson maternal health, child health and HIV/AIDSare relevant to sexual and reproductive health. 169 Other international initiatives recognising reproductive health rights for indigenous peoples include the outcome document of the 2014 World Conference on Indigenous Peoples, 170 and the conclusions of the World Conference on Indigenous Women in Lima, Peru, in 2013. 171 The 59th session of the Commission on the Status of Women that took place in New York from 9 to 20 March 2015 focused on the current challenges that affect the implementation of the 1995 Beijing Declaration and Platform for Action. 172 In addition, a number of regional treaties have been adopted affirming, expressly or indirectly, women's reproductive health rights. Particularly important is the Inter-American Convention on the Prevention, Punishment and Eradication of Violence Against Women , 173 which affirms that 'violence against women constitutes a violation of their human rights and fundamental freedoms' and establishes the state duty to protect woman against violence. 174 Although women's reproductive rights are not listed in Article 4 of the Convention, that provision is not supposed to be exhaustive. As such, the Convention can be interpreted to also protect women's reproductive rights in its provisions concerning the protection of women against violence. 175 Within the Council of Europe framework, the Convention on Human Rights and Biomedicine requires that an intervention in the health field cannot be carried out without the free and informed consent of the person concerned. This must be done on the basis of objective information, provided without any pressure, regarding the nature and consequences of the medical intervention and its alternatives. 176 Perhaps the most significant regional treaty concerning the right to reproductive health is the Protocol to the African Charter on Human and Peoples' Rights on the Rights of Women in Africa . 177 This Protocol entered into force on 25 November 2005 and has so far been ratified by 36 African Union States. 178 The Protocol recognises 'health and reproductive rights' for women in Africa including 'the right to control fertility' and 'the number of children and spacing of children.' 179 It seems clear from the foregoing that there is a relatively well-developed body of international norms and instruments recognising the right to reproductive health as a human right, although the majority of these instruments are non-binding. CEDAW is one of the few exceptions as an international treaty recognising the right to reproductive health and autonomy of women. However, one of its key weaknesses is that, from the broader perspective of population policies, CEDAW has a more limited scope as it aims to affirm women's rights, and hence forced male sterilisation and forced child removals are not expressly outlawed under CEDAW. To conclude, the rights to self-determination, to health and to reproductive health provide the essential substantive basis for the affirmation of indigenous peoples' reproductive autonomy. The next section assesses the extent to which there are mechanisms under international law for these rights to be adjudicated and protected. --- PROCEDURAL RIGHTS FOR ASSERTION OF INDIGENOUS PEOPLES' RIGHT TO REPRODUCTIVE AUTONOMY The substantive norms discussed in the previous sections could be easily disregarded by States if there were no effective legal and extra-judicial mechanisms for enforcement and settlement of disputes. This section examines the procedural rights that exist internationally, regionally and nationally for the assertion of indigenous peoples' reproductive rights and for reprehension of government population policies that infringe those rights. --- THE RIGHT TO PRIOR AND INFORMED CONSENT AND TO PARTICIPATION IN DECISION-MAKING UNDER INTERNATIONAL LAW International law recognises the principle that indigenous peoples should be consulted in decisions made by national authorities and others that could affect them. 180 The right to consultation and to prior and informed consent should serve as a basis for indigenous peoples to reject policies that subject them to birth control programmes. Indeed, the right to free, prior and informed consent is at heart of cases regarding reparations for victims of involuntary sterilisation programmes. 181 The right to consultation is enshrined in ILO Convention 169 182 which employs different standards ranging from consultation to participation and informed consent. 183 According to the Convention, consultation must be undertaken 'in good faith and in a form appropriate to the circumstances, with the objective of achieving agreement or consent.' 184 Indigenous peoples' participation in the broadest level of governance must not supplant local participation in connection with specific issues or projects that impact on their well-being. 185 This right is further recognised in the UNDRIP, which defines the right of indigenous peoples to participate in decision-making in matters that could impact on their rightsthrough representatives chosen by them in accordance with their own proceduresas well as the right to maintain and develop their own indigenous decision-making institutions. 186 Before adopting and implementing legislative or administrative measures that may affect them, 187 States must consult and cooperate in good faith with the indigenous peoples concerned through their own representative institutions in order to obtain their free, prior and informed consent. In particular, the free, prior and informed consent of indigenous peoples needs to be defined in accordance with their customary laws and practices. 188 Both the ILO Convention 169 and the UNDRIP thus recognise that consultation is an obligation when indigenous peoples' fundamental rights are concerned. 189 Those developments have led Anaya to argue that 'widespread acceptance of the norm of consultation demonstrates that it has become part of customary international law.' 190 The extent of this duty to consult has been intensely debated. In particular, it is contended that the right of indigenous peoples to participate in decision-making should include the right to veto decisions affecting them . 191 Aside from the circumstances that would involve relocation, 192 the ILO Convention 169 generally falls short of requiring the consent of indigenous peoples, and instead establishes the principle of consultation and that they have the right to participate in decision-making. 193 The Convention also incorporates indigenous peoples' right to participate in decision-making in matters that would affect their rights 194 before the adoption and implementation of such legislative or administrative measures. 195 The international jurisprudence and State practices advancing the right to prior and informed consent for indigenous peoples has been developed particularly in the context of relocation and the right to participate in the context of extractive industry projects. This right must also be recognised in the context of other coercive policies impacting on indigenous peoples well-being, including sterilisation and birth control programmes. Indigenous peoples' participatory rights regarding state health and population policies are expressly recognised in the 1994 Programme of Action to the ICPD, which called for governments to 'incorporate the perspectives and needs of indigenous communities into the design, implementation, monitoring and evaluation of the population, development and environment programmes that affect them. ' 196 Beyond those international initiatives, there are specific regional instruments recognising the principle that a medical intervention in to one's health cannot be carried out without the free and informed consent of the person concerned. 197 There remain many barriers to implementation of the right to prior and informed consent in national legislation, which include inadequacy of laws and regulations, lack of articulated community procedures and the lack of will to facilitate access by some local communities. 198 The requirement that consent be 'free' suggests that it should be obtained without coercion, intimidation or manipulation. Secondly, 'prior' suggests that the consent is sought sufficiently in advance of any authorisation or commencement of activities and respect time requirements of indigenous consultation/consensus processes. Finally, 'informed' requires that there be sufficient information regarding the nature, size, pace and scope of any proposed project or activity. 199 The need to obtain a patient's prior and informed consent has been paramount to the development of international jurisprudence concerning forced sterilisations. For example, in V.C. v. Slovakia the European Court of Human Rights held that 'where sterilisation was carried out without the informed consent of a mentally competent adult, it was incompatible with the requirement of respect for human freedom and dignity.' Therefore, it was in breach of Article 3 ECHR on the prohibition on torture and inhuman or degrading treatment; and Article 8 ECHR on the right to private and family life. 200 Moreover, as discussed above, the lack of free, prior and informed consent were paramount to the findings by the Peruvian Human Rights Ombudsman Office that the Peruvian sterilisation programme violated indigenous women's reproductive rights. 201 The procedures for obtaining indigenous peoples' consent should be 'culturally appropriate,' in line with the international initiatives concerning the right to health. 202 In this regard, language is often a barrier in the context of access to sexual and reproductive health services. 203 For example, it has been suggested that in some languages there is no term for fistula , making open discussion difficult and perpetuating stigma and discrimination and restricting access to services. 204 It is thus recommended that mutual forms of communication be implemented between service providers and users, taking into account local languages and customs. --- THE RIGHT TO ACCESS TO JUSTICE: THE JUDICIAL AND EXTRA-JUDICIAL AVENUES FOR ASSERTION OF INDIGENOUS PEOPLES' REPRODUCTIVE HEALTH RIGHTS The right of individuals and communities to bring effective mechanisms for justice and redress is a fundamental aspect of the evolution of international human rights law. International courts generally fail to establish a right of standing to individuals, groups of individuals or corporations to bring legal proceedings for defence of individual or collective rights. In particular, the International Court of Justice only has jurisdiction over disputes between states. 205 Despite this, this section discusses a number of other international, regional and national forums for indigenous peoples to bring claims for violations of their reproductive health rights. This section examines in particular their rights to access to justice before international human rights treaties monitoring bodies; the ILO Convention 169 Compliance Committee; two regional human rights systems; international and national criminal courts; as well as alternative mechanisms to litigation, such as truth and conciliation commissions. --- UN Human Rights Treaties' Supervisory Bodies The supervisory or monitoring bodies established under specific human rights treaties offer a level of protection against violations of indigenous peoples' reproductive rights. One example is the CEDAW Committee, which may be called upon in cases involving violations of reproductive health and indigenous women's rights. In addition to recommendations to State reports, Optional Protocol to CEDAW, 206 introduces a system whereby communications detailing violations of the Convention may be submitted to the Committee. Both Australia and Peru are parties to this Optional Protocol. The CEDAW Committee has focused on problems faced by indigenous women in its observations to State reports, which have included recommendations regarding sexual violence against indigenous women. 207 The disrespect for indigenous women's human rights and violence against them are part of a general pattern of gender injustice and neglect by governments, while the international indigenous rights movement itself often turns a blind eye to the issue. 208 On August 2006 the CEDAW Committee found for the first time that restrictions on women's reproductive freedom constituted a human right violation, and decided that Hungary had violated the Convention by not ensuring that a Roma woman who was sterilised in January 2001 had given her fully informed consent. 209 The Committee considered that she has a right protected by article 10 CEDAW to specific information on sterilisation and alternative procedures for family planning in order to guard against such an intervention being carried out without her having made a fully informed choice. 210 The Committee found that a failure by the State -through the hospital personnel -to provide appropriate information and advice on family planning constituted a violation of the complainant's right under article 10 CEDAW. 211 The Committee further recommended that Hungary introduce legislation on informed consent. 212 As Peru only ratified the Optional Protocol in 2001 . communications could not be submitted to the Committee. 213 However, in its observations on Peru's report the Committee explicitly condemned the sterilisation practices in August 2002, expressing its concern over the 'numerous cases of sterilisation of women without prior informed consent, using psychological violence or the promise of financial incentives, thus affecting women's right to decide the number and spacing of their children' and its regret that 'those responsible have not been punished.'' 214 In a more recent case that demonstrates CEDAW's scope in dealing with reproductive health rights of minorities or ethnically discriminated groups, the Committee held that the lack of access to quality medical care during pregnancy was a systematic problem in Brazil. It found that an Afro-Brazilian woman who died due to the lack of adequate emergency obstetric care had been discriminated against because of her ethnic status. 215 In February 2014 the Brazilian Government reached a settlement with the victim's family in the case, agreeing to pay compensation. 216 Another example is the work of the Human Rights Committee against Torture. In its 2006 Report on Peru expressed concerns at reports of women undergoing forced sterilisation and considered that the State failed to take steps to prevent acts that put women's physical and mental health at grave risk and that constitute cruel and inhuman treatment. 217 While not dealing explicitly with indigenous peoples' reproductive rights, the Human Rights Committee has continued to favour an interpretation of Article 27 ICCPR that includes strong indigenous land, cultural and language rights. 218 The HRC has heard a number of cases regarding reproductive health, although not specifically concerning the rights of indigenous women. For example, in K. L. v. Peru the HRC asserted that it is the 'state responsibility to ensure access to legal abortion services.' 219 An important development concerning indigenous women's reproductive rights was the HRC's 2009 call on the Australian Government to 'adopt a comprehensive national mechanism to ensure that adequate reparation, including compensation, is provided to the victims of the Stolen Generations policies.' 220 The UN Committee on Economic, Social and Cultural Rights has defined the minimum standards for delivering the right to health in goods, facilities and services, including the underlying determinants of health care, in particular their availability, accessibility, acceptability and quality. 221 The ESCR Committee also emphasised the participation of indigenous peoples in decisions relating to the right to health. 222 In a decision of 28 November 2014, the Committee called on the Romanian Government to revise its laws and policies to ensure access to reproductive health care for women and adolescents and recommended the implementation of a national strategy on sexual and reproductive health. 223 Also significant has been the role played by the Committee on the Elimination of Racial Discrimination . CERD has intensified its monitoring of indigenous issues and encouraged many States to review their policies concerning indigenous peoples. In this vein, CERD has used the 'Urgent Action Procedure' in order to address the discriminatory policies of some States. 224 --- ILO Convention Compliance Committee State parties to the ILO Convention no. 169 are required to report on measures taken to ensure the implementation of ratified ILO conventions and any problems encountered to the Committee of Experts on the Application of Conventions and Recommendations . This Committee may take specific action against non-compliance, including the power to submit 'observations' and make 'direct requests', 225 yet the Committee's powers to impose sanctions are more limited. 226 The ILO Compliance Committee has contributed to indigenous peoples rights jurisprudence, including by recognising that consultations must be held when indigenous peoples interests are involved. 227 Although the Committee has not decided a case involving Peru's sterilisation programme, 228 in Guerrero and Oaxaca it has received complaints and reports of investigations, observations and recommendations regarding alleged violations of indigenous peoples' reproductive rights. It was alleged that members of public health institutions have performed vasectomies on indigenous men and fitted indigenous women with intra-uterine devices as a method of birth control, without their free, informed consent, in two Mexican states. 229 Moreover, a local study alleged that the health system for indigenous communities was precarious and referred to the inhumane and discriminatory treatment of indigenous persons in health-care centres and to the practice of forced sterilisation of women by tying their fallopian tubes. 230 In response, the Mexican Government indicated that its health institutions have no record of judicial or administrative complaints concerning alleged violations of sexual and reproductive rights of the indigenous population and that their freedom of choice had been fully respected. The Committee did not find this response entirely persuasive. As such, the Committee requested Mexico to supply information on the steps taken to guarantee that the decision to take permanent contraceptive measures is indeed a free choice and to ensure that the persons concerned are fully aware of the permanent nature of the sterilisation measures. Moreover, the Committee requested Mexico to supply information on the extent to which indigenous peoples participate and are consulted regarding reproductive health and family planning. 231 In a the report delivered to the Committee in September 2013, the Mexican Government indicated that the procedure implemented by the Opportunities Programme of the Mexican Social Security Institute is of informed and shared consent. 232 The Government further noted that in 2012, the Gender Equity Action Programme with the Indigenous Population consolidated its initiatives in relation to the Indigenous Women's Forum and developed projects for indigenous women's organisations addressing violence and sexual and reproductive health. 233 The Committee invited Mexico to continue providing information on the manner in which informed consent concerning sexual and reproductive rights has been included in programmes intended for indigenous communities. 234 Despite the fact that, as yet, no overwhelming evidence was presented to the Committee showing that the Mexican sterilisation programme failed to respect women's right to free and informed consent, the Guerrero and Oaxaca case is particularly significant. The significance of this case relates to the fact that it shows the potential for international human rights law to condemn and ultimately revert an existing population policy programme that violates indigenous peoples' reproductive rights. --- Regional Human Rights Systems A State party may declare that it recognises as binding the jurisdiction of the Inter-American Court of Human Rights on all matters relating to the interpretation or application of the ACHR. 235 Twenty-one member States of the Organization of American States have accepted the Court's jurisdiction. 236 It is well-established in the IACtHR's jurisprudence that a State's omission to provide effective protection against threats to indigenous populations -including those to their health -can lead to the liability of the State itself. For example, in Sawhoyamaxa Indigenous Community v. Paraguay, 237 the Court ruled that Paraguayan legislation failed to provide an effective judicial remedy aimed at protecting legitimate land claims made by indigenous communities, which constituted a per se violation of the ACHR. Moreover, the Court has stated that the displacement and expropriation of indigenous' lands amounted to a violation of the right to life. 238 Hence, the IACtHR takes the view that States have both positive and negative obligations in respecting the right to life: [i]n order for this positive obligation to arise, it must be determined that at the time of the occurrence of the events, the authorities knew or should have known about the existence of a situation posing an immediate and certain risk to the life of an individual or of a group of individuals, and that the necessary measures were not adopted within the scope of their authority which could be reasonably expected to prevent or avoid such a risk. 239 As there was no effective remedy or system through which individuals could claim redress for violations of their human rights under Peruvian law during the Fujimori Government, the Inter-American Commission was expected to play an important role in the redress of the indigenous and rural communities that were impacted by Peru's sterilisation programme. 240 However, one significant limitation of the Commission is that it can issue recommendations but not binding decisions. 241 It should be noted, however, that after consideration by the Commission the presumed victims may request that their case be submitted to the Court. 242 In the Mamerita Mestanza Chavez case, the IACommHR found Peru responsible for the forceful sterilisation and death of the victim as a result of the operation. 243 Following this decision, a 'Friendly Settlement' was signed in August 2003 by Peru and the victim's family. In this agreement, the Government acknowledged international legal responsibility, and agreed for the first time to compensate victims of the programme and agreed to modify and implement recommendations made by Peru's Human Rights Ombudsman. However, the agreement does not foresee compensation for other victims. This Friendly Settlement helps to explain why the agreement has not reached the Inter-American Court. 244 Although Peru withdrew its acceptance of the Court's jurisdiction in 1999, it was renewed in 2003. 245 Despite the fact that the IACommHR has taken up the cases of over 200 women sterilised under the Fujimori-era program and recommended their compensation, 246 at present Mestanza is the only case that has led to compensation to a victim, although other deaths have been reported. 247 These are, however, only some of the most damaging consequences of the Peruvian sterilisation programme, and compensatory remedies must also to be available to other victims. Moreover, the IACtHR has upheld the principle of prior, free and informed consent in other cases involving allegations of forced sterilisation. 248 It should also be noted that the Mamerita Mestanza Chavez case is an exceptional one in which the right to health was the central issue in a complaint brought before the IACommHR for violation of indigenous rights, without there being a closer link with the indigenous right to land and property. One explanation for this is connected to the fact that the ACHR does not explicitly protect the right to health, although this right could be regarded as integral to the promotion and fulfillment of the civil and political rights enshrined in the Convention. 249 Indeed, the Court applied this reasoning when it drew connections between a reproductive health right and the rights to privacy and family life in the Artavia Murillo et al v. Costa Rica case concerning the ban by Costa Rica on invitro fertilization . 250 The Court found Costa Rica in breach of the ACHR, ruling that the State's ban on IVF violated the rights to privacy, to liberty, to personal integrity, and the right to form a family, in conjunction with the right to be free from discrimination. 251 The Court concluded that there were less restrictive ways to accomplish the State's objective and to reconcile the interests at stake. 252 The Court finally held that the outright ban on the practice of IVF constituted an arbitrary interference and a restriction incompatible with the AHCR on the exercise of the right to a private and family life in article 11. 253 In Xakmok Kasek community, 254 the Court's reasoning was not limited to a consideration of allegations of violations of indigenous peoples' property rights. Indeed, the Court examined whether Paraguay had violated the provision of basic services, including health services, with respect to the State's obligation to protect the right to life. The Court held that there was a violation of the right to medical care as the State 'had not guaranteed physical or geographical access to health care establishments for the community.' 255 Within the African human rights system, the African Court on Human Rights 256 is yet to decide on a case involving indigenous peoples' rights. 257 Moreover, the African Commission on Human and Peoples' Rights has dealt with only a few cases regarding indigenous peoples' rights, 258 but none of these have dealt specifically with reproductive health rights. The right to indigenous peoples' health more broadly was addressed in the Social and Economic Rights Action Center and the Center for Economic and Social Rights v Nigeria. 259 In this case Nigeria was found to have violated several articles of the African Charter for violating the health rights and livelihood of the Ogoni people. 260 This case concerned the negative health and environmental impacts of oil exploration in Ogoniland and the contamination of water with lead and mercury affecting community health, particularly that of the children. The Commission called for compensation to victims of human rights violationsincluding relief and resettlement assistance to victims of government raidsand comprehensive clean-up of lands and rivers damaged by oil operators. 261 The African Union's legal framework could more specifically protect indigenous peoples' reproductive health rights, in particular since the adoption of the Protocol to the African Charter on Human and Peoples' Rights on the Rights of Women in Africa. 262 The Protocol recognises 'health and reproductive rights' for women in Africa including 'the right to control fertility' and 'the number of children and spacing of children.' 263 In this regard, the African Commission has expressed concerns over States' failure to implement their obligations under the Protocol. In particular it has condemned the practice of forced female circumcisions in many States and expressed concerns about the problems relating to reproductive health care and the quality of services available to women in Africa. 264 --- International and National Criminal Courts The more serious and abusive violations of indigenous peoples' fundamental reproductive rights could also potentially be established as international crimes. As such there is potential to engage the criminal liability of State officials and others acting on behalf of the State before the International Criminal Court . The ICC's jurisdiction established by the Rome Statute, 265 arises automatically over international crimes committed in a State party or by a national of a State party. Sterilisation programmes that target one specific group could amount to acts of genocide and to a violation of Article 2 of the UN 1948 Genocide Convention and Article 6 of the Rome Statute, thus potentially attracting the ICC's jurisdiction. The Genocide Convention prohibits 'measures intended to prevent births within a group' with 'intent to destroy, in whole or in part, a national, ethnical, racial or religious group.' 266 Australia's removal of aboriginal children from their families violates Article 2 of the Genocide Convention, as well as Article 6 Rome Statute, which prohibits 'forcefully transferring children from one group to another group.' 267 However, the crime of genocide requires evidence of specific 'intent to destroy a group', which has proved difficult to establish in specific cases. 268 Although they appear to meet the actus reus element of the crime, it would be challenging for the ICC Prosecutor to establish that these coercive assimilationist policies intended to 'destroy a group', unless the law evolved to recognise and criminalise more broadly acts of 'cultural genocide.' It must be noted that although Australia is a party to the ICC, the Rome Statute only applies to crimes occurring after its entry into force in 2002. In relation to the Peruvian family planning programme, criminal investigations into it were terminated by the Peruvian chief prosecution services in 2009. The public prosecutor in charge of the investigations decided that the over 2,000 registered cases of forced sterilisation did not constitute a severe violation of human rights and thus the statute of limitations prevented criminal investigations. 269 However, formal criminal investigations were reopened by the Attorney General in 2011, in which it is suggested that the violations amounted to crimes against humanity and, therefore, that the alleged offences are not covered by the statute of limitations. In an apparent misinterpretation of the rules of international criminal law, in January 2014 the prosecutors closed the case as it was argued that 'crimes against humanity' only fit the situation of Mamerita Mestanza Chavez . At the time of writing, no criminal prosecutions have been brought against State officials or the health practitioners involved in the Peruvian programme. 270 This decision by the public prosecutor is in direct violation of the Friendly Settlement reached in the Mamerita Mestanza Chavez case, in which the Peruvian Government undertook to investigate and punish those responsible for the reproductive health violations. 271 --- Reparations, reconciliation and redress under national law Beyond the liability of the State and individuals for human rights violations as recognised by international and regional courts, the role of national court systems cannot be understated. In this context, in a landmark ruling in 2007, South Australia's Supreme Court awarded compensation for the first time to a 'Stolen Generation' victim. 272 This decision opened the way for a number of other legal actions claiming reparations for the 'Stolen Generation' before Australian courts. 273 Another notable development is the legislative proposal for a South Australia Stolen Generation Bill 2014. 274 The Bill foresees the reparation of indigenous Aboriginal and Torres Strait Islander peoples removed from their family as children, including the payment of compensation to individual Aboriginal and Torres Strait Islander persons removed from their families as children under previous government policies to of a lump sum of up to AUS$50,000. 276 The focus on monetary compensation under this Bill sets it in contrast with the earlier 2010 Stolen Generations Reparations Bill, which recognised a broader range of reparation measures. That earlier Bill foresaw reparation measures such as funding for healing centres, community education projects, community genealogy projects, and funding for access to counselling services, health services, language and culture training. 277 Moreover, the current draft of the 2014 Bill foresees the monetary compensation of descendants of an Aboriginal removed from his or her family, but not for other family members and parents who have suffered harm as a consequence of the removal. 278 This would have recognised the fact that the government-sponsored child removal policies were particularly traumatic to the families. It would also be in line with the 'Bringing them Home' report recommendations. 279 The Bill is currently being debated by the Australian Parliament -if adopted, South Australia will be the second Australian state to set up a tribunal to compensate the 'Stolen Generation' victims. 280 Reconciliation can play an important role in healing wounds from past crimes, as applied in the case of truth and reconciliation commissions. 281 Even though Peru has set up a Truth and Reconciliation Commission with the purpose of '[...] clarifying the process, facts and responsibilities of the terrorist violence and human rights violations produced from May 1980 to November 2000 [...]', the Commission has been criticised for failing to address the state-sponsored sterilisation programme that took place between the mid to late 1990s. 282 In contrast. Australia has not set up a such a commission. A significant development happened in 2008 when Kevin Rudd, then Australian Prime Minister, apologised for the tragedies suffered by aboriginal families of the 'Stolen Generation', breaking with his predecessor's disregard for this issue. 283 The apology is not only symbolically significant, as the more concrete measures for reparation of indigenous peoples have been proposed subsequently in the South Australian Stolen Generations Reparations Tribunal Bill and the South Australian Stolen Generation Bill . 284 Another positive development came in April 2009, when Rudd's Government endorsed the UNDRIP, despite Australia's vote against it in the General Assembly under the former Howard Government. 285 It has been recognised that the apology was only a 'first step by acknowledging the past and laying claim to a future that embraces all Australians' 286 and as Jenny Macklin, who was then the Minister for Families, Housing, Community Services and Indigenous Affairs has reportedly put it: 'we have serious inequalities between indigenous and non-indigenous Australians. The apology is symbolic, but there's a lot of hard work to be done to reverse those inequalities.' 287 Indeed, indications that considerable work lies ahead are the suggestions that aboriginal child removals still currently happen under Australian welfare legislation 288 and that maternal and infant mortality rates for indigenous women and children in Australia are still much higher than the national average. 289 In this regard, Julia Gillard, former Prime Minister, apologised on 21 March 2013 for another Government policy from the 1950s to 1970s that led to thousands of unmarried young mothers who were coerced or deceived into giving up their babies to adoption by married couples. 290 It will be interesting to follow how the Government's subsequent response to the apology for these past policies, which were applicable to Australian women irrespective of ethnicity, will compare to the reaction that followed the Government's apology to the 'Stolen Generation.' This review of the jurisprudence of international courts and UN treaty-monitoring bodies, as well as of two regional human rights systems, suggests that under international law States have an obligation to ensure the protection of indigenous peoples' reproductive rights and that States can be held liable for their failure to protect these rights. Moreover, international, regional and national courts and procedures have played important roles in recognising the right to reparation schemes and a measure of redress from past government policies violating indigenous peoples' reproductive rights. The Guerrero and Oaxaca case before the ILO Compliance Committee is particularly significant. The significance of this case stems from the fact that it illustrates the role that international human rights law can play not only in addressing the reparation of victims and their relatives , but also to condemn -and potentially end -an existing national policy of population control that violates indigenous peoples' reproductive rights. --- CONCLUSION The oppressiveness of post-colonial State systems has confronted indigenous peoples with the terrifying prospect of inter alia genocide, birth control and sterilisation programmes. As such, what type of governmental system is necessary to accommodate indigenous claims for some degree of independence in order to rule their own communities and enable them to control their fertility? The right to reproductive freedom is a fundamental tool for indigenous peoples to decide their own reproductive strategies. There are grounds for concern that rapid population growth in some continents and forced migration arising from environmental factors, could lead to the reassertion of coercive national population policies , and that these policies could impact disadvantageously on indigenous peoples . This is so not only because population policies tend to be top-down and disregard the rights of minorities, but also because the institutionalisation of these policies may have a racist or discriminatory intent. International and national human rights law play a central role in giving effect to the right of reparation to indigenous peoples that have suffered violations of their reproductive autonomy from past government-sponsored policies. Recent history and events have shown that forceful birth control programmes can pose a significant threat to indigenous peoples. The long-term survival and generational continuity of distinctive ethnic and cultural groups requires that there is effective protection of their reproductive health rights, in particular in countries such as Peru and Australia where there has been well-document evidence of violations of indigenous peoples reproductive rights. These rights must also be recognised and made effective under international human rights law. While securing reproductive autonomy may pose an even bigger challenge for coming generations of indigenous peoples in light of future challenges including dislocation and climate change, the main threat to indigenous peoples' survival arise from national policies that disregard their right to health more broadly, including policies that are complacent or condone the invasion and expropriation of their lands. Certainly the dispossession of indigenous lands and the impact that this has on their health pose a major threat to their survival in many parts of the world. Therefore, the right to life and to health of indigenous peoples need to be effectively protected and must be made enforceable under international human rights law. This is recognised under the 1994 Plan for Action Programme, emphasising that population policies impacting indigenous peoples need to be considered within the broader context of indigenous peoples' fundamental rights, including the right to health, to land and to natural resources. 291 Although States have renewed their commitment to improving women's reproductive health and to strengthening indigenous peoples' rights in the 2014 ICDP , it is notable that a number of States did not declare the protection of reproductive health rights for indigenous women to be a current priority. Yet it is the indigenous identify and the survival of distinctive indigenous cultures that may be the biggest long-term threat for indigenous peoples, in particular government policies that aim to assimilate them. Some of these assimilationist policies may in fact be part of a family planning programme, as in the case of the Aboriginal child removal policies in Australia. Therefore, it is crucial that indigenous traditional cultures and languages that have evolved over centuries are also effectively protected under international law. 292 The perception of 'indigeneity' by broader society and indigenous peoples themselves, defines the contours of the relationship between cultural identity, population politics and human rights. This paper concludes that the right to reproductive health must be seen as an integral part of the right to self-determination and to health under international law, which delineates the degree of autonomy of indigenous peoples to self-government, 291 1994 Plan for Action Programme, 6.21 292 See Siegfried Wiessner, ' The Cultural Rights of Indigenous Peoples: Achievements and Continuing Challenges' 22 European Journal International Law 1. On the other hand, it is rightly suggested that it is not for human rights law to advocate that indigenous cultures are kept 'frozen in time', but to allow indigenous peoples to develop in their own way and protect their right to enjoy their own traditional culture. See also, Pereira and Gough . including reproductive strategies. Moreover, the right to indigenous peoples' reproductive autonomy must be given effect by procedures for prior and informed consent and to consultation in the context of family planning programmes. These rights must be recognised in the post-2015 Millennium Development Goals framework, The right to self-determination, to health and to prior and informed consent provide the essential legal basis that must be made enforceable under international law in order for indigenous peoples to triumph over assimilationist and other neo-colonial practicesof which forceful child removal policies and birth control programmes are just some of the unfortunate examples.

P e r ei r a, Ric a r d o 2 0 1 5. Gov e r n m e n t-s p o n s o r e d p o p ul a tio n p olici e s a n d in di g e n o u s p e o pl e s: C h all e n g e s fo r in t e r n a tio n al h u m a n ri g h t s law. N e t h e rl a n d s Q u a r t e rly of H u m a n Ri g h t s 3 3 (4) , p p . 4 3 7-4

Introduction Concepts of 'cybernetics', 'systems' , 'complexity', 'chaos', 'catastrophe', and 'oscillation' have long been part of discourse in the behavioral, social, and biological sciences. Without the ready availability of mathematical models for the social and biological sciences, a nonmathematical approach to systems theory was prominent in early biofeedback work, and was well articulated by Schwartz and colleagues . However, a mathematical approach for aircraft communication and control systems had been articulated by Wiener as early as the 1940s ). This model served as a heuristic for describing behavioral, economic, biological, and astronomical systems, among others . In recent years, mathematical versions of systems theory have been applied in biology and behavioral science, as described below. Because Wiener's theory had limited predictive power, it was soon combined and supplanted by theoretical systems with greater complexity, such as Shannon's information theory , which incorporates concepts of channel capacity, noise, and entropy , as well as chaos theory , which emphasizes deterministic nonrandom but complex nonlinear relationships among a large number of co-occurring processes, for which statistical prediction is possible. Applications of both information theory and chaos theory have been made to the biological and behavioral sciences, and the present article draws upon insights from this work. For the purposes of this article, we define a system as a variety of elements that interact with one another to form a whole entity. A system's distinct parts are not isolated from each other, thus the characteristics of the whole entity cannot generally be deduced from each of its components. A stable system retains important characteristics even in the face of perturbations that disturb the behavior of specific elements. Dynamical systems show changing patterns of action over time, but retain characteristics of an integrated whole. Cybernetics is the study of such regulatory processes, including system characteristics like stability, feedback, adaptation, information, and the relationships among them. Dynamical systems have been described as "organized" or "disorganized" and of greater or lesser complexity. Simpler organized systems can often be described and modeled in linear terms, although nonlinear approaches usually are more appropriate for more complex systems. A control system is defined as a stable system, in which system elements interact to preserve system stability, both for internal control and in response to perturbations caused by external influences. Systems can be modeled as closed-loop or open-loop. An open-loop system involves only a system response to an outside event. It does not provide an internal mechanism for monitoring system performance or preserving system stability. For example, open-loop models have been used to biomimetically describe the flying patterns of drosophila melanogaster after stimulation . Behavioral and chemical interventions designed to change system behavior are examples of open loop control processes, such as when we teach a person to lower muscle tension or blood pressure directly, rather than by using natural feedback mechanisms. A closed-loop model includes internal regulation, whereby system responses to outside stimulation are processed through internal feedback loops that monitor and adjust the response of the system itself. As a discipline, biofeedback is particularly concerned with closed-loop, internal regulatory systems, and tries to maximize their effectiveness. A good example of a closed loop system is the baroreflex , as will be described in more detail later in this article. However, when biofeedback or relaxation is used as a strategic intervention to control acute symptomatology, the resultant process can be described as an open-loop system. A common control mechanism leading to system stability in closed loop systems is the negative feedback loop. Negative feedback loops preserve stability by activating opposing processes that act to modulate change, though importantly, stable systems tend to contain both positive and negative feedback loops , as will be described below. Negative feedback systems usually have inherent delays caused by the time needed for opposing system elements to effect change in each other. Negative feedback loops with delays cause oscillations at particular frequencies Cinquin & Demongeot, 2002). A very simple negative feedback system is the thermostat, which responds to changes in ambient temperature by switching a heating or air conditioning system on and off. Here the delay is dependent on such factors as outside air temperature, insulation, and volume of inside air. Thus, room temperature oscillates around the thermostatic setting, but is rarely constant. The negative feedback loop in the baroreflex system also contains delays, as will be described below. Oscillatory properties have been observed in systems as small as the cell, and as large as societies. For example, at a cellular level, oscillations have been used to map various negative feedback loops in protein concentrations, and circadian gene expressions in bacteria . In social systems oscillations describe patterns of voting, marital relationships, and mood. The baroreflex system is a good example of a stabilizing oscillatory process. It produces periodic fluctuations in blood pressure, heart rate, and vascular tone that reflect modulatory control. External influences on blood pressure induce blood pressure changes that, without a control mechanism, could burst blood vessels or cause circulatory insufficiency, depending on the direction of blood pressure change. The baroreflex responds to blood pressure increases by slowing heart rate and causing vasodilatation, which, in turn, produces a decrease in blood pressure. Each time blood pressure decreases, the same mechanism produces vasoconstriction and heart rate acceleration, which causes blood pressure to rise again. Changes in blood pressure are delayed by a number of factors, including inertia in blood flow and blood vessel plasticity . The delay in blood pressure changes following heart rate change averages about five seconds, thus yielding the ubiquitous heart rate oscillation of about 10 seconds , i.e., five seconds for the baroreflex to cause an increase in blood pressure when the baroreflex responds to a decrease in blood pressure, and five seconds to cause a decrease in blood pressure following a baroreflex response to a blood pressure increase. In actuality, however, the baroreflex system is composed of two closed linear systems with negative feedback loops -it also contains a slower vascular tone loop with a rhythm of approximately 0.02-0.03 Hz, which has a slower response time than the heart rate loop . It should be noted that a full description of baroreflex control would not be restricted to purely mechanical properties of the reflex. The baroreflex system shows characteristics of neuroplasticity. Dworkin demonstrated that the baroreflex can adapt to changing environments through classical conditioning. Consistent with Dworkin's findings, Lehrer and colleagues demonstrated that regular stimulation of the baroreflex through biofeedback can increase baseline baroreflex gain, suggesting an improvement in regulatory capacity over blood pressure changes . --- Complex control of systems Most psychological and biological systems are much more complex than thermostats, with multiple control mechanisms, and multiple overlapping oscillations, the quantitative characteristics of which depend on the frequency and phase of each component oscillator, as well as the interactions among them . Changes to one system component affect others, so the individual control mechanisms are interdependent, thus promoting biological vitality and complexity. Additionally, it should be noted that closed loop systems with negative feedback loops often require simultaneous open loop processes in order to function properly, to provide sufficient stimulation to stimulate negative feedback loop activity . This adds to the complexity. Simple control, such as control of room temperature, can typically be described with linear mathematical models, but when multiple controllers operate, each with a characteristic frequency, attendant mathematical computations become more complex. Existence of multiple overlapping control systems assures system stability in the face of various perturbations by providing multiple "backups". A high degree of entropy, reflecting a high degree of nonlinearity and/or a need for multiple linear formulas to describe the system, is often related to health and biological system stability. Conversely, simple periodicity or random variability often underlies pathology in disordered systems , reflecting, respectively, either the presence of only a single control oscillator, or absence of any regulatory activity. Oscillatory patterns with greater complexity, such as those that occur when a number of oscillatory patterns overlap, are described as "chaotic". Chaos reflects the simultaneous operation of numerous control processes. Although the healthy cardiovascular system can be effectively modeled using linear statistics , it also has chaotic properties, which diminish when cardiac function is impaired or an individual is in danger of dying from physiological decompensation . Similarly, healthy adjustment is characterized by chaotic rhythms in appetite and mood , among other biobehavioral dimensions. Also, negative feedback loops often work in concert with open loops as described above, and with positive feedback loops and resonance characteristics, as described below, to maintain optimal functionality. --- Oscillation, control, and health Many oscillatory systems in the body, such as neural , cardiovascular , and circadian tend to degrade with age, reflecting a decrease in adaptive capacity. In addition, there are also pronounced age-related decreases in baroreflex gain , while blood pressure oscillations tend to increase with age , perhaps reflecting degrading of control over this otherwise tightly regulated function. Generally speaking, oscillatory patterns in a variety of biological control systems tend to be a characteristic of youth, fitness, and health. Oscillatory properties do not only represent activation of specific control reflexes . They provide information as well. Periodic stimulation is required for system control in feedback networks. This principle has been proposed for cellular behavior as well as macro biological and behavioral systems, as in discussions of open loop processes described above, and both resonance and stochastic processes, described below. External stimulation "perturbs" a control system causing it to oscillate. Without such oscillations, and hence, without noise , systems may fail to get the information they need to function properly. The tendency for noise to stimulate system oscillation is called stochastic resonance. There is evidence that interoceptive and exteroceptive noise enhances the sensitivity of a variety of control processes, from micro to macro levels . Studies of stochastic resonance show that noise can facilitate control functions, as in optimization of monosynaptic afferent reflexes in the motor system of the cat by noisy stretching of a synergistic muscle . In humans, extreme absence of system perturbation, as in sensory deprivation, can cause regulation go awry, resulting in psychotic-like sensory and emotional symptoms . One could, therefore, speculate that psychological stress may have some beneficial effects, by introducing noise into neural regulatory processes, just as exercise helps maintain many physiological processes. Since, in biological systems, oscillations tend to reflect the operation of reflex systems involved in modulation and self-control, presumably a complete lack of stimulation would deny exercise to these reflexes, causing them to atrophy. Under normal circumstances, however, the environment provides sufficient perturbations in biological and behavioral systems to preserve their normal function. In psychophysiological theories of stress effects, this formulation overlaps with the more recent concept of "allostasis" and "allostatic overload" . Allostasis refers to stability achieved through variability, and is analogous to the activity of oscillating control processes stimulated by stochastic noise, or general unorganized environmental stimulation. A well functioning personality or biological system responds well to moderate environmental stressors, and, in fact, may be energized by them, as in Yerkes and Dodson's model of stress and performance, wherein moderate stress enhances performance in moderately difficult tasks, while very low or very high levels of stress can be detrimental . Allostatic overload occurs when environmental stresses are either too great or too prolonged, or when various control reflexes become exhausted. --- Resonance: A source of stimulation and simplicity in system oscillation In addition to oscillation, another characteristic of negative feedback systems with a delay, is resonance . When outside stimulation causes a system oscillation at a particular frequency, and no other forces are at work to dampen the oscillations, negative feedback loops can themselves destabilize the system by causing increasingly large resonance frequency oscillations, to the point where information from other frequencies no longer gets processed. A common example of a 'run-away' negative feedback loop, resonance effect is the "Larsen effect" that occurs when a microphone is placed near a speaker, causing a high-pitched squeal at a single frequency . In such a case, a single high-amplitude oscillation, triggered either by noise or by pulsatile stimulation close to the resonance frequency, can obscure the effect of otherwise meaningful perturbations. In the human body, resonance effects in the postural control loop have been implicated in postural impairment in Parkinson's disease . Resonance has been similarly implicated in age-related impairment in gait control . In terms of psychophysiology, it is conceivable that resonance effects might inhibit the fight-or-flight reflex because a single homeostatic process may overwhelm reflexes needed to confront an environmental stressor. One could speculate that resonance effects, stochastic and otherwise, might deprive the system of information from various internal control mechanisms, as when resonance-frequency oscillations are at such a high amplitude that they obliterate information from reflexes operating at different frequencies. Thus, it is possible that resonant oscillations in the heart rate baroreflex system at 6/minute could weaken the effects of other control reflexes, by greatly diminishing their relative size and depriving the system of information. This is a hypothesis for future research. To prevent a 'run-away' resonance process, systems must be dampened . Dampening effects on heart rate amplitude during heart rate variability biofeedback probably stem from inherent limitations in the ability of the heart to respond at extreme levels. The literature on heart rate variability biofeedback, however, shows that not all resonance effects are detrimental. Resonance in the cardiovascular system produced by a rhythm in the baroreflex is the basis of the beneficial effects of heart rate variability biofeedback. The baroreflex provides a negative feedback loop for controlling blood pressure, such that heart rate falls when blood pressure rises, and vice-versa when blood pressure falls, thus modulating blood pressure fluctuations. The resonance frequency of this system appears to be related to the individual's blood volume , with a lower resonance frequency among people with a larger blood supply . The resonance frequency appears to be determined by changes in hemodynamic inertia resulting from fluctuating heart rate. When this resonance system is activated by a respiration rate close to the resonance frequency, oscillations in heart rate become very large. This high-amplitude stimulation of the baroreflexes appears to strengthen them , and is perhaps the cause of the various salutary effects of heart rate variability biofeedback. Heart rate variability biofeedback has been shown to restore autonomic control that has been acutely repressed by experimental exposure to inflammatory cytokines , and appears to ameliorate a number of disorders characterized by autonomic and/or emotional dysregulation , including hypertension , asthma , anxiety/stress , depression , and chronic pain , while improving athletic performance . We might, however, theorize that the effects of constant breathing at resonance frequency would not be advantageous. Chronic resonance frequency breathing could theoretically weaken or obstruct reflexes dependent on oscillations at other frequencies. Thus, individuals practicing heart rate variability biofeedback are instructed to practice for a relatively brief period of time daily and to use the technique strategically when symptomatic. --- Positive feedback loops The body also contains multiple positive feedback loops, whereby change in a particular direction facilitates greater change in that direction. Although positive feedback can provide a stimulating effect in an open loop system, it also plays a role in closed loop systems. In psychophysiology, positive feedback loops often are involved in maladaptive processes . A prime example is anxiety and depression's propensity to amplify their inherent symptomology by triggering greater sensitivity to anxious or depressive thoughts, thus creating a vicious cycle. On the cellular level, positive feedback loops may be necessary to stimulate modulatory negative feedback loops . However, positive feedback loops can also play an important role in enhancing emotional control. The psychological augmenting effects of positive feedback may add to oscillation amplitude during HRV biofeedback, or the sympatholytic effect of muscle relaxation. This may result from a spiraling relaxation effect. Just as anxiety begets more anxiety by sensitizing anxiogenic brain circuits , heart rate variability biofeedback or muscle relaxation therapy may compound relaxation effects. Positive feedback loops can also play a role in creating system complexity, which may increase stability by creating a multiplicity of stable states , as well as by maintaining oscillation amplitude during frequency adjustment . Positive feedback loops have been shown to promote load compensation in control of movement , and to be necessary to prevent dampening of oscillations in cellular function . Positive feedback loops also are important for the maintenance of the ovulatory cycle , and for propagating dopaminergic signaling . Without some form of regular perturbations, oscillations will gradually decline in amplitude and disappear. Positive feedback loops can provide these perturbations. --- Simplicity, complexity, and randomness The dimensions of simplicity and complexity are important for understanding psychobiological control, and can be usefully incorporated into theories of applied psychophysiology. Simplicity in a biological or behavioral control system can have multiple sources. In addition to reflecting the effects of resonance, simplicity may result from system fatigue due to allostatic overload , or from biological damage to system components, such as when heart failure leads to diminished entropy in heart rate . Simplicity in heart rhythms as a sign of cardiac pathology was discussed earlier. Emotional rigidity, or simplicity, is well described in DSM-IV as a sign of psychopathology, which is characterized by a tendency to respond to a wide variety of situations with a stereotyped response: sadness, anxiety, suspiciousness, anger, etc. However, just as simplicity is a sign of poor adaptation, so is random variation, suggesting lack of modulatory control. Pathological examples of random fluctuations include manic or depressive episodes in bipolar disorder, as well as certain cardiac arrhythmias such as preventricular or preatrial contractions, asthma exacerbations, and so on. Work in the tradition of chaos and information theories has provided mathematical tools for describing organized complexity of temporal patterns in terms of entropy, as described in information theory, including Shannon's "spectral entropy" calculations , Pincus' "approximate entropy" ), Lempel and Ziv's method for evaluating "randomness" . These measures represent various approaches to calculating relative unpredictability of fluctuation patterns in a time series, or the number of dimensions, or fractals necessary to describe a data set mathematically. --- Autonomic regulation Healthy regulation is often characterized as sympathetic and parasympathetic activity converging in a limit cycle around a critical value, such as is the case of cortisol-vasopressin or acetylcholine-epinephrine reactions, where each process induces a compensatory reaction in the other . Thus, sympathetic activity may simultaneously suppress parasympathetic activity, but increase parasympathetic reactivity, thereby producing an oscillation . During extremely stressful stimulation, an individual may show wild oscillations in various autonomic functions, both sympathetic and parasympathetic , although these oscillations are sometimes paradoxically suppressed in severe emotional reactions ). Heart rate changes in response to stimulation appear to be important modulators of systemic response. Reduced cardiac variability in response to stimulation is related to a variety of pathophysiological states , while increased heart rate variability and vagal influence on the heart are positively correlated with recovery after physical exercise stress . Thus flexibility in cardiac response to stimulation, and higher vagal influence on the heart are related to better adaptation to interoceptive and exteroceptive demands . This is another example of ways in which complexity and oscillatory function are related to health and adaptability. --- Thermoregulation Thermoregulation is another system in which several systems converge around a critical value. Until the mid 20 th Century, stability and circadian variability in core body temperature were considered homeostatic responses to environmental triggers, such as light intensity or daily routine. In fact, core body temperature is partially regulated by an endogenous clock in a central control mechanism, which integrates multiple systems including thermoreception of ambient temperature, systems for conserving or dissipating heat, as well as those for generating heat, and endogenous circadian regulators. As with heart rate variability, circadian thermal variability decreases with age . A relationship between circadian thermal oscillation amplitude and general homeostasis can be hypothesized, but further evidence is needed to confirm its existence. --- Oscillation, complexity, and disease Heart rate variability complexity is indicative of cardiac and neurocardiac flexibility and adaptability, and may be diminished significantly by pathology. A healthy heart can respond to various moment-to-moment physical and emotional demands, while the diseased heart does not exhibit such flexibility. Although oscillations may persist in the diseased heart, aging and disease cause increasingly more regular oscillatory patterns reflecting a decrease in mechanisms of cardiac control . Such noncomplex patterns of variability ultimately predict death in the critically ill . Similarly, a disturbance of complementary oscillatory patterns in pancreatic islet beta cells is associated with impaired glucose tolerance and diabetes mellitus . Under normal conditions, oscillations in nervous output from pancreatic ganglia cause pancreatic islet beta cells to secrete insulin in a coordinated pulsatile manner, at two distinct periodicities, 1) in small amplitude oscillations with a periodicity of 5-15 minutes, and 2) in larger amplitude ultradian oscillations ranging from 80 to 150 minutes, particularly in response to the ingestion of food or intravenous glucose perfusion . This oscillatory rhythm operates through an insulin-glucose negative feedback loop, with a delay between plasma insulin and hepatic glucose production. Compared with steady-state infusion, this oscillation promotes more efficient glucose utilization . When these oscillatory rhythms are disturbed, such as in diabetes mellitus type II, decreased oscillatory complexity is associated with system decompensation . --- Mental illness Mood and behavior appear to be governed by many of the same concepts used to describe oscillatory behavior in physiological processes. As with physiological systems, organized complexity is the hallmark of healthy psychological behavior. Both hyper and hypostabilities may characterize a breakdown in normal mental functioning, and leave a person psychologically vulnerable to the deleterious effects of stress. Bauer and Whybrow found that while participants with bipolar disorder showed some brief periods of fairly well defined cycling in mood, their overall pattern of mood changes were less complex than controls. Paulus et al. compared patients with schizophrenia to healthy individuals in a binary choice study, and found that participants with schizophrenia showed a simpler and more predictable pattern of response, apparently reflecting decreased adaptability in cognitive processing . A study of laterality in electrodermal activity found decreased activity and complexity in the left hand, compared with the right, among depressed patients, and decreased activity and complexity in the right hand, compared with the left, among patients with schizophrenia, while there were no laterality differences among healthy individuals . These results are consistent with theories of right hemisphere dysfunction in depression , and left hemisphere dysfunction in schizophrenia. However, even some psychopathological states appear to possess chaotic organization. Tschacher, et al. recorded occurrence of psychotic symptoms among 14 patients with schizophrenia over time, and found a complex, nonlinear time course in eight of them. Others showed more random psychotic behavior, suggesting that the patients were responding to environmental cues, not modulated by control processes. --- EEG oscillations reflect system functions Systems theory also applies to neurons' activity in the brain, whose functioning has been shown to be interdependent with other nearby neurons and even neurons in different brain regions. The complexity and chaotic characteristics of brain function, particularly as studied by electroencephalogram , has been thoroughly reviewed elsewhere . Briefly, EEG reflects summations of excitatory postsynaptic potentials that neurons emit as discrete events. When neurons fire synchronously, relatively high-amplitude slow waves occur, in a relatively simple pattern; when they fire desynchronously, the numerous potentials can both augment and cancel each other out, generally resulting in loweramplitude, more complex signals at higher frequencies . Fast-firing cells in the cortex may contribute to these effects, as does the interaction between excitatory and inhibitory processes. Faster, low-amplitude oscillations tend to predominate during states of arousal, attention, and cognitive processing , and in general, during periods of greater cortical activity . Oscillations in EEG signal reflect the interplay of various brain processes involved in cortical integration, such that specific kinds of mental activity may be reflected in definable patterns of activity at specific frequencies, and at specific electrode locations . Slower waves from the cortex reflect decreased arousal at the measured EEG sites . Low levels of alpha and beta activity and/or increased waking slow-wave activity are associated with a lack of inhibitory control over behavior . Similarly a greater frontal theta beta ratio is related to decreased emotional response inhibition . Greater delta and theta activity and lower alpha activity have been found in the EEGs of children with fetal alcohol syndrome . This pattern is also found in attention deficit disorder, lower cognitive abilities, impaired motor control, hypoglycemia, and antisocial behaviors. It is even found during deep relaxation and sleep, while exposed to hypoxia, fasting, and in sexual arousal and orgasm . Conversely, delta, theta, and lower-frequency alpha activity are inhibited when people are awake and cognitively active. A recent study found that, during tasks with increased cognitive demand, increases occurred in the EEG frequency with the greatest activity across electrode sites, but a decrease in EEG entropy . Oscillations are not smoothly distributed across the frequency spectrum. Resonance structures have been identified at frequencies of 4, 10, 20, and 40 Hz with the largest body of research on the 10 Hz rhythm, the center of the alpha frequency band . Resonance properties can be determined by a variety of methods. For instance, one can stimulate the system at specific frequencies with photic stimulation, and determine the frequencies at which high-amplitude oscillations are obtained . One may also use computer simulations based on known frequency characteristics of particular nerve cells or from evoked potentials . Alternatively, one can stimulate an individual with audible noise and measure the obtained frequency peaks . Resonances at these frequencies suggest the existence of specific positive feedback reflex arcs at each resonance frequency, related to particular neural processes contributing to each resonance frequency. Thalamocortical processes have been studied and modeled, and rhythm bands containing each of these frequencies have been described in a large literature . It has been theorized that delta oscillations reflect activity of motivational systems, while theta oscillations reflect emotional regulation, and alpha oscillations reflect inhibitory processes . Some theorists have related specific resonance frequencies to the period length of experimentally evoked action potentials, event-related potentials, synchronies, or asynchronies . Patterns of theta and alpha-frequency resonances triggered by memory consolidation processes have been identified . It is also known that memory consolidation occurs during sleep , which is characterized by high-amplitude slow EEG waves primarily stemming from the brain stem, with cortical modulation of lower centers, perhaps from a single oscillator . Phase relationships in EEG activity among various brain areas have also been studied, appearing to vary systematically. For example, occipital phase synchrony in the alpha rhythm tends to be greatest during relaxation with eyes closed, and decreases with greater arousal . Experience of "pure consciousness" in transcendental meditation appears to be characterized by alpha phase synchrony from the frontal and central areas . Although phase synchrony may reflect tight coordination among brain sites, more complex forms of coordination have also been modeled, often involving varying frequencies, varying orderly phase relationships, and nonlinear relationships, where frequent changes in phase relationships among brain areas reflect complexity of cognitive processing . Measures of chaos and complexity applied to EEG data suggest a simpler pattern of brain organization, with, fewer control processes, among patients with schizophrenia and depression . Pezard et al. found that EEG patterns became more complex among depressed patients, whose depression improved, becoming indistinguishable from those among healthy individuals, while the pattern of decreased complexity persisted among individuals whose depression did not improve . Other studies have found risk for autistic spectrum disorder and the presence of attention deficit hyperactivity disorder to be related to decreased average EEG complexity. --- Social systems Humans live as couples, in families and in neighborhoods or larger social and economic networks. Systems theory has been applied to social systems, and, indeed, oscillatory patterns of social system behavior can reflect negative feedback loops, which could contribute to stability. Stability in a social system generally represents a stable pattern of interaction, implying the action of mechanisms that resist change. Oscillatory patterns of openness and closedness to new ideas have been described in societies , as have variations in societal norms about expression of emotion , although presumably extremes in either of these characteristics trigger a consequent tendency in the opposite direction. Voter patterns appear to have oscillatory characteristics, which, in the United States, have been quantified as thirty-year oscillatory patterns in party alignment . The cause for this oscillation period may reflect generational change. Perhaps people see the problems caused by results of their parents' voting patterns, and react to this by changing course. Although delays, chaos, and resonance characteristics in social feedback loops have not been measured, the existence of specific frequency patterns in some of them suggest that these characteristics may be present. Dyadic systems may also have oscillatory qualities. Gottman and his colleagues have taken a mathematical approach to assessing marital systems . They have applied concepts of nonlinear dynamics and catastrophe theory, which posit complete periodic restructuring of systems when perturbed by certain forces. They propose that marriages have one or more "set points" for nature and style of interaction, some of which may be favorable to marital satisfaction and stability, and others unfavorable. From this perspective, they have developed formulas that predict marital stability and divorce. They record verbatim transcripts of marital interactions, where they note the number of times that individuals respond to each other's content with specific affect, including such variables as threshold for negativity, and frequency of positive and negative interactions. In marriages characterized by reciprocal negativity and paucity of positivity in interactions, couples can be trained to modify their individual behavior. Thus, based on positive feedback loops, marital systems, wherein negativity begets negativity, may be altered so that positivity begets positivity. The overarching goal is to move the interaction system from a negative steady state to one that is more positive. Repair and dampening functions have been calculated, which can modify the positive or negative influence of one partner on another in various steady state conditions. The investigators report conflicting data about whether rigidity in behavior predicts poor marriage outcome, but show how intervention can affect the system of interaction and, ultimately, marital satisfaction. Their model includes the interaction of perception of well-being in the relationship, the flux over of negative and positive behaviors, and physiological responses. Systems theory has also been applied in analysis of corporate structures , but concepts of oscillation, delay, and influences of positive and negative feedback loops have not been studied. Corporations are systems, just as those defined above. They are constantly bombarded by changes in market conditions, competition, etc., and need to maximize the adaptiveness with which they respond to changing conditions, while still maintaining their structure. Information and communication, both from outside of and within the organization, has been proposed as a medium for achieving these ends. Utilization of feedback from information allows an organization to monitor both external events and internal processes. For example, if an organization's norms do not allow for transmission of bad news to its CEO, then poor, uninformed decisions are likely to be made. Various feedback mechanisms can prevent change from happening . Although we have found little research using time series analysis to discover patterns of oscillations in negative feedback loops comprising organizational structures, examples of feedback loops in organizational behavior have been described . Presumably oscillations occur in organizational communication as well as in other control systems, and might be quantifiable on a number of dimensions, such as commands to change operational procedures to adapt to external pressures, oscillating with instructions to follow internal protocols, pressured work versus relaxation, and supportive versus critical communications from superiors. --- Implications for research In psychophysiology, linear models are frequently employed to predict one variable from another. However, because physiological systems are so complex, such prediction does not truly characterize physiological states. Time series models have become available that can be used to better model these systems. The complexity of biological and behavioral systems and the fact that complex systems can be mathematically modeled present a challenge to psychobehavioral research. This paper proposes that we re-examine the systems approach, in light of methodologies recently proposed. Cybernetic systems approaches in biology and psychology present practical challenges such as in the assessment and measurement of multiple systems, and large quantities of time series data. Nevertheless, quantification of complex systems involved in personality, illness, marital or social harmony, and stress reactivity may give us greater insight into ways of promoting health and managing disease, and are thus amply justified. --- Hypotheses generated by this review This review would lead us to posit certain hypotheses. These include: 1. Any stable system can be modeled using time series analysis. --- 2. Internal control mechanisms can be modeled as closed negative feedback loop systems with oscillatory properties. Decreases in oscillatory properties will be associated with less adaptability to perturbations , and greater vulnerability to system failure. --- 3. Stable systems are complex, indicating a multiplicity of control mechanisms; however, they are not random. 4. Adaptability of a system can be modeled from the effect of perturbations on its various closed-loop systems of internal control. Changes in any elements due to system perturbations will return toward a pattern of stable equilibrium. --- 5. A stable system does not necessarily have a single level of stability. This may change depending on demands placed on the system. 6. Disease, injury, stress, and older age all reduce adaptability, and will be reflected in decreases in oscillations of control systems, complexity of oscillations, and ability to return to equilibrium after perturbation. They may also be characterized by greater periods of randomness. --- 7. This model can be applied to all systems: biological, behavioral, cognitive, emotional, social, organizational, etc. Elements of systems could be quite varied, and include such diverse phenomena as behaviors, communications, and biochemical or physiological functions. --- Particular implications for applied psychophysiology and biofeedback The present analysis indicates that further investigation of patterns of oscillation in various psychobiological systems is warranted. Better identification of oscillatory systems may lead to the elucidation of homeostatic or allostatic control mechanisms in such systems, and may ultimately lead to the development of ways to exercise and strengthen reflexes that strengthen allostatic capacity. Ideally, these exercises could render an individual more resilient to effects of environmental stressors, improve performance, and strengthen resistance to psychological and physical disease. Systematic use of various quantitative approaches to assessing oscillation and complexity could contribute to advances in this field. Work in this direction has already begun in studies of heart rate variability and EEG biofeedback. Heart rate variability biofeedback is based on the explicit mechanism involving stimulation and exercise of modulatory reflexes by stimulating resonance characteristics of an oscillating system. For this method, the theory proposed in this paper predicts that only periodic practice of the technique would be beneficial, while constant practice of the technique might actually be iatrogenic, since the resonance frequency oscillations would prevent chaotic variability in heart rate, and possibly decrease adaptability, if they were continuously present. However, strengthening of homeostatic reflexes, with a consequent increase in chaotic variability, appears to be beneficial. Periodic stimulation of a reflex seems to improve its efficiency. We have found this to be the case with the baroreflex . As yet, no studies have reported changes in chaotic variability in the resting state after a course of heart rate variability biofeedback; although other measures of variability do appear to increase. It is possible that biofeedback used to increase other sources of periodic variability in the cardiovascular system might also be explored. One such possibility is biofeedback targeting the vascular tone baroreflex, following research by Vaschillo and colleagues . Neurofeedback is another biofeedback modality that explicitly addresses oscillations by stimulating and exercising modulatory reflexes in the central nervous system. Alpha rhythm and theta: beta rhythm biofeedback have been used to enhance relaxation, memory consolidation, and attention, and appear helpful for such disorders as attention deficit hyperactivity disorder , and anxiety disorders , as well as intellectual performance enhancement . This feedback may strengthen reflexes involved in conscious attention and memory consolidation, depending on the particular frequencies and brain regions that are trained. Pulsatile stimulation by electricity and light flicker have also been used . Chaotic and oscillatory properties of various psychophysiological measures suggest that there may be many more potential control reflexes that can be trained, perhaps one on each fractal plane. This raises a host of other possibilities for future biofeedback research. Future research may bear fruit by examining various sources of oscillation in the body and in behavior, to determine whether periodic stimulation of these oscillations may similarly exercise modulatory reflexes, and improve health. In particular, resonance and positive feedback loops could be considered, such as in heart rate variability biofeedback, wherein positive feedback can promote complexity and maintain oscillatory activity caused by negative feedback, helping individuals adapt to environmental demands. Applications of systems theories to social and organizational structures, other than for Gottman's approach to marital interactions, have tended to be nonmathematical, focusing primarily on norms, beliefs, and behaviors that preserve a particular pattern of social interaction. Time series analysis of these interaction patterns may yield important information that could be used as instruments for increasing social stability or social change.

Systems theory has long been applied in psychology, biology, and sociology. This paper applies newer methods of control systems modeling to the assessment of system stability in health and disease. Control systems can be characterized as open or closed systems with feedback loops. Feedback produces oscillatory activity, and the complexity of naturally occurring oscillatory patterns reflects the multiplicity of feedback mechanisms, such that many mechanisms operate simultaneously to control the system. Unstable systems, often associated with poor health, are characterized by absence of oscillation, random noise, or a very simple pattern of oscillation. This modeling approach can be applied to a diverse range of phenomena, including cardiovascular and brain activity, mood and thermal regulation, and social system stability. External system stressors such as disease, psychological stress, injury, or interpersonal conflict may perturb a system, yet simultaneously stimulate oscillatory processes and exercise control mechanisms. Resonance can occur in systems with negative feedback loops, causing high-amplitude oscillations at a single frequency. Resonance effects can be used to strengthen modulatory oscillations, but may obscure other information and control mechanisms, and weaken system stability. Positive as well as negative feedback loops are important for system function and stability. Examples are presented of oscillatory processes in heart rate variability, and regulation of autonomic, thermal, pancreatic and central nervous system processes, as well as in social/organizational systems such as marriages and business organizations. Resonance in negative feedback loops can help stimulate oscillations and exercise control reflexes, but also can deprive the system of important information. Empirical hypotheses derived from this approach are presented, including that moderate stress may enhance health and functioning.

Introduction COVID-19 is the third known animal coronavirus after Severe Acute Respiratory Syndrome and Middle East Respiratory Syndrome [1]. Although COVID-19 can be transmitted through droplets, close contact , aerosols and possibly oral-fecal transmission, patients can also transfer the virus to other people in the incubation period [2,3]. It is associated with SARS and MERS but shows some specific pathogenic, epidemiological and clinical features that are not fully understood to date [3]. There have been 26 468 031 confirmed cases of COVID-19 globally, including 871 166 deaths in the world and 382 772 confirmed cases of COVID-19 with 22 044 deaths in Iran based on the report of the World Health Organization on 5 September 2020 [4]. Patients with mild symptoms of COVID-19 without chronic conditions that are aware of the disease can be cared and treated at home by compliance with all precautions. Family caregivers should apply the principles of home care to these patients. Despite the existence of numerous epidemiological studies on the outbreak of disease, there is little information for supporting family caregivers of patients with COVID-19 [5]. The effects of COVID-19 may have impacted caregiving intensity, which is defined as the amount and type of care provided by informal caregivers [6]. It also may have affected caregivers' caregiver burden, which is defined as the impacts on physical and mental health, and healthrelated quality of life may also have changed due to the pandemic [7]. One study found that informal caregivers are likely at greater risk of adverse effects on their health due to the pandemic as they have been found to have particularly more stress and suffer from more depression and physical problems [8]. The different nature of COVID-19 and the need to maintain public safety will limit the physical presence with patients and, therefore, caregiving in this situation will be unique for everyone involved. The importance of maintaining social distance to control the COVID-19 pandemic probably increased isolation, loneliness, psychological stress and other adverse health problems [9]. Social and family relationships are disrupted not only for patients but also for their caregivers [10]. Mirzaei et al. indicated that caregivers of patients with COVID-19 face greater challenges compared to other caregivers because of limited training and resources at hand and their lack of knowledge about this emergent disease and the way to care for the patients [11]. The results of a review showed that sleep disorders, fatigue and inadequate self-care were recognized as common physical health problems in family caregivers [12]. Other qualitative and quantitative studies show that family caregivers may also grapple with anxiety, depression, apathy, disappointment, loneliness and isolation because of reduced social interactions and social exclusion while providing that care [13,14]. COVID-19 conditions have overwhelmed family caregivers, as well as others in the community, with new economic stressors, such as job loss by the household head and high medical expenses. Many family caregivers have other roles in addition to the caregiving role, such as managing their job, housekeeping, child care and additional responsibilities and concerns about their children's education due to school closures from COVID-19. They also must strive to protect themselves and other family members from virus transmission, and all of this is more difficult to manage than caring for other diseases or in other situations [15]. Abendroth et al. also argued that the anxiety about caring for a patient at home could be reduced by seeking information in caregivers as an adaptive strategy [16]. Given the importance of home care follow-up and selfcare education to patients and families, counseling centers and home care services are obliged to provide the necessary training to families to provide care services for patients at home according to the guidelines of the Ministry of Health and Medical Education [17]. Under WHO guidelines, any patient with a severe acute respiratory infection should be triaged and admitted to the hospital. However, people who have mild diseases do not require hospitalization and, to continue the treatment and care in the home, can be referred to health centers [18]. The acceptable criteria of WHO for home care of patients with COVID-19 are as follows: patients who do not tend to stay in the hospital, patients with mild symptoms and those who do not have systemic diseases, such as immune deficiency, cardiovascular disease or the risk of developing complications. The patient and family should comply with the treatment, including isolation at home, hand washing, respiratory hygiene , cleaning the environment and observe safety-related issues [19]. Evidence shows home-based caregiving effects on caregivers' quality of life and life satisfaction. Due to the combination of social distancing recommendations, stay-at-home orders, limitations on gatherings and the disproportionate impact of COVID-19 itself on mortality and morbidity among older adults, there is a critical and immediate need to understand the specific challenges and changes experienced in family caregivers [7]. Considering the peak of COVID-19 in Iran and the lack of human resources, physical space, protective and therapeutic equipment in the hospitalization of patients, many COVID-19 patients with mild to moderate symptoms are obliged to receive home care services. Since most of the studies have focused on the experiences of patients or health care staff, and few studies have been conducted on family caregiving experiences in Iran and at the global level, so the consideration of caregivers' experience and their problems is necessary to plan services and support them. Therefore, a better understanding of how the pandemic has impacted home caregivers can inform future interventions to address stress and health problems stemming from caregiving through the pandemic along with the everyday issues normally encountered. This study was conducted to explain the lived experiences of home caregivers in the families of patients with COVID-19. --- Methods --- Study design and setting A phenomenological approach was developed by Taylor and Bogdan in 1975 to the social sciences [20], but Diekelmann et al. applied hermeneutic phenomenological approach in health and nursing education [21]. So, this qualitative study was carried out using Dickelman's method. This study is used when the researcher needs the revelation of the little-known or unknown phenomenon through an in-depth examination of the lived experiences of people involved in that phenomenon [22]. The present study was conducted in the Center of COVID-19 Control at the Khoy University of Medical Sciences, Iran. --- Study participants The study participants consisted of 23 family caregivers who provided care for patients with COVID-19 at home and willingness to participate in research. Participants were selected using purposeful sampling method. Inclusion criteria are participants who have patients with COVID-19 at home based on laboratory tests and home caregivers who have cared for a patient with COVID-19 for at least 14 days and have not been affected by COVID-19 during the care of a patient. Exclusion criteria included participants who showed clinical signs of COVID-19, home caregivers who cannot take care of the patient, incapable of verbal communication and any psychological disorder that causes a lack of transfer the individual experiences. Sampling continued until data saturation was achieved when the researchers perceived that no new themes were emerging. Saturation in qualitative studies constitutes the completion of code levels, and new conceptual information requiring new code or expansion of the existing code is not obtained [23]. --- Data collection The semi-structured interviews were used for data collection. After explaining the goals and importance of research to the participants, the informed consent was obtained to participate and record the interviews. Participants were assured that the obtained information would be used only for the research and would remain confidential. The interview was planned and implemented concerning the qualitative questions of research and focusing on the experiences of home caregivers of patients with COVID-19 . Interviews were conducted individually with 23 participants at the private rooms of COVID-19 Control Headquarters, where there is no or minimal employees commuting to the place, having proper ventilation and a window to the outside of the building, while maintaining a social distancing of at least 2 m with the participant, and using gloves and masks. Each participant was interviewed only once, the total number of interviews was 23 and there was no conflicting data. The interviews lasted between 45 and 60 minutes and then interviews were transcribed by putting --- Analysis Phenomenological philosophy is based on believs of Taylor and Bogdan have dominated in the social sciences [20], but in the health and medical sciences especially in nursing used Dickelman's hermeneutic approach for interpreting data [21]. This approach used when the research purpose is understanding the human lived experience. In this way, the researcher categorizes the data by interpreting the discourse of participants [21]. So in this study, the interviews were analyzed using the principles of hermeneutics phenomenology as developed by Diekelmann. This qualitative approach was selected because it gives a high level of access to participants' lived experience [21]. The analysis was done by a six-member interpretive team and involves Dickelman's seven-step analysis . --- Trustworthiness Credibility, confirmability, transferability and dependability by Lincoln and Guba were employed to achieve the trustworthiness and validity of the study [24]. Credibility means that the findings of the study are real and that the findings reflect the research objective and social reality of the participants. To ensure the credibility of the findings, there was a prolonged engagement with the participants and data. The findings were shared with some of the participants. For this purpose, the transcribed interview and the extracted primery codes were presented to the participants, and they commented on its accuracy and, in case of any discrepancies, they were considered. To ensure the confirmability of the findings, the interviews and all procedures taken to conduct the study were recorded and complementary comments of research team were applied in data analysis. To check the dependability of the findings, an external reviewer who was familiar with both the clinical setting and qualitative research was asked to review and confirm the results. The transferability of data was achieved by maximum variation of the samples, such as different levels of age, education, occupation and family relationship with the patient. --- Results Participants' demographic details and their relationship to the patient are presented in Table 3. The themes emerged from --- First Reading and rereading descriptions -Transcription was read as a whole to gain an overall understanding and impression of the narrative. -Then, the narrative was reread line by line. --- Second Writing interpretive summaries and coding for emerging themes -Short interpretive summaries were written, and the common themes with exemplar quotations from the narrative were identified to support the interpretation. -Clusters of themes were merged and these classifications were named. --- Third Analyzing selected transcripts as a group to identify themes -Collaborative analysis was performed, to do these, meetings were held biweekly to discuss interpretations for similarities, differences or contradictions. --- Fourth Returning to the text or to the participants to clarify disagreements in interpretation and writing a composite analysis for each text -If conflict occurred or further interpretation was needed, the group returned to the original text and reread all texts to identify hidden meaning and link themes. -Or consulted the original interviewers for clarification Fifth Comparing and contrasting texts to identify and describe Shared practices and common meanings -The researcher compared and interpreted the relational themes to uncover the constitutive patterns existing within the relational. The constitutive pattern expressed the relationship of all the previously identified themes Sixth Validation of the interpretations -Draft themes and constructive patterns were observed by the researcher's supervisor. -All responses and suggestions received were integrated into the final draft. -The individual texts and audiotapes were also reviewed many times to allow the researcher to become further immersed in the hermeneutic circle, gain an in-depth understanding of each participants' experiences and further validated the interpretations. --- Seventh Eliciting responses and suggestions on a final draft -Excerpts from the participants' own words that reflected the strong meaningful transactions were included in the final written report. quotes of family caregivers . The codes extracted from the interviews led to the emergence of 13 primary concepts, 9 sub-themes and 3 main themes . --- Irrational fear of disease --- Fear of getting infected Because of the permanent intellectual engagement about the transmission of the COVID-19 to the caregiver and family members, frequent washing and disinfection of hands and home appliances was done. As some caregivers were suffering from respiratory allergies but, due to obsessive thoughts and fear about getting infected with the virus, could not emancipate themselves from these behaviors. Indulging in strength of immune system Some caregivers had too much sensitivity to strengthen the immune system through which they could reduce the risk of affliction with the disease. They tried to consume diverse herbal teas, the properties of some of which have not been approved, to strengthen the immune system. --- Increasing the burden of care in caregivers --- Lack of support Many patients and caregivers expressed that the government and medical team were unaware of their care problems at home. In most cases, patients are cared for in a shared room at home with other family members, and there was no separate room or suitable place in the house to care for them. Besides, necessary training about patient care was not given to the caregivers. Moreover, lack of financial resources, high cost of treatment and job interruptions caused financial burden in families and caregivers. --- Caregiver's family challenges Patient care by the family caregivers provides a context for the neglect and inattention of other family members. This issue is a predisposing factor for damage to family members. Therefore, many caregivers assumed that the pressure and stress they endure are more than patients. That is because the caregiver must not only meet the patient's needs but also be responsible for a lot of family problems and issues. --- Self-reinforcement --- Highlighting positive features Putting the caregivers under the challenging conditions of care provision caused the revelation of positive human characteristics, such as patience, strength and resilience in the face of hardships, calming down, increasing a sense of altruism and helping people in need. --- Turning to spirituality Given the exposure of caregivers to stressful events, such as caring for a patient with COVID-19, spiritual growth emerged in them and made them more resistant to the hardships caused by the care. There were things like a change of attitude in life, deepening the relationship with God, appreciation for blessings, being thankful for God's blessings and hoping in God in the direct quotes of these participants. --- Discussion The themes obtained from the interview with the participants were an attempt to achieve the main purpose of the study that was to explore the experience of home caregivers of patients with COVID-19. --- Statement of principal findings The results of this study showed that long-term care, performing tasks related to patients 'daily activities and patients' psychological problems, along with the physical fatigue of caregivers, had caused mental disorders, such as burden, obsessive behaviors and fear of disease transmission in caregivers. Of course, in addition, the participants had positive experiences, such as tending or serving to strengthen itself. Therefore, it can be said that the disease had positive and negative consequences on caregivers. The development of obsessive behaviors, increasing the burden of care in caregivers and high human achievements are important themes of the present study. --- Interpretation within the context of the wider literature The irrational fear of disease was one of the main themes. Participants stated that the thoughts of fear of transmitting infection and disease automatically enter their minds. Due to these obsessive-compulsive behaviors, the duration and frequency of hand washing and sterilization of equipment were dramatically enhanced. Caregivers said they were too sensitive to boost the immune system to help the patient to overcome the virus and prevent other family members from becoming infected. In this respect, the study of Brooks et al. demonstrated that the fear of disease transmission and concern of getting infected with the COVID-19 increases dramatically in critical situations and epidemics. This concern in families with small children or pregnant women is much more than in other families [25]. Perhaps many people are not familiar with the nature of COVID-19, and infection is stigmatized as the patient not respecting their health and not following proper hygiene practices, therefore developing obsessive behaviors. et al. also revealed that some factors could influence increasing the care burden. Some of these factors are concerns associated with the risk of afflicting disease, ambiguity in the status of career prospects, lack of income-generating resources, lack of adequate facilities at home in the period of home care and the lack of sufficient support of the medical and governmental system of patients after discharge [28]. Asgari et al. identified several challenges in family caregivers of patients with COVID-19, the main themes of the study were 'captured in a whirlpool of time' and 'feeling helpless' and 'loneliness'. A part of these experiences was due to careers, to protect other family members and friends against getting the virus, isolating themselves due to a fear of others becoming infected [29]. Thus, providing relevant information on symptoms, medical and protective care, support in preventing disease complications and identifying the factors that cause the burden of care on the caregiver is essential [30]. Many participants had economic concerns. The families of patients who are cared for at home and can not go to work suffer from great economic pressure [10]. The research performed by Brooks et al. indicated that financial concerns are among the adverse psychological outcomes of closure because of home quarantine. Governments and policymakers should make the necessary mutual trust to people in the community and offer financial compensation for their losses [25]. Furthermore, the findings of the study conducted by Hertz-Palmor et al. indicated that the most significant psychological problem that arose after the recovery of SARS in China was the reduction of financial resources and household incomes so that it was regarded as the highest predictor of reducing mental health [31]. One of the most prominent positive consequences caused by caring for family caregivers was self-reinforcement. The caregivers felt that they had become human beings who were more insightful and mature in the care process. They expressed that caring has led to a deeper relationship with God, appreciation of blessings, gratitude for God's blessings, better acceptance of situation and a positive direction in life. Investigations have also indicated that those who use positive strategies to cope with stress decrease adverse outcomes and increase positive results [32]. Religious behaviors and beliefs have a positive impact on the meaningfulness of life. Behaviors such as trust in God, worship and pilgrimage can bring inner peace to individuals through creating hope and encouraging positive attitudes, and they can incur less damage in the exposure of stressful life events [33]. Besides, divine fate has been taken into consideration from two perspectives. Some people consider it as a God bless; others believe that these events awake people and are opportunities for people to be prepared for the present situation [29]. --- Strengths and limitations One of the limitations in this study was the personal characteristics and mental concerns of the participants affecting their response. Therefore, by explaining to the participants before the interviews and conducting in-depth interviews, the researcher tried to minimize this limitation. In the case of lacking preparedness of participants, another time was considered for the interview. One of the strengths of the present study is variation in sampling. In order to consider all influential factors in the occurrence of the phenomenon, sampling of caregivers in different age, gender, cultural, social and economic areas was performed. --- Implications for policy, practice and research Health care providers are required to consider the role of family members in providing care for patients and their home care challenges. Also, it is recommended that interventions such as education to the patient and his/her caregiver, counseling, family therapy and referral to financial support groups to reduce the pressures of caring should be considered. This can both promote the quality of patient care and guarantee the physical and mental health of caregivers as unknown patients. It is recommended for the health staff to check the health of these people regularly via telephone and, if possible, provide in-person visits daily. Besides, sharing positive experiences of family caregivers allows the researchers to design interventions to decline negative consequences. --- Conclusion This study implies that much attention should be paid to psychosocial, physical, mental and family relationships health of caregiver to provide high-quality care to the patient. We can conclude that the level of care pressure in caregivers of COVID-19 patients is high. It is possible that these pressures reduce the level of patient care and also jeopardize the physical and mental health of caregivers. It can also turn the families of these patients into vulnerable families who require intervention, counseling and follow-up over time. the optionality of participating in the study, the confidentiality of the information obtained and the possibility of withdrawal from the study at any stage were provided to the participants. --- Data availability statement The data that support the findings of this study are available from the corresponding author upon reasonable request. --- Quotation Primary concepts 'Every day, I have to disinfect the washbasin, the bathroom, the room surface, the door handles, and the dishes several times. On the other hand, I can't wash my spouse's clothes in the washing machine, and I have to wash them into the basin and pour disinfectant liquid on them. Doing all these things really takes a lot of my time, and I get tired. However, if I don't do these things, the children or I will get infected'. 'My health has changed a lot since my husband took the corona. I use masks and gels regularly and wash my hands frequently with soap and water'. 'Whatever I touch, I wash my hands, for example I disinfect the door handle regularly. I'm really afraid that the disease will be transmitted to me'. 'We washed our hands a lot with soap and alcohol, we became obsessed. When we go out on the street, if we do not shake hands, we will go home again after washing and wash our hands. Well, one is afraid'. Frequent washing of hands 'Because of the constant washing of hands and disinfection of furniture and kitchen utensils with Javel Water before each use, I am suffering from lung problems, even my child has an obsession and washes his/her hands constantly'. "Because my husband has a heart disease, both he and I were very upset about Corona. I even disinfect the inside and outside of the car and then I get in the car. I disinfect the phones several times a day. All of the things that I buy from the outside, first take out the bags and then, if they are washable, I wash all of them or disinfect their surface. 'Previously, if I washed the fruit once, now I put the fruits in a pan, disinfect them with vinegar and salt, leave them for half an hour, and then wash them again'. --- Continuous disinfection of surfaces 'Permanently, I give my daughter strengthening foods and a variety of effervescent tablets at home to eat them and recover sooner because it is believed that Vitamin C and D are excellent'. 'The doctor said that I should give fruits that contain high vitamin C to my son at home because it strengthens the immune system. I also buy oranges, tangerines, lemons, grapefruits, oranges and kiwis wherever I saw, and then I would take the juice and let it eat'. Excessive use of vitamins 'My wife and me drink ginger and lemon herbal tea every day so that our bodies to be strong. I became highly sensitive to the immune system of myself and my family'. 'I knew that they get Corona, should drink plenty of warm fluids, so I went to a store of medicinal herbs and bought mint, thyme, chamomile and a few other herbs in combination and gave them to my parents to drink every day in a form of herbal tea'. --- Consumption of herbal teas 'Since my husband got afflicted with COVID-19, he does not go to work for about two months, and I think someone else was replaced to work in the factory. I don't know what to do; we have a lot of loans and installments that we must pay for them. On the other hand, we have to think about the monthly rent of the home and the cost of our needs'. 'Corona has caused part of the cost of living to be spent on prevention supplies. But now my husband cannot go to work due to illness. I also spent all the money we had saved during this time and I do not know what will happen next. We were really suffered financially'. --- Economic concerns 'There are strict rules in the hospital that do not let you see a doctor directly; additionally, they are always busy. They do not matter at all what my problems are concerning the care of my mother? And how do I deal with the problems? In addition to her illness, my mother is also depressed. She does not talk to anyone. These do not care for the medical staff'. 'Well, in terms of the potential of finding out, understanding, or effective use of health information, everyone is different. Some have enough information and skills, and others do not. I went to search for it and obtained the information; no one taught me, but many people, especially the elderly, do not have the condition and cannot do it'. Lack of responsiveness of the medical team after discharge 'During this period, I have taken after my father and done his work since I am the only son in the family ; however, because of quarantine, I do not have access to my wife and child, and I cannot go to them. This has caused problems in my family'. 'I take care of my husband who took the corona in our house and sent the children to their grandmother's house. I was worried that would get a corona and that my children would be without a mother. It has been about two months since I did not huge my children, fed them or slept with them'. 'I was worried about my mother having diabetes and heart disease. I had nightmares that my mother had taken a corona. I am very dependent on my mother and I am always worried about losing her'. Isolation of the caregiver 'The patient's caregiver is more engaged in the situation and is under pressure than the patient himself because the caregiver follows up all the patient's work and needs. On the one hand, he is looking for drugs; on the other hand, looking for the cost of living. Moreover, he has to do all the sick work at home. Now, judge for yourself'. 'During this time I went to work and took care of the patient at home. I felt I had no strength left. Most of the relatives would call to ask for help, but I did not even have enough time talk to them'. --- Quotation Primary concepts 'We did not tell any of our relatives or neighbors about my father's disease, for we knew that if they found out, they were bothering us, would constantly ask how he was afflicted with the disease, why he did not respect, and they look at us completely differently as if we have a leprosy patient at home'. 'Community perception of COVID-19 is not just a disease and is not limited to a person with the disease, but is identified as a person who most likely did not have health behaviors and as a result of these unhealthy and careless behaviors, they have get corona. So we did not want anyone to know about the disease'. Hiding the disease from relatives and others 'During this period, I became stronger and more patient. I think I can tolerate difficult situations more easily. Difficult situations can be in terms of finance, loneliness, and problems that can happen to humans'. 'My husband is a touchy and irritable person. He has made a thousand excuses since he got sick and I take care of him at home. But I think God has increased my patience and strength so that I can take care of my husband and children. Now I am facing the problems of life alone, well, this disease, which is not permanent, we have to increase our patience and endurance'. --- Patience in difficult conditions 'Now, I have become more discerning, I have realized and understood my mother's love for myself more. I have found that if I am at this position at present, it is because of my mother's forgiveness and sacrifice'. 'Nothing makes me feel better and happier than seeing my wife calm and well, thank God I am by her side and I can help her'. Opportunity for love 'Now, many things have changed a lot, I appreciate life more, I appreciate what God has bestowed me, and I appreciate my health so much more than before, maybe I was a little unaware of it, I thank God for all this'. 'In general, caring has given me knowledge, the knowledge of the path I am going to take, where I have made mistakes in life and where I have used the blessings of God correctly, it has made me aware that I am always grateful for the health that God has given me. In a word, I want to say that I became another human being'. --- Appreciation for blessings 'This disease influenced me a lot. I got closer to God, the opening out of a heart for God makes me very calm, and it is the reassurance of my heart. I am less ungrateful, and I always think He knows the best of everything'. 'It is true that I did not leave anything for my son at home, but I was also worried about his condition. Every time I prayed and recited the Quran, and I even vowed that God would cure my son of this disease. God has helped me everywhere and I am sure that he will never leave me alone'. Due to the fear of jeopardizing the family's health, participants were trying to strengthen their immune systems. In this regard, Caleo et al. indicated that individuals in critical conditions are incredibly concerned with their family situation and are continuously seeking extreme care for them [26]. Moreover, Louca et al. study showed that a number of micronutrients have been shown to play key roles in supporting immune function and in reducing the risk of respiratory infection; also, they observed modest but significant association between the use of dietary supplements and lower risk of testing positive for SARS-CoV-2. However, the results were significant only in women but not significant in men [27]. Considering that, it is reasonable to hypothesize that vitamin supplementations may enhance host immune responses against COVID-19. --- Nearness and recourse to God The problems of patients, lack of support for the patient and family, are the basis of many stresses that targeted the participants' souls. Hence, one of the themes derived from this study was the increased burden of care in caregivers. Wang --- Contributorship All authors participated in the process of the initial drafting of the article, review, presentation of the initial idea and design, collection of the data, analysis and interpretation of the data, and all of them take responsibility for the accuracy of the material contained in the study. --- Ethics and other permissions This paper is part of the results of a research project approved by Khoy University of Medical Sciences, Iran, which has been approved by the ethics committee of that university . Participants were given enough information, and each person signed an informed consent form to participate in the research and audio recording during the interview. Moreover, the necessary explanations about

Background: Following the spread of COVID-19, many patients received home care services for recovery. The family is one of the informal caregivers who provide daily physical and psychological support to the patient, and they have the most contributions in the care of patients. Objective: This study explains the experiences of home caregivers in the families of patients with COVID-19. Methods: This is a qualitative study with a hermeneutic phenomenological approach. Purposeful sampling was applied, and semi-structured interviews in the Center of COVID-19 Control in Khoy, Iran, were used to collect the information. Twenty-three family caregivers, who had already provided care for patients with COVID-19 at home, participated in the study. The themes emerged from interviews. The interviews were recorded and transcripted and then analyzed using Dickelman's interpretive phenomenological approach. Results: Thirteen primary concepts, six sub-themes and three main themes were extracted from the data analysis, including 'Irrational fear of disease' with sub-themes: fear of getting infected and indulging in strength of immune system; 'Increasing the burden of care in caregivers' with sub-themes: lack of support and caregiver's family challenges; and 'Self-reinforcement' with sub-themes: highlighting positive features and turning to spirituality. Conclusion: Understanding the complexities, experiences and beliefs of family caregivers about living with a COVID-19 patient provides a comprehensive perception of the psychological and physical consequences of care. Executive decision-makers, health care personnel and mental health professionals can also take the necessary strategies to support and manage home caregivers and interdisciplinary cooperation.

Introduction Is economics a science or a social science? Arguments have been made for both orientations, i.e. that economics is a social science , a science and even a moral science . In terms of sciences it has been linked with physics, with many physicists doubling as, or transforming into, economists and a subpart of physics, econophysics, specifically evolving which deals with economics . There are also links with biology . Frey argues economics is a social science as it is part of those sciences which deal with actual problems of society, i.e. it is the subject matter which makes it a social science. However, he also points out that in practice economists tend to fill the journals with axioms, lemmas and proofs, i.e. they adopt what they perceive as a scientific, and particularly a mathematical, methodology. Mayer argues that economists act as though economics is a hard science and this is reflected by their sophisticated use of mathematics. However, he then goes on to argue that econometrics is not sufficiently advanced to enable us to test theories as a hard scientist would. With respect to this argument, it is worth noting that econometrics has moved a long way since 1980, when Mayer wrote his paper and perhaps that objection is not as valid today as it was then. In any case, part of the reason for this disagreement as to the nature of economics may be because the subject matter of economics deals with people, whether as agents in firms or consumers, either as individuals or aggregates. This will tend to place it as a social science. But in terms of methodology it has arguably closer links with the sciences. It is essentially a quantitative discipline. Theoretical models are built to replicate and simplify some real world situation. Empirical work is focused on econometrics and many of the techniques used and developed by economists are also used outside the discipline and often in the sciences, such as biological sciences. Economics has of course evolved over the years. Fourcade et al. argues that for much of the period since 1945, a theoretical approach based on rigorous mathematical modelling was the main path to establishing scientific purity in economics. However, in the 1990s and 2000s there occurred a shift in emphasis from theory to empirical work. Einav and Levin potentially link this shift to a substantial increase in data availability. They argue that this may change the emphasis on setting up and testing hypotheses to one of searching for the best possible explanation. It may also further move economics away from the emphasis on theory, all of which may in turn impact on how economists perceive themselves and the relationship of economics with other disciplines. The quantitative nature of most economics research is in contrast to the qualitative methods that characterize the work of many other social scientists. Mankiw epitomises this perception, when he admitted he thought of himself as, and sought to project the image of, a scientist. He admitted that this might amuse physics academics, but he would remind them that economists formulate theories with mathematical precision, collect huge data sets1 on individual and aggregate behaviour, and exploit the most sophisticated statistical techniques to reach empirical judgments that are, potentially at least, free of bias and ideology. However he also thought that economists could be likened to engineers, having been put on earth not to propose and test elegant theories, but to solve practical problems. This perception that economics is different from the social sciences may also have contributed to the insularity of the discipline, as reflected for example in the relative dislike of economists for interdisciplinary work when contrasted with other social scientists , as well as a reluctance to cite work from these other disciplines . There have been a number of attempts to map academic disciplines on the basis of citation or similar data. It is often done for specific disciplines, rather than science as a whole. Thus McCain analyses co-citation data, 2 using cluster analysis and multidimensional scaling, to produce maps of highly cited and co-cited documents in clusters representing interrelated research activities. The focus is on mapping economics journals, and is able to identify several major groups comprising: the 'core of economics', a West European set of journals, formal journals, basically key theory and econometrics journals, public finance journals and economics journals. 3 A number of studies do attempt to map the full range of disciplines. Rafols et al. using a matrix of similarity measures based on correlations of different pieces of information items, conclude that science is a fragmentary structure composed of both solid clusters and empty spaces. This gives support to the concept of independent, but potentially linked disciplines, rather than a continuum of study with somewhat artificial boundaries. The analysis is based on citing similarities among ISI subject data. There is an important biomedical research cluster and a major physical sciences cluster, comprising engineering, physics and material sciences. There is a rather diffuse social sciences cluster, which Rafols et al. argue is due to relatively low citation rates. Economics is part of this cluster, but with some links to mathematics and computer science. Bollen et al. use click stream data to construct a journal network that outlines the relationships among various scientific domains. Economics is on the periphery of the social sciences cluster. But it is also one side of a bridge connecting the social sciences to physics and engineering via production and manufacturing. Hence these mappings tend to loosely tie economics to the social sciences, but often with some linkage to parts of the sciences. However most of these analyses are based on citations or similar measures, rather than the similarities in the structure of academic work. We have already emphasised that economics analyses social science issues, but that it is the methodology or the style of working which differentiates it from the social sciences and links it to the sciences. It is almost inevitable that economics' citations will be greater in social sciences, than the sciences. But this tells us relatively little about the similarities between styles of working. There is little hard evidence on subject linkages based on different styles and in order to provide the beginnings of a perspective on this, we compare the journal title characteristics of the different disciplines. The characteristics we will focus on include title length, average word length, the proportion of papers containing a colon, the proportion containing a question mark, the average publication date and the number of authors. In particular we will extract the economics' footprint as reflected in these title characteristics. We will then compare this footprint with those of other disciplines. We will argue that each discipline tends to adopt a certain style in terms of these characteristics. Lewison and Hartley argue, for example, that there are substantial differences in title length, use of colons and question marks, etc. between disciplines. Nagano also emphasises different disciplinary conventions with respect to the title, distinguishing between the hard and soft sciences. Different disciplines also tend to differ in the number of authors and the extent to which journal papers are important compared to, for example, books. If disciplines have their own style, then it seems reasonable to assume that there are also similarities between contiguous disciplines. For example, we would expect, as suggested in the literature, the characteristics of a social science discipline to be more closely aligned with other social science disciplines than ones in the sciences. Following on from this, we will then use cluster analysis to place the different disciplines within groups to see how well they fit into the social science, science and art and humanities categories. In the process we will also identify the group and subjects to which economics most closely belongs. In doing this analysis we make use of a unique data base obtained from the Research Excellence Framework4 of 2014. This is the latest in a series of exercises seeking to evaluate the quality of research done in UK universities across a range of subject areastermed units of assessment . In this REF there were 36 such UoAs, as listed in Table 1, including one for economics and econometrics. 5 Table 1 also gives the number of papers on which our analysis is based. In economics there were 2600 individual pieces of work submitted. Not all were journal papers and a proportion were submitted more than one once to the economics panel, i.e. by definition they were jointly authored papers. This left 2246 separate papers to analyse. Each entry to a UoA was scrutinised by a sub-panel working within the framework of 4 main panels-A-D. Economics was in Panel C along with the other social sciences. The main scrutiny was on the quality of academic research contributions, primarily journal articles and books, which were individually looked at and assessed by panel members. Each participant in the REF could submit up to four research outputs . Each submission was also accompanied by a research environment statement and a number of case studies which highlighted the impact of the university's research on society or the economy, widely defined . The evaluation thus comprised three components: the quality of the four research outputs submitted by individuals, the research environment statement and the impact of the research on society or the economy. Our focus is on the publications. Hence, and this is an important prerequisite for the study, the definition of the discipline of the sciences and social sciences is based on the REF's subject-classification system. In the next section we present the summary data on title characteristics. We then use cluster analysis to group the different disciplines together and finally conclude the paper. --- The differing characteristics of disciplines in terms of journal titles Table 1 shows the characteristics of the titles, i.e. median title length, median word length, proportion of papers containing a colon or a question mark, the average date of submitted papers, etc. These calculations are done for the main subject groupings. The data relate to journal titles only and these form the basis for the cluster analysis. 6 Focusing on journal papers, the penultimate two columns show that in economics 8.05% of papers were also submitted in other disciplines, whilst 6.38% of papers were submitted at least twice within the economics UoA. In the latter case they would most likely have been submitted by coauthors in different institutions. Multiple submissions of the same paper across disciplines were particularly common in the health and engineering UoAs. They were much less common in the arts and humanities. Multiple submissions within a discipline were relatively common in the sciences, particularly physics, and also business, but less common in engineering UoAs. Apart from the data on multiple submissions, the other columns are calculated on papers rather than submissions. Hence if a paper was submitted more than once to the economics panel, it would still only be counted once in calculating the statistics Columns median character length of title, median author numbers, average publication date, % using a colon, % using question mark, % of submissions which are journal papers, median word length, % of multiple publications in same UoA, % of multiple publications in other UoAs, number of journal papers in UoA which form the sample for our analysis. Source: calculated from data on journal papers in the REF for these columns. The number of papers this left for analysis is shown in the final column of the Table . The median number of authors in journal articles for economics was 2, this is on the low side with a median across all disciplines of 3. The average was much higher at 11, being pushed up by the very high average number in physics, 92.4. But in general across the sciences and health disciplines the number of authors tends to be high. The average can be heavily influenced by one or two outliers. In physics for example, the most number of authors is 3269, for a paper on the Higgs boson. In the REF, this paper was submitted 13 times. This is why in this current analysis for the number of authors and also title length, average word length and the number of citations we work with the median. That is we calculate, for example, the median title length of all the titles in the sample and use that instead of the mean title length. The average date of submission also shows systematic variations across the disciplines. It tends to be earliest for the sciences and engineering and latest for the arts. Economics is quite late. This may reflect the ease of writing a new paper in the different disciplines. In the sciences a grant is often needed and also the team of people tends to be larger than in the non-sciences. Thus in the non-sciences it may be easier to write a new paper if current REF output is considered unsatisfactory. However in order for the ease of writing new papers to explain the timing of submissions, we also need the starting date for research to be influenced by the REF cycle . The other columns provide us with information on the way economists work in comparison with other disciplines. Firstly the proportion of submissions which are journal papers was quite high at 91.79%. This reflects the importance economists place on journal papers. Other submissions include books and also working papers. However, the proportion of journal papers is still higher across much of the sciences and engineering. But it does tend to fall substantially for other disciplines such as law, history and languages. Thus in classics, less than a third of submissions are journal papers. There is also information which helps us compare economics with other disciplines in the structure of the title. Economists tend to be parsimonious in title length relative to other disciplines, with a median number of characters of 64, only philosophy having a shorter median title length. 7 The longest titles tend to be in the health disciplines and sports sciences, which is linked to health. Economists are however less parsimonious in their use of colons. The heaviest users of colons tend to be the arts and social sciences. In sociology over 66% of submissions feature a colon and in business it was almost 53%. Colons were least common in mathematics and electrical engineering. The same is true with respect to question marks. Finally economists tend to use shorter words than the science and health disciplines and longer ones than the arts and social sciences. 8Hence we can see that different disciplines have different styles and practices with respect to journal titles, together with the number of authors. The sciences, including the health sciences, tend to be characterised, for example, by a large number of authors and be parsimonious in using question marks and colons, even though their titles tend to be relatively lengthy. This raises the possibility of categorising economics with respect to other disciplines, at least with extent to title characteristics, the number of authors and the submission dates for the REF, by the use of cluster analysis. The basic assumption is that people in contiguous disciplines work in a certain style which tends to characterise those disciplines. Two disciplines which are close will tend to have similar characteristics in terms of numbers of authors and the characteristics of the title. Apart from providing information on different title characteristics across disciplines, it is also worth noting that characteristics such as title length, use of colons and question marks have been found to impact on citations . --- Grouping disciplines We use cluster analysis to group disciplines based on title length, number of authors, date of submission, use of colons and question marks, word length of journal titles and proportion of submissions in the form of journal papers. The basic idea of cluster analysis is to group, first, the two disciplines that are closest using a Euclidean distance criterion. In some techniques the squared Euclidean distance is used. The Euclidean distance is formed by finding the difference in the values of the first variable between two disciplines and squaring the result, repeating for each variable, adding the all the squared differences together and taking the square root of this. Out of all the disciplines in the study, the pair that are closest join first and lowest in the dendrogram . When groups are being joined together, some measure of the collection of values of a variable is needed and various possibilities are discussed later. We consider only journal articles, and not other publications such as books which may in themselves follow a different format in their title than do journal papers. The percentages of joint papers shown in the penultimate two columns of Table 1 were not included in the cluster analysis as they will tend to reflect the number of authors, and hence this would be to an extent double counting this characteristic. In addition, the number submitted outside the discipline partly depend upon how the boundaries are drawn by the REF itself, rather than inherent characteristics of the discipline. We standardised the data by transforming each variable into a Z variable. This is frequently done when the different variables are substantially different . This was done in STATA and the algorithms used to calculate the different clusters using different assumptions were also those which are given in STATA. Figure 1 shows the dendrogram calculated using the method of average linkage. As its name suggests, this uses the average over all papers in a cluster for each variable as the measure of cluster location. The average linkage method tends to work well in many situations and is reasonably robust . The diagram shows the Euclidean distance. Two subject matter areas that have highly similar characteristics join low in the diagram, hence aeronautical and general engineering are highly similar. At higher values of the vertical axis, clustered pairs are joined by a third area, or group of areas, with largely similar characteristics to the pair. Other areas that have somewhat similar characteristics are also paired. Moving up the diagram, the number of clusters reduces and each contains more heterogeneous, although somewhat similar, areas. Economics is initially grouped with philosophy, although quite late in the clustering process, hence the linkage is not that strong. At the same approximate level of dissimilarity is a second group involving architecture, geography, sports science, business, social work, education, sociology, anthropology, area studies, law and politics. These correspond closely to the social science disciplines and REF panel C. This group is subsequently linked to a second grouping comprising modern languages, media studies, history, theology, English language, music, the arts and classics, which largely correspond to the arts and humanities and REF panel D. It is these two combined groups that economics and philosophy eventually are linked. There is then one further large grouping. This is made up of sub groupings of allied health, agriculture, biological sciences, psychology and earth systems, which largely come from Panel A, the health sciences, and the four engineering disciplines which come from panel B, the sciences. These two sub groups merge and are then joined by chemistry, physics, a sub group of maths and computer science, and finally clinical medicine and public health. To a large extent the clusterings are intuitively plausible. But there are some surprises. For example, the clustering of sports science with architecture and geography. However this does take place quite late which suggests that the linkage with sports sciences and the other two disciplines is not that strong. Even so there are plausible connections, with geography forming a bridge between the other two. The full title of sports science is sport and exercise sciences, leisure and tourism, and the latter two areas make the linkage with geography plausible. Similarly with geography and the last two elements of architecture, built environment and planning. But the linking of psychology, psychiatry and neuroscience with earth systems and environmental sciences is more difficult to rationalise. However the subsequent linking of this subgroup with allied health, biological sciences and agriculture, veterinary and food science is easier to rationalise, e.g. the link of earth systems with agriculture and biological sciences. Hence we can see that groupings based on journal paper characteristics categorise the different disciplines into the sciences, social sciences and the arts in a manner which largely supports the conventional groupings, as can be seen from Table 2, where these clusters are listed. We choose seven clusters9 as this allows the main clusters to emerge, given that some UoAs such as clinical medicine will tend to be on their own, not linking to other UoAs. In this case both the social sciences and the arts and humanities are well identified. Most of the sciences in both panels A and B form a single cluster, with the exceptions being clinical medicine, public health, chemistry and physics. Mathematics and Cluster A fitted by average linkage; B, ward's linkage; C, single linkage; D complete linkage. E is fitted by average linkage excluding the number of authors computer science also form their own sub-group. Economics does not fit in well with the other social sciences, nor indeed does it fit in with the sciences. It fits best with philosophy, although the fit is not that strong and if we move to nine clusters it is on its own, not linked with philosophy. If we use other methods 10 to identify the clusters, the results viz a viz economics change relatively little as can also be seen from Table 2. For example, when using Ward's method, there is a clearer split identifying panels A and B, and the social sciences form two clusters-which join together again when we reduce the number of clusters to 6. The dendrograms give more insight than the simple identification of clusters, but what this Table also reveals is that while the contents of a cluster might be identical or almost identical under different linkage assumptions, the ordering of clusters depends on the algorithm and other assumptions. The final column shows the clusters using average linkage when we omit the number of authors. The results are similar to before, although social sciences now has two sub clusters and in panel B, chemistry and physics are now part of the other sciences, but with maths and computing excluded. But still the clusters are based around the different REF panels and hence these results are not being driven by large differences in the number of authors between disciplines. The same is true if we omit the proportion of papers which are journal papers. Fig. 2 Dendrogram based on average linkage with sub-groups for economics 10 Ward's method is based on increasing the overall within-cluster variance to the smallest possible degree. Single linkage is based on the minimum distance and complete linkage is based on the maximum distance. 11 If we repeat the analysis but using the raw data, not the standardised one, then there are some changes. The clusters still identify the different panels well, but now economics is weakly linked with mathematics and computer science. Again there is no evidence of a linkage with either the rest of the sciences or the social sciences. --- Splitting up economics In this section we divide the economics papers into three groups, those published in theory journals, those published in econometrics journals and all the remaining papers. We used the same categorisation principle as Hudson , i.e. a journal was counted as a theory journal if some version of the word 'theory' appeared in the title and similarly for econometrics with respect to variations on 'econometrics'. For this exercise, the journal Econometric Theory was counted as an econometrics journal. Table 1 shows the summary data on these three sub groups of economics. There are substantial differences, with titles of both theory and econometrics journals having significantly, at the 1% level, lower proportions of papers with both colons and question marks. The titles of papers in theory journals are exceptionally parsimonious, whilst their average word length is high relative to other disciplines. Although not as high as papers in theory journals, word length in econometrics journals also tends to be high relative to other disciplines. The penultimate column shows that a submission which spans disciplines is relatively uncommon for papers in theory and econometrics journals, but slightly more common in the rest of economics. Redoing the cluster analysis in a manner consistent with that shown in Fig. 1 results in the dendrogram shown in Fig. 2.12 Papers published in econometrics journals are linked in with maths and computing as subsequently are papers published in theory journals, they thence join with the sciences, although at a high level of dissimilarity. The rest of economics now links with business and thence to the social sciences. This result does not change if we use the other methods for clustering. The clustering algorithm is a hierarchical one and expanding the disciplines has had only limited repercussions on the structure of the other clusters, and this involves mainly business and philosophy. The close tie of econometrics to mathematics is not that surprising, although that with computer science a little more so. Perhaps the most surprising conclusion is that ''other economics'' ties with business and follows a completely different path up the dendrogram to that of theory and econometrics. It is this second part which is linked with the social sciences. These differences suggest that economics is a broad discipline compared with some others, although it is also a discipline with a degree of heterogeneity. It is also not surprising that these clusters are different to those that we get using citation analysis . The latter will relate primarily to subject matter rather than to academic style of working. The REF data provides more information on the linkages between economics and other subjects via the submission to other UoAs of papers also submitted to the economics UoA. The proportion of joint submissions, shown in the penultimate column of Table 1, for the theory and econometrics journals combined is significantly lower, at the 1% level, than for the rest of economics. 13 Ignoring Business UoA 'partners', for theory and econometrics seven out of twelve 'partners' were in mathematics or computer science. For the rest of economics the spread was wider. Out of 71 non-business UoA 'partners', nine were in the mathematics or computing UoAs, with the other largest groupings being public health , history , social work and social policy and education . These linkages will partially reflect closeness of subject matter, as in citations, but also possibly closeness in working styles. 14 The results, albeit with small samples, again suggest that the more quantitative part of economics is closely linked with mathematics and computing, but that the rest of the discipline is spread more widely across parts of the sciences, the social sciences and also the arts. --- Conclusions This paper represents one of the first attempts to identify, using quantitative analysis, the linkages between different disciplines based on the stylistic characteristics of journal papers. We are able to identify subjects, as defined in the REF, into clusters which closely resemble those which form the four main REF panels and also more generally what are thought of as the sciences, the social sciences and the arts and humanities. This indicates both that contiguous disciplines have similar style characteristics evident in the title of the journal papers in those disciplines, and that using these we are able to link the different disciplines together in a manner which broadly is reflective of the social sciences, sciences and arts and humanities as they are generally understood. Thus linking cluster 3 to panels A and B, cluster 5 to Panel C and cluster 6 to Panel D as in column A of Table 2, correctly identifies 29 of the 36 disciplines. The disciplines of physics and public health however, are difficult to categorise and maths and computer science form a separate group from the rest of the sciences. Economics as a whole is linked with philosophy, although not strongly so. Only area studies is allocated to a 'wrong' panel. The fact that we have been able to successfully allocate disciplines to clusters which so closely resemble the REF panel groupings suggests that our methodology is sound. On the basis of our analysis, it seems possible that most of economics does not closely link with any of the other disciplines. Economics tends to have relatively few authors, but more than the arts and humanities, tends to be parsimonious in title length and to an extent, particularly when compared with the other social sciences and the arts and humanities, in the use of colons. When we look at sub-disciplines within economics, we find that econometrics first links with maths and these are then joined by economic theory and computer science. Business and the rest of economics link together and thence with the rest of the social sciences, although at a relatively high level of dissimilarity and hence the linkage is not that strong. However there is little evidence of a link with the sciences per se. This is not unexpected, large grants involving several labs have become increasingly important in the sciences, at least as represented by the leading papers . This may have tended to increase author numbers and this too will leave a footprint on title styles . Of course grants are important in economics too, but not to the same extent as in big science. With respect to the question posed in the title of this paper, our view, in part informed by these results, is that the subject matter of economics places it clearly as a social science, but many in the discipline act as if it were a science. Economics has gradually developed into an interdisciplinary subject that uses scientific methods, for example setting out and formally testing hypotheses, and particularly a mathematical approach, to solve social science problems. This dual aspect may explain why it fails to fit neatly with either as in fact it combines elements of both. It is a social science pursued with more quantitative rigour than much of the rest of the social sciences. This emphasis on technique then pushes economics and economists away from the other social sciences and in some sub-disciplines towards maths. Placing economics as a science, emphasises technique above subject matter. But there is then a danger that technique becomes an end in its own right, rather than a means to an end, i.e. a vehicle which allows a more refined analysis of economic issues which have wider relevance. Often, for example when developing a new econometric technique which subsequently becomes widely used, this is justified. But sometimes this is not the case, which helps explain the relative isolation of the discipline as noted earlier. This does not mean that economists should abandon their quantitative focus, but that they should always strive to ensure that their work contributes, directly or indirectly, to an understanding of a real world issue. This is consistent with Frey's observation that economics is part of those sciences which deal with actual problems of society, but that most economists attempt to imitate the sciences and that economics can be regarded as a branch of applied mathematics. It is also consistent with the view that that economics, particularly neoclassical economics, is the most mathematical of the social sciences . There are two further implications of this, both of which might be explored in further work. Firstly, modern universities have mainly organised their structure along the 'tree of knowledge' type model , according to which knowledge is split into branches , then into major disciplines and thence into subdisciplines and specialties. This has tended to result in economics being linked with the social sciences or business schools. Seldom are they linked with any of the sciences. This may lead to problems in being involved with administrators and academics who do not really understand the way economists work, nor their values. Secondly, the ethos economists perceive themselves as being part of tends to set the confines within which successful economists must work. If economists perceive themselves as mathematicians, then this sets out the theoretical style we see in many journals. If they see themselves as scientists then the emphasis moves to hypothesis testing . If, on the other hand, they see themselves as social scientists this will focus attention more on policy outcomes and a greater readiness to use the results of the other social science disciplines. The analysis has focused on papers submitted to the REF. Whilst this provides a large number of papers to analyse, they are biased towards the better, higher quality papers. Further analysis could extend the data set to cover a more random set of papers across a range of journals in each discipline in order to determine whether the writing characteristics of these high quality papers translate to the discipline as a whole. In addition, further work could estimate clusters based on other aspects of papers, such as length and the typical structure of a paper. Abstracts too differ between disciplines. In many journals, it is a single paragraph of between 100 and 250 words. But in some of the science disciplines the abstract too is divided into sections. Other possible characteristics include the numbers of equations, footnotes or endnotes, quotations and use of explanatory diagrams or flow charts, use of regression analysis and the ratio of the length of the concluding section to the rest of the paper. 15 Finally, we have divided the subject along into theory, econometric and other areas. This could have been taken further, dividing by subject subfields, such as health economics, labour economics and macroeconomics, and then studying these subfield's attribution by means of semantic similarity. Defining the subfields could be done, e.g., by looking at title papers in journals recognised as representing the different disciplines.

Different academic disciplines exhibit different styles, including styles in journal titles. Using data from the 2014 Research Excellence Framework (REF) in the UK we are able to identify the stylistic trends of different disciplines using 155,552 journal titles across all disciplines. Cluster analysis is then used to group the different disciplines together. The resulting identification fits the social sciences, the sciences and the arts and humanities reasonably well. Economics overall, fits best with philosophy, but the linkage is weak. When we divided economics into papers published in theory, econometrics and the remaining journals, the first two link with mathematics and computer science, particularly econometrics, and thence the sciences. The rest of economics then links with business and thence the social sciences.

INTRODUCTION One of the major avenues through which the advancement of Artificial Intelligence is affecting society and its people is by redistributing economic opportunities and earning prospects. Such redistribution can be equitable and inclusive, or biased in favor of certain groups that get to benefit from economic power concentrating in their hands [9]. Up until recently, many technologists believed that technological progress always "lifts all boats" and automatically leads to shared prosperity once the economy has gone through an adjustment period. This has left many to ignore the consequences of their inventions for economic inequality. However, since the 1980s, technological progress has been accompanied by a significant increase in inequality. For example, [4] surveys a significant body of economic literature that finds that automation in the US has led to a polarization of the labor market, whereby middle-income jobs that used to perform routine tasks were replaced by lower-income jobs, while those at the top of the income distribution experienced significant gains, leading to an increase in economic inequality. [14] document a similar phenomenon across the OECD, a club encompassing the richest countries of the world, over the past two decades: that median wages -the compensation of a typical worker -grew at a slower pace than overall productivity, and that the gap between the two has been increasing. In recent years it has -fortunately -become common to see demands that "AI should be human-centered, with transparency and accountability as paramount features" . However, those are often immediately -and unfortunately -followed by statements like "Overwhelmingly, AI will disrupt labor markets and the economy around the world, " which are taken as immutable facts. There seems to be a big difference in attitude towards AI violating notions of fairness, accountability and transparency, versus the prospect of AI disrupting the economic order, increasing inequality, and endangering the livelihood of millions of workers. The former is justifiably seen as unacceptable, while the latter is often taken as a given and treated as an unavoidable consequence of AI advancement. The difference in attitude seems in part driven by the ethical foundations used to evaluate advances in AI. Under traditional deontological foundations, developing an AI system that operates and performs all its actions while complying with all of society's traditional ethical rules is considered fair game, even if the system has the side effect of triggering general equilibrium effects in the economy that lead to the displacement of millions of jobs. By contrast, under more consequentialist foundations, it is more natural that such massive job displacement would be considered a violation of ethical norms. The distinction between the two approaches also has significant implications for how society is expected to respond to job disruption: if it is considered fair game for the developers and deployers of AI to disrupt labor markets and impose losses on millions of workers, then the burden of adjustment is on workers: it is their duty to face their losses, and they must continuously upskill in order to remain relevant in the labor market that is being reshaped by AI advancement. The implication for governments is that they are well-advised to strengthen social safety nets, expand training programs and prepare for benefits like a Universal Basic Income, all financed by taxpayers, not the perpetrators of the disruptions. In other words, it is society and its people, not the AI developers, who are expected to bear the burden of ensuring that AI advancement does not cut people off their sources of income. By contrast, if it is considered unethical for the developers and deployers of AI to disrupt labor markets, then the burden is on the AI industry to consider the impact of the technologies they are shaping on labor markets and employment opportunities. And it would be incumbent upon the AI industry to guide its decisions around the development of AI to avoid massive elimination of jobs and make the adjustments to a new technological and economic reality less burdensome and costly for workers and taxpayers. This paper builds on the recent economic literature discussing the impact of automation technology on labor demand and lays out a framework for systematically evaluating and predicting the impact of new AI-based technologies on labor income. We propose a step-by-step procedure that can be used by interested stakeholders in the AI industry and research community without prior knowledge of economics to evaluate which AI applications are likely to "lift all boats" and support shared prosperity, and which ones are likely to contribute to greater inequality. Evaluating this question throughout the AI development process is a critical step towards assuring that AI advancement is ethical and creates a fair and inclusive economic future for humanity [7]. We hope that such an evaluation will inform the choices of researchers, developers, innovators, investors and consumers in the field of AI, and that it will help to practically execute the idea of steering the direction of AI progress to benefit the workers instead of displacing them [3,10]. A decrease in labor demand can manifest itself as reduction in wages, employment, or both. Hence gauging the change in labor demand allows to understand the direction and magnitude of impact on workers' incomes and financial well-being. We should note that assessing AI's impact on labor demand does not necessarily capture the impacts that technology has on other aspects of well-being and overall job quality, like safety, level of physical strain, schedule predictability, freedom from surveillance, etc. However, job quality is frequently correlated with wages levels as higher labor demand also gives workers more bargaining power to ask for better job conditions. The rest of this paper is structured as follows: Section 2 discusses why the current trajectory of AI advancement risks exacerbating economic inequality; Section 3 presents ways to practically incorporate the consideration of economic inequality into the AI research and development process and proposes a high-level framework for directing AI progress for Shared Prosperity; Section 4 makes a case for AI researchers and developers to recognize the responsibility of ensuring that their creations support economic inclusion; Section 5 outlines the framework's limitations and open questions for future work; Section 6 concludes. --- ADVANCES IN AI AND EMPLOYMENT Traditional economic theory views capital and labor as the main factors of production in the economy. Capital includes machines, equipment, etc. The factor labor can be distinguished by geography or by levels of education, e.g. into lower-skilled, medium-skilled and higher-skilled labor. Depending on the application, it may also be useful to distinguish workers along specific occupations that may be differentially affected by an innovation. Technological change may boost the returns on some or all of these factor owners. But some types of technological change lead to starkly diverging impacts on different factor owners [9]. For example, they may increase the returns on capital but not labor, or of higher-skilled workers but not lower-skilled workers, or -to offer a very specific example -of AI engineers but not radiologists. By implication, they may shake up who in society has access to gainful employment opportunities and for whom those opportunities become harder to obtain as a consequence of falling demand for their skills. AI may change workers' access to economic opportunity even more starkly than the redistributions generated by past waves of technological progress. Let us provide a clear example of the type of redistributions that AI may generate: an e-commerce business that scales up might displace a large number of local mom-and-pop stores, resulting in a concentration of gains from selling goods to consumers. To provide another example, a shift towards consuming news through social media may disrupt traditional patterns of news consumption and may cause local newspapers to lose advertising revenues, leading to large job losses at local newspapers and, again, resulting in a concentration of gains in the hands of a few large companies. AI may also generate significant redistributions across countries [11]. In the following, let us discuss three common types of technological change that are frequently discussed and that may have large effects on workers' earnings opportunities: automation, skill-biased technological change, and human augmentation. Automation. Automation displaces human labor with machines. Frequently, higher-skilled workers are employed to design systems that automate away the jobs of lower-skilled workers. A cursory look might suggest that automating a job always has negative effects on workers while creating jobs always has positive effects on workers. But it is more complicated than that. Automation of human labor is not, in itself, undesirable. The history of progress since the Industrial Revolution is a story of relentless automation that has contributed to rising living standards. Moreover, there are many examples when automation allowed us to make work less dangerous and less physically taxing. If the automation of human tasks is accompanied by the creation of new tasks for humans, then the adverse effects on labor demand are offset [2]. Empirical evidence from the US suggests that while tasks automated and reinstated by technological change used to balance out during the four decades following WWII, the past three decades have seen task displacement significantly outpacing reinstatement. AI is expected to continue this trend and may even accelerate it [3]. Moreover, if the new tasks that AI advancement creates require a much higher level of skills or educational attainment compared to the tasks being displaced, these new tasks will bring little to no economic relief to the workers whose jobs get automated, even if the new jobs match or exceed the displaced jobs in volume. --- Skill-biased technological change. A type of technological change that disproportionately benefits those with comparatively high levels of educational attainment is referred to as skill-biased. Skillbiased technological change is poised to exacerbate society's structural inequalities, especially in countries with low economic mobility. Technological change does not have to be skill-biased. In fact, the first wave of the Industrial Revolution is generally viewed as having been biased in favor of unskilled workers, who suddenly had much greater earning opportunities. Likewise, there is nothing inherent about AI that makes its applications necessarily be skill-biased, or labor-saving. AI can be used to complement lower-skilled workers, making their labor valued more highly. It could also be used to expand the economic possibilities of people who previously had limited access to training and education, or for whom education was prohibitively costly. In other words, AI advancement could be economically inclusive, but making it so requires deliberate action. Human Augmentation. It is commonly suggested that developing AI systems that augment human workers instead of displacing them would be a good recipe to ensure that advances in AI benefit workers. Some AI firms have even started to use this language when describing their products in promotional materials and sales pitches. While "human-augmenting AI" as a generic goal seems more desirable than measuring the progress of AI by how well it automates away humans, it is important to note that human-or labor-augmenting AI can still result in the displacement of workers and reductions in wages, because labor-augmenting AI may still be labor-saving -it all depends on whether the firm will cut its prices and how much demand will respond to the price cut. For a simple example, suppose that a firm employs workers to produce a product that is sold to consumers. Suppose further that the firm develops and deploys labor-augmenting AI technology that allows each worker to increase their output per hour by 10%. With this technology, 10% fewer workers are needed to produce the same amount of output. If the firm cannot increase its production volume or the number of units it can sell per year, it may have to make redundant those 10% of its workers. By contrast, if the firm lowers the price of its product as a result of its reduced labor costs, and if consumers are willing to buy a lot more of the product at the lower price, then the firm might keep or even grow the size of its original workforce to satisfy the increase in the demand for its product. Whether this happens or not depends on how responsive consumer demand for the product is to the changes in its price . In a competitive market, the firm is indeed likely to lower its price when its costs decline. And even in a monopolistic market the firm might lower its price, for example if it aims to keep its markup constant. This stylized example illustrates that developing AI that is "augmenting humans" is not enough to ensure that employment and wages are not reduced. Because labor-augmenting technology allows firms to employ fewer workers to produce the same amount of goods, something needs to compensate for the reduction in labor demand that such technology will create in order to avoid job losses. These three examples illustrate that the effects of new technologies on labor markets are not always straightforward to assess, and that a systematic economic framework is needed to sort through the different effects. In the following section, we propose the outlines of such a framework, allowing researchers, companies, and policymakers to evaluate the overall impact of a new technology on labor markets. As we discuss the effects of technology on wages and inequality, we do not take a stance on what particular social welfare function to embrace, but our results on what redistributions a technological innovation will give rise to are relevant for any inequality-averse social welfare function. --- FRAMEWORK FOR EVALUATING AI'S IMPACT ON LABOR DEMAND This section describes a set of questions that represent a heuristic to systematically evaluate the overall, or general equilibrium, impact of an AI application on labor demand and, consequently, on employment and wages. We will use customer service chatbots and autonomous grocery delivery vehicles as illustrative examples when describing the logic of the framework. Customer service chatbots are applications frequently powered by Natural Language Processing machine learning models that simulate the behavior of human customer support agents. For the purposes of this discussion, we will refer to applications that respond textually or vocally to textual or vocal customer service requests as chatbots. In today's world, human customers frequently encounter such chatbots, for example when texting or calling customer support number, responding to a "How can I help you today?" automatic pop-up prompts on commercial websites and in other contexts. Autonomous delivery vehicles are driverless vehicles with a space to transport goods, but not people. At the time of writing, autonomous delivery is not widely commercially available, but a few startup companies are piloting them in a limited set of locations. 1 For the purposes of this discussion, we will examine autonomous grocery delivery, while noting that autonomous delivery vehicles could be used to deliver other types of goods as well. The framework that we propose in the following does not provide precise quantitative estimates of the magnitude of effects. Often mapping relative magnitudes and the directions of effects is in itself already very instructive and permits to understand the direction of the overall impact on labor demand. However, we hope that the proposed framework will inspire follow-up work on how to flesh out the described effects in more detail and at a more quantitative level [12]. Depending on the particular AI application, we also note that some steps may be skipped if the magnitude of the described effects is not deemed significant enough to merit a deep investigation. --- Direct Effects The direct effects reflect the workers who are directly hired or displaced because of the introduction of a new AI application in a company, while holding everything else fixed. We capture these direct effects via the following two questions: Which types of workers will be displaced by the introduction of a new AI system, and in which geographies are they located? It is important to capture the skill level and geographic location of those workers in order to understand if any of the compensatory effects below will be of relevance to them. For example, if an AI application displaces workers without a college degree in one set of locations and creates an equal number of jobs requiring advanced degrees in another location, the displaced workers will not be able to compete for the newly created jobs. In the chatbots example, the workers directly displaced are customer service associates located around the world, with major centers in the United States, India, and the Philippines. Those are predominantly formal sector jobs with predictable schedules and earnings. Skill requirements vary. The workers directly displaced by autonomous grocery delivery vehicles are delivery associates often employed as independent contractors, also referred to as gig workers. Their jobs are frequently precarious and lacking earnings predictability and benefits, but the work can be physically demanding and prone to injury. There usually are no degree requirements associated with delivery jobs. Notably, autonomous delivery vehicles would also "displace" unpaid work by households who shop for and bring their groceries home by themselves, without calling a delivery person. For what types of workers will new demand be created by the introduction of the AI application, and in which locations? The introduction of chatbots will create demand for software engineers customizing and maintaining the bots, chatbot platform sales and customer success people, product and marketing managers. Most of those jobs will require a college or advanced degree and will likely be geographically concentrated. They will also likely be fewer in number compared to the number of displaced customer 1 See, for example: https://venturebeat.com/2021/01/27/starship-raises-17-million-to-send-autonomous-delivery-robots-to-new-campuses/ and https://medium.com/nuro/california-dmv-grants-nuro-first-ever-av-deploymentpermit-ca424ebd2 service associates identified in step 1, because one chatbot development company can service many corporate clients. Introduction of autonomous grocery delivery vehicles will create jobs similar to the above at companies producing those vehicles, but in addition would create a need for workers assembling grocery orders and loading them onto the autonomous delivery vehicles . Those jobs will be similar in quality and skill requirements to the jobs of displaced delivery gig workers but might be safer and higher in volume if a significant number of households switch from shopping for and delivering their own groceries to using an autonomous delivery vehicle-powered service. --- Demand Effects Will the innovation increase demand for the company's products because it will lower its prices or increases the quality of its products? Will the increase in product demand translate into higher demand for workers? For what categories of workers and in which geographies? A company that cuts its costs by replacing some or all of its human customer support agents with an automated chatbot might lower the costs of its main product because of the lower expenses on customer support. Moreover, it might offer more customer support services and choose to provide it in contexts where it was not used before . This might not have much impact on how many human customer support agents it employs, but it may somewhat raise labor demand for all types of workers across the company. If grocery deliveries are made easier and less costly by the introduction of autonomous delivery, households will demand more grocery deliveries, increasing the revenue and associated employment of pickers at grocery stores. However, there are likely few effects on the overall volume of groceries purchased since grocery consumption is quite inelastic and quickly reaches a point of saturation. --- Vertical Effects How does additional product demand affect labor demand along the supply chain of the innovating company? In the chatbot example, the innovating company will need fewer workstations for customer support, less office space, and fewer office services, resulting in lower labor demand among those suppliers. However, if the company can expand the demand for its products because of lower prices or better marketing, then it will also raise the demand for its intermediate inputs, which may add labor demand to the economy. For the grocery delivery vehicles, we do not anticipate significant vertical effects since total demand for groceries will not change significantly. --- Horizontal Effects The horizontal effects capture how companies that produce goods and services that are substitutes or complements to the innovating company will be affected. Economists refer to goods or services as substitutes if they can easily be used for the same purpose, e.g. taxi rides or Uber rides. Goods or services are complements when an increase in the consumption of one makes it more desirable to consume the other, e.g. coffee and cream. How will competing companies producing substitute products/services be affected? Will they need to downsize their workforces? In which geographies and for which categories of workers? For the company introducing chatbots, lower prices may undercut its competitors and lead to cuts in their workforce. In the case of autonomous grocery delivery, services associated with getting to and from the grocery store might be impacted, resulting in a decrease in car sales and in the use of public transportation, impacting the associated employment. The customer service chatbots are not anticipated to lead to significant effects on complementary goods. If a significant number of households switch from buying their own groceries to using delivery services as a result of the introduction of autonomous delivery vehicles, it might affect the volumes of purchases of services that rely on the physical presence of customers at grocery stores, for example small vendors in front of the store. The effects are not likely to be major. --- Factor Reallocation How will wages adjust to reflect the new balance of labor demand and supply resulting from these changes? Our final question considers how equilibrium in the labor market is affected. The workers displaced by technological progress will, after an adjustment period, be redeployed in different companies or sectors, increasing overall output and wealth in the economy. However, when workers need to compete for new jobs, they will push down wages in their sector. And conversely, for workers who are in higher demand, wages will rise. This implies income redistributions that have the potential to exacerbate -or to mitigate -inequality, depending on the specifics of the situation. In our example of chatbots, the layoffs of customer service representatives will put downward pressure on the wages of unskilled workers, whereas the hiring of additional AI engineers will put upward pressure on their wages. Moreover, capital owners will benefit from the greater returns that the company earns. Ultimately, the described effects are likely to increase inequality. For the grocery delivery vehicles, the results will be similar, leading to a redistribution of income from unskilled workers to AI engineers, although the losses of unskilled workers are likely mitigated by the fact that many of the displaced drivers can be redeployed as pickers in grocery stores. It is important that the field of AI introduces a practice of systematically evaluating these effects to gauge the impact of their inventions on the income prospects of different groups of workers and to ensure a fair distribution of the gains from progress. We view it of particular importance for AI developers to pay attention not to disadvantage those groups of society who already are marginalized and who have limited access to retraining opportunities. --- RESPONSIBILITY OF THE AI COMMUNITY Progress in AI is unlikely to bring about an inclusive economic future if the direction of AI development is determined solely by market forces, which tend to favor efficiency but do not ensure that the gains from progress are distributed equitably. For example, in recent decades the labor share of income in the US has decreased, and the wages of non-college educated workers have stagnated and have not shared in the productivity gains generated by technological progress [5]. Market prices do not always reflect the true social costs and benefits generated by a business activity and therefore may provide misguided incentives. For example, in a free market, prices do not reflect the costs of environmental damage produced by a carbon-emitting enterprise. As a result, the costs are not borne by the polluter but instead by society as a whole. To make the polluting enterprise internalize the full costs of its operations, economic theory suggests taxing activities that produce externalities . But even in absence of carbon taxes, companies still have a moral responsibility to take on voluntary commitments to cut their emissions and minimize the un-internalized cost to society from their operations. Likewise, market prices do not reflect distributive concerns. When companies produce technologies that induce undesirable shifts in labor demand and, consequently, reduce the labor share of national income, there is a cost to society in the form of a more unequal income distribution, disrupted livelihoods, distressed families and communities and sometimes even "deaths of despair" [6]. These are costs that companies do not internalize, akin to environmental externalities. In absence of a regulatory framework to internalize the cost of job-displacing innovation, it is up to developers to behave responsibly in how they handle the redistributive power of AI. However, aside from voluntary action on the part of AI developers to consider the economic interests of workers in the AI development process, government can also take steps to ensure that the regulatory environment does not provide incentives for excessive automation and ever greater concentration of earning opportunities. For example, the incentives for innovators are affected by tax policies -current tax regimes that favor capital over labor [1] and policies that limit labor mobility [13] create strong incentives to develop AI applications that focus disproprotionately on labor-saving use cases. If these policies remain in place, AI advancement might bring about levels of automation well above what is socially optimal, to the disproportionate detriment of the economically vulnerable workers with limited access to retraining opportunities. The risk of excessive automation is increasingly recognized by scholars on the "future of work. " But unfortunately, the current discourse places too much of the burden of adjustment to the changing technological landscape on workers and governments, including developing country governments. Questions around the role and responsibility of the AI industry and the AI research community remain relatively neglected. --- LIMITATIONS AND CALL FOR FUTURE WORK The set of heuristics described in Section 3 is an early attempt at defining the questions that the AI industry and research community ought to be asking in order to evaluate the likely impact of their choices on inequality and availability of gainful employment opportunities for workers, especially the more economically vulnerable workers with lower levels of educational attainment and limited access to retraining opportunities. This attempt has both gaps and limitations. A non-exhaustive list of those is below. First, the economic framework we outlined is only the beginning of a research agenda to develop models of the impact of advances in AI on earning opportunities. Many different approaches are possible, and we hope that many will be considered, further developed, and refined by both the AI ethics and the economics community. Second, there is often a considerable amount of uncertainty when it comes to estimating the magnitude of the likely impacts of AI advancement on labor demand, especially so for the secondand third-round economic effects, such as effects that propagate internationally through trade, global supply chains or other mechanisms. AI developers and scientists working on basic research which can subsequently be used to enable many different kinds of applications face an additional layer of uncertainty. The heuristics described above could be improved by introducing weights that reflect the associated levels of uncertainty about each of the effects being considered. But even when the level of uncertainty seems overwhelmingly high, we want to caution against the AI industry and research community excusing themselves from the responsibility to think about the likely impact of their actions and choices on economic inclusion. To return to our analogy to environmental effects, just because climate science does not yet allow us to precisely attribute climate effects to the energy management and waste management choices does not mean that our choices should be reckless. The same holds for the effects of AI on inequality. It is incumbent upon members of the AI community to think about the economic future that they are so powerfully contributing to for societies around the world. Third, the framework presented above focused on labor demand and market wages as a measure of shared prosperity and ignored other aspects of job quality. On the one hand, this position is justified because for the majority of the human population, labor is the main asset that allows them to earn income. AI applications that reduce the demand for human labor undermine the value of what is the main and, for much of the world's population, the only asset which people have. On the other hand, "quality" of labor demand, not only its volume, matters to human well-being. If an AI-induced change to production processes replaces 1 million "good" jobs with 3.5 million jobs of equal pay but starkly lower job quality , it would be difficult to argue that this was overall beneficial to workers and advanced shared prosperity. Lastly, even though paid labor is currently the main source of income for the majority of humans, we do not want to rule out the possibility that far greater shared prosperity and human flourishing may be possible in the future of humanity, if all labor is replaced by sufficiently intelligent AI systems and machines [15,16]. This would entail a transformative change in our society and would bring about a stark set of novel challenges, including challenges in the way in which income is distributed [8]. We believe that in the short and medium run, while paid labor is the main source of income for the majority, our proposed framework to develop AI for shared prosperity offers one of the most promising ways of distributing the gains from technological progress. If we did approach a situation in which novel AI systems turn all human labor into a redundant technology, we believe -in the same spiritthat it would be incumbent upon the developers of these systems to develop new ways of sharing the economic gains generated by their inventions that do not rely on compensating labor in order to avoid mass immiseration and advance shared prosperity globally across human society. --- CONCLUSION Advances in AI have the potential to produce far-reaching impacts on workers and carry the risk of exacerbating long-standing inequalities within as well as between countries. AI developers have a moral responsibility to think about the economic impact of their creations, whom they might benefit and whom they might harm. We argue that the Responsible AI community needs to make an effort to develop frameworks and heuristics for thinking rigorously about these impacts and to steer AI development choices away from applications that deepen economic inequality and instead into directions that fulfill the potential of AI to generate shared prosperity.

Future advances in AI that automate away human labor may have stark implications for labor markets and inequality. This paper proposes a framework to analyze the effects of specific types of AI systems on the labor market, based on how much labor demand they will create versus displace, while taking into account that productivity gains also make society wealthier and thereby contribute to additional labor demand. This analysis enables ethically-minded companies creating or deploying AI systems as well as researchers and policymakers to take into account the effects of their actions on labor markets and inequality, and therefore to steer progress in AI in a direction that advances shared prosperity and an inclusive economic future for all of humanity.• Social and professional topics → Employment issues; Automation; Economic impact; Codes of ethics; Socio-technical systems.

Mexican-born persons account for the majority of deportations from the United States. In 2009, Mexican nationals accounted for 86% of the 613,003 persons apprehended in the United States, and ~72% of the 393,289 persons removed from the country . Deportees comprise a diverse population, which includes persons apprehended while crossing the border, during workplace raids, or during prosecution for other criminal charges, including drug offenses . In 2010, ~11% of Mexicans who were deported to Mexico from the United States were women . Mexican migrants and return migrants, including deportees with drug-related criminal records, might experience increased risk for HIV transmission from continuing drug use and related risk behaviors such as syringe sharing . Lifetime drug use is highest among Mexicans who have ever migrated to the United States , possibly because of increased exposure to drugs through social networks in border communities and in the United States . Specific subgroups of Mexican migrants in the United States, including injection drug users , experience a documented risk for HIV . Although women who migrate exhibit lower rates of drug use than men, women experience additional vulnerabilities to HIV , including sexual exploitation . Among persons who inject drugs in Tijuana, Mexico, male deportees had four times the risk of HIV infection; this elevated HIV risk was not found among women . In contrast, another study found that Mexican women who injected drugs also reported frequently trading sex for money or drugs . Significant gaps in the literature prevent a thorough understanding of how involuntary migration, including deportation, can shape HIV risk among vulnerable groups, including drug-using women. Little is known about the relationships between deportation, social and economic vulnerability, drug abuse , and engagement in HIV risk behaviors among women who inject drugs. Thus, we conducted an exploratory qualitative study of deportation experiences among women who inject drugs in Tijuana. Many Mexican deportees are returned or travel to border cities such as Tijuana, Baja California, which is adjacent to San Diego, California. Together, both cities form the busiest land border crossing in the world. In 2010, ~40% of Mexicans deported from the United States arrived to ports of entry in Baja California, with Tijuana receiving the largest share of deportations, at >126,000 persons . Tijuana is the largest and fastest-growing city on Mexico's northern border with the United States, with an estimated 1.4 million people in 2005. The city lies on drugtrafficking routes for heroin, cocaine, and metham-phetamine destined for the United States . Tijuana supports a large population of injection and noninjection drug users . Increasing HIV prevalence, particularly among high-risk populations of IDUs, female sex workers , migrants, and deportees, is a significant concern for health officials in Tijuana . HIV among substance-using migrants is also concerning because of the potential for "bridging" otherwise unconnected populations via their sexual and drug use practices, which may occur in diverse geographic regions. Among women in a cohort of IDUs in Tijuana, the adjusted HIV prevalence was recently estimated at 3.1%; notably, in the larger cohort, the majority of IDUs diagnosed as HIV positive were unaware that they were HIV positive . Persons who are unaware of their HIV serostatus are at high risk of transmitting HIV , implying that the sexual and substance-using networks of undiagnosed binational populations, such as deportees, might experience heightened risk for infection. Given this setting and the lack of research on deported women who inject drugs, we drew on Rhodes' risk environment framework ) for understanding the social and structural production of HIV risk. Within this framework it is posited that drug users' risk for HIV is shaped by exogenous physical, social, economic, and policy risk factors at the macro-, meso-, and micro levels , in addition to endogenous individual characteristics and risk behaviors such as unprotected sex and sharing of syringes ). Using this framework, in our exploratory study we aimed to describe the experiences and risk environments within which deported, drug-using women in Tijuana lived throughout their migration and drug-use trajectories. --- Methods --- Study Population & Recruitment We recruited women from a prospective study examining HIV risk behaviors among IDUs living in Tijuana, Mexico. As previously described , in April, 2006, outreach workers used respondent-driven sampling to recruit participants >18 years old with evidence of recent injection drug use . For our exploratory study sample, we first generated a list of IDUs in the prospective study who reported U.S. deportation. Of the 1,056 prospective study participants, 15% were women, 36 of whom reported U.S. deportation. In October, 2008, outreach workers invited all women who were not lost during the follow-up period to participate in our exploratory study. After obtaining voluntary and informed consent, we enrolled 12 women in our exploratory study. We reimbursed women with US $20 for their time for undergoing qualitative interviews. The Human Research Protections Program of the University of California-San Diego and the Ethics Board of the Tijuana General Hospital approved the study protocols. --- Data Collection During October and November 2008, trained bilingual interviewers conducted semistructured qualitative interviews in private interview rooms in Tijuana's zona norte , an area of the city where drug use is widespread. Based on our theoretical framework, we developed a semistructured interview guide to query women on their migration, deportation, drug, and sexual experiences within various risk environments in the United States and Mexico. Sample questions included "Tell me about your drug use in the United States," and "Describe what happened the most recent time you were deported." Because many participants experienced multiple deportations, we asked women to focus on their most recent deportation, a strategy that the study team developed to increase the participants' ability to recall the conditions surrounding this event. Interviews ranged in duration from 60 to 120 minutes, and were conducted in English, Spanish, or both languages, according to women's preferences. We digitally recorded and transcribed interviews but did not translate the transcriptions because we wanted to preserve participants' language, which was often bilingual and contained terminology specific to the economic and drug subculture of Tijuana. --- Data Analysis We analyzed data for concepts by identifying major categories in participants' experiences . We created an initial coding scheme based on the major themes and concepts in the interview guide. Next, we read several cross-sections of interviews, revising the coding scheme to include emergent concepts and categories. We independently applied these codes to five interviews and discussed and resolved all discrepancies between coders by refining code definitions. We continued applying codes and memos to the transcripts. When new codes emerged, we updated the coding scheme and reread all transcripts according to the new structure. We employed ATLAS.ti, a qualitative data analysis computer software program, to attach codes and memos to transcripts . Throughout the analysis, we discussed emergent themes and compiled quotations illustrative of key concepts. Quotes presented in this article were translated from Spanish to English as necessary by the first and last authors, and follow the original wording of the participants' statements. --- Results We identified several important themes relating to the various social and physical risk environments surrounding women's drug use before, during, and following U.S. deportation. We found that most women's social risk environments in the United States contributed to their initial drug use , escalating drug abuse, and injection drug use. Drug-abuse experiences in the United States were intertwined with women's criminal histories and eventual deportation to Mexico. Following U.S. deportation, women reported heightened social and physical risk environments in Tijuana, including lack of social networks, greater financial and physical insecurity, emotional distress, increasing drug dependence , and lack of access to drug treatment and other health services. In the following sections, we review these themes in greater depth and provide illustrative examples of women's experiences. --- Migration Experiences The median age and educational attainment in our sample were 37.5 years [interquartile range : 32-41] and 9 years , respectively. Women were born in four states in Mexico, including Baja California , Jalisco , Chihuahua , and Guerrero . Women arrived in the United States as children , either migrating with parents or being sent by parents to live with U.S. relatives. Most women completed primary and secondary education in the United States and spoke English proficiently or fluently at the time of the interview . Women lived in cities in Southern California and suburban or rural areas , and rural areas in Northern and Central California . Participants reported a median of 2.5 lifetime deportations , with the most recent deportation occurring a median of 5 years before the interview . --- Drug Use Initiation in the United States Women reported first consuming alcohol and marijuana during adolescence , when most were already living in the United States. Three women who traveled to the United States during later adolescence recalled seeing family members in Mexico use alcohol or drugs, but initiated their use of illicit substances after migrating to the United States. Several women reported trying cocaine or heroin in high school in the United States, and began injecting these drugs during late adolescence. Women's social risk environments in the United States facilitated their initial drug use. Women explained how family members friends, and boyfriends in the United States introduced them to drugs . Women also attributed early experimentation with drugs to an underlying curiosity because of the high prevalence of drug use in their communities and social networks, as described by a woman who migrated to California as a young child: When I was sixteen, I wanted to see how it felt. My whole family [used], well not my whole family, but my brothers and sisters. … I used to see them get high or whatever, and I wanted to know how it felt, I wanted to join the club. I thought it was cool. … Anything that they had I used. … PCP, acid, rock, heroin, crystal, coke. Some women recalled experiencing pressure to use drugs by gang members , sex partners , and acquaintances. One woman who used cocaine and methamphetamine in the United States explained, "My husband's friends introduced me [to drugs]; they were very possessive people who wanted me to use, even though I didn't want to." Although women attributed much of their initial drug use to social relationships, two women also described practical or employment-related reasons for starting to use drugs. One woman described how methamphetamine could help her maintain sufficient energy for two factory shifts: "I felt enthusiasm, strength. … I was no longer sleepy [or] in pain, and I worked well." --- Drug Abuse in the United States For nearly all women in our sample, occasional drug use during adolescence in the United States transitioned into regular, daily consumption of drugs, eventually contributing to physical dependence on heroin and/or crack cocaine . Experiences with methamphetamine in the United States were limited, and alcohol use was likely underreported because it was not a focus of our interviews. Some women discussed periods of polydrug use in the United States, particularly in the context of social relationships and parties. Women described how friends, spouses, and sex partners enabled their continued drug abuse. For example, one woman attributed her increasing dependence on heroin to a close friend who regularly supplied her with the drug: She would give me a bit more, and more, and several months passed by, and I felt shivers. My bones in my arms ached, and she would tell me, "Maybe you are addicted to that stuff [heroin]," because I was using more and more. Despite increasing drug abuse and dependence, five women described using drugs in the United States within an environment of relative stability and social support. When asked about daily life in the United States before deportation, one woman explained, It was good because I was working to pay my bills, help my mom. … I wasn't ripping nobody off. I was doing good, living a normal life. Not so normal because I was using drugs, you know what I mean? But I was stable in my life. While living in the United States, family members provided economic and childcare support despite women's escalating social and legal problems resulting from drug abuse: "My husband supported me, gave me money. … My family saw I was crazy [from drugs], but they helped me with the rent, helped me with my daughters and everything." Despite the economic stability and social support in the United States, four women described negative consequences of their increasing use of heroin, crack cocaine, and methamphetamine. Losing control over one's life was a theme reported in many interviews; for example, one participant lost her job when caught with illicit drugs at work, and another described selling and trading sex for drugs at night in a dangerous neighborhood. Women also described social problems resulting from their drug abuse. For example, one woman hid her drug use from friends and family in the United States out of fear of being ostracized or punished, and two women described having to "live on the streets" after being kicked out of families' homes. --- Injection Drug Use in the United States Nearly all women reported injection drug use in the United States . Eight women injected regularly for many years, and two reported injecting in the United States occasionally or "just to see what it was like." Similar to the social risk environment influencing their initial drug use and experimentation, women reported that friends , boyfriends , and family members in the United States taught them how to inject drugs. Seven women explained that they regularly shared injection equipment with friends and boyfriends because of physical dependence and withdrawal, apathy or not caring about the consequences, and not knowing where or how to access sterile injection equipment. One woman who started injecting heroin in the United States when she was 16 years old explained that she knew sharing syringes could be dangerous but did so anyway, "out of necessity, because I didn't care, you know? … Because it's a pretty ugly feeling when you're sick." Women bought syringes in the United States from drug dealers or at liquor stores, and one respondent reported obtaining syringes from her mother, a nurse: I know it broke her heart, supplying me with [syringes]. "If you're going to be using that shit, at least promise me that you're going to use clean needles. And if you run out, please don't use nobody else's needle." It was something that I was raised with. I mean, raised, no, that's not the word. I got used to, you know, clean needles. Two women reported rarely sharing syringes because they typically injected alone, hiding their drug use from their social networks or avoiding other IDUs for safety concerns. --- Deportation and Criminal Justice System Experiences For all participants in our sample, women's most recent deportations were preceded by lengthy criminal histories that often included "serving hard time" in U.S. federal and state correctional facilities. Several women served lengthy prison sentences . Reasons for women's most recent deportations included being arrested, violating parole, and having a record of unauthorized entries into the United States. Criminal charges ranged from minor drug possession charges to armed robbery, auto theft, and assault. Some women experienced feelings of defeat and hopelessness with regard to U.S. criminal justice and immigration systems, particularly when contesting their immigration charges became more difficult because of past drug offences. One woman explained that she was prohibited from ever returning to the United States because of her accumulated drug and immigration charges: Immigration is hard on drug trafficking, which is why I got so much [prison] Several women recalled being informed by immigration officials that they were prohibited from reentering the United States for periods of 10 to 20 years, or that they were "banned for life." Another respondent remarked that she was assumed to be Mexican and perhaps was singled out for being Mexican, despite having little connection to or knowledge of the country: [Immigration] started looking for Mexicans. … They interviewed me, and they told me that because of my drug history, I couldn't be in the United States. I had to get deported, you know? They asked me, "Are you afraid that somebody's going to hurt you in your country or something?" And I told them no, because I consider the [United States] as my country, you know? Not anywhere else. … I don't know anybody [in Mexico]. Women described "waiving" their rights and agreeing to voluntarily return to Mexico in exchange for an expedited removal process without additional prison time. Women described different levels of drug accessibility in prison and detention facilities, with 9 women using drugs in prison and only 1 woman using drugs in an immigration detention center. Women described obtaining drugs and injection equipment from visitors or other inmates while in prison, and 1 woman sold drugs. Many women described intense physical withdrawal symptoms and "getting clean" in prison and detention facilities, followed by relapse or bingeing upon release. One woman who served consecutive federal sentences for drug charges and unauthorized reentry into the United States, an immigration offense, noted, "Two or three years before I got out [of prison] I stopped [using drugs] completely. I wanted to get clean, you know, but I didn't last very long after I got out." --- Postdeportation Milieu Following their most recent deportations, women were released into Tijuana or migrated to Tijuana from other Mexican ports of entry in search of economic opportunities, drugs or the local party scene, or existing social networks in the city. Women experienced heightened physical risk environments immediately following deportation, including difficulty communicating with family and friends in the United States and Mexico, finding shelter and safety, and securing employment. Some women shared the cost of temporary housing with other deportees in the zona norte. Women lacking social networks in Tijuana described feelings of fear, isolation, and disorientation with the neighborhoods or street culture of Tijuana and other border cities into which they were released. Some women were deported and released into Mexico very late in the evening and experienced increased physical vulnerability immediately following deportation, including lacking money and shelter . Although some women shared the cost of hotel rooms with other deported women, one woman described immediately engaging in sex work because she needed money for food and shelter, and her deportee acquaintances "showed [her] the ropes." Another woman met a man in the zona norte who initially offered help but actually abducted her. For several months, the man injected her with heroin, which she had never injected regularly, and forced her to sell sex. She eventually escaped, but remained dependent on heroin and continued sex work. At the time of the interview, half of our sample reported that sex work was their primary source of income ; others were underemployed in the informal sector. Four women reported an average monthly income of less than US $250, indicating high levels of financial need. Another immediate postdeportation concern was reentering the United States. Several women expressed an interest in returning to the United States for family reunification and improved financial stability , but women who were "banned" from returning to the United States worried about additional criminal charges and incarceration. None of the women had attempted to return to the United States following their most recent deportation because they lacked the resources to hire a coyote , or were afraid. One woman explained, "A lot of people die when they cross into the United States and it makes me scared, and I no longer want to return." --- Postdeportation Drug Abuse Locating drugs was a major concern for many women immediately following deportation . Two women described meeting men who were willing to provide them with shelter or drugs, at least temporarily, in exchange for sex. One woman explained how this quickly led her back into a life of "working the streets" to support her drug use: When I was sent here, I was unfamiliar with Tijuana. Well, I knew Tijuana, but not this drug scene. So I arrived with a little money; it ran out and I had to convince a man here and it was simple [to trade sex] to get high. … And I fell lower and lower. Many women relapsed into old drug "habits" ; some immediately began using the same drug they had used in the United States, and others experimented with new or rarely used drugs. Methamphetamine, a new drug for many women postdeportation, was injected in combination with heroin or by itself . Two women tried to abstain from drugs because they had been "clean" in prison or detention before their deportation. They reported that they were unsuccessful in abstaining from drugs because of the ubiquitous presence of drugs in the zona norte. Women described increasing drug dependence in Tijuana following deportation, which several women attributed to the lack of control over their lives. --- Postdeportation Injection Drug Use Several important themes emerged regarding injection drug use in women's postdeportation risk environments. First, the 4 women who rarely or never injected drugs in the United States began injecting regularly following deportation. Women described beginning to inject because of drug dependence, lack of self-control, and the influence of their social networks and neighborhoods, in which drug abuse was pervasive. Second, women reported frequently sharing syringes in Tijuana, particularly with people from whom they sought help injecting, including boyfriends and "hit doctors" . Although many women described efforts to avoid sharing injection equipment because of fear of acquiring HIV/AIDS or developing abscesses , sharing injection equipment was common with sex partners. Two women described their need to be high on drugs "when it's business," or to make sex with clients easier. A final emergent theme relating to injection was the fear of being apprehended or harassed by police for carrying syringes, "being a known user," or lacking proof of identification . Even though carrying syringes and having track marks are legal in Mexico , some women reported never carrying syringes because they were afraid of or had experienced police harassment or detention. Instead, women reported injecting in the same locations where they bought drugs, hoping that sterile syringes would be available. One woman described how she tried, unsuccessfully, to hide her track marks from the police, which she showed the interviewer: I [used to] inject my arms and everywhere. My legs [were] all fucked up, you know? See all those black spots? I used to burn myself so there wouldn't be [anything], no tracks or whatever, and the cops wouldn't see it. But I'm still "track looking" because they've got holes, see? I burned myself. They hurt a little bit, but I didn't care at the time. Respondents bought sterile syringes from pharmacies and liquor stores, obtained free syringes from a local organization, or borrowed used syringes from other drug users in picaderos. --- Postdeportation Drug Treatment Seven women in our sample had experience with drug treatment programs in the United States or Mexico. At the time of the interview, nearly all women wanted to "get clean," and several described strategies to reduce their drug consumption or frequency of injecting. However, only 5 women had attended drug treatment programs in Tijuana, which they described as unsanitary, strict, involuntary, and overly religious. Four women felt fearful or skeptical of existing programs in Mexico because of their poor reputations. Financial access was a major barrier to drug treatment. Six women who were interested in starting drug treatment did not know how to afford entrance fees. Four women had knowledge of a methadone clinic in the zona norte that charged a prohibitive daily fee; of these, 2 women ultimately found long-term methadone treatment to be unaffordable. One woman described purchasing methadone once a week or less, depending on how much money she could save. --- Postdeportation Emotions Nine women described feeling lonely and sad following their most recent deportation, often because they were separated from children and other family members in the United States and elsewhere in Mexico. Several women discussed "losing everything," as exemplified by a woman who was deported 10 years prior to the interview and was unable to return to the United States: I felt alone. I felt abandoned. …. Before my deportation, I did use drugs but not like right now. …. I had my house, my children. My children were young. I still have my children but they are older now, married. … Well, the deportation affected me a lot since in the past, I did not use drugs on the street. I feel humiliated. Women also believed that they had been rejected by family members in the United States, and reported feeling ashamed about their current lives and struggles with drug dependence: I mean, right now, I haven't seen [my family] in about a year already, because my habit is getting too much, you know? And I don't want them to see me like this because I don't want them to lose the little respect they still have for me, you know? And that's why I don't want to bring them here. They want to come and live with me, but no, no, no, no. I tell them, "No, I'm not ready for you guys, yet." Women described wanting to discontinue sex work, although supporting themselves was perceived to be much more difficult in Tijuana than in the United States. One participant characterized life in Mexico as a daily struggle to earn enough money for food and drugs. Two respondents mentioned their desire for a stable routine, although they did not believe they could achieve such stability without help. One woman described her hope "to find a good man" who could help her "get a job, get clean, and live in a house." --- Discussion Our exploratory study, which was guided by a risk-environment framework ), provides an important characterization of the range of migration and drug use experiences, physical and social vulnerabilities, and HIV risk factors among women who are deported from the United States to Mexico. We identified several key themes relating to the social and physical environments that influence women's drug abuse and engagement in other risk behaviors for HIV transmission . Overall, women described that social networks influenced their initial drug use in the United States, which was followed by escalating drug abuse, incarceration, and eventual deportation to Mexico. Key themes supported the notion of heightened risk environments following deportation, including greater financial and physical insecurity, increasing drug dependence, emotional distress, and lack of access to drug treatment and other physical and mental health services. Overall, we found that deported drug-using women in Tijuana reported extensive experiences with U.S. migration, drug use, incarceration, immigration detention, and deportation. Previous studies have found drug use among Mexican American adolescents and young adults to be associated with the "Americanization" process, particularly among young women with high levels of acculturation . Most women in our study immigrated involuntarily to the United States as minors and thus had little or no exposure to drugs in Mexico. Mimicking drug-initiation trends among adolescents born in the United States, women began experimenting with drugs with friends and family members during their middle-and high school years. In our sample, all women became heavy drug users by early adulthood while living in the United States. In addition to exposure to U.S. culture, the social and physical risk environments of women's lives in the United States facilitated their drug initiation and continued use. Our study findings support arguments by Valdez and coauthors that culturally relevant theories of social networks and drug use must consider the influence of family members in addition to that of peers and sexual partners within social networks . Our study supports previous findings that characteristics such as familismo and collectivismo can influence behaviors within druginvolved Mexican American families . Women in our study also described supplementing their family's financial support by engaging in criminal activities in the United States, including drug selling, theft, and occasional sex work. Women's U.S. social networks might have encouraged or accepted these criminal activities . However, exchanging sex for drugs could simultaneously increase autonomy from families and traditional gender roles , suggesting complex relationships between social and economic influences on women's drug abuse in the United States. Women's legal problems relating to drug abuse were often described as the primary causes of their deportations. Women developed extensive criminal records in the United States, often starting in adolescence and resulting in lengthy periods of incarceration and detention. Drug use continued during incarceration , but injection behaviors were rare, and some women described "getting clean" during these periods of their lives. Following deportation, women described struggling with drug dependence and quickly relapsing into drug abuse and other risky behaviors. These findings are consistent with research on the economic, social, and drug-related challenges that incarcerated drug-using women face when released into communities where they lack social support and financial opportunities . The experiences of women in our sample support arguments that the U.S. criminal justice system needs evidence-based drug treatment services for minority women , including drug treatment and other health and social programs for parolees reentering communities . Our findings suggest that women could have benefitted from sustained drug treatment services in U.S. prisons/detention centers, and following their deportations; however, none of our respondents had accessed drug treatment or other reentry services in the United States or Mexico. The women's stories highlight the need for programs to recognize the multiple risk environments that deportees experience, and identify ways to integrate or reintegrate deported women into Mexican communities. It is unsurprising that women described postdeportation life in Tijuana as being more difficult, uncertain, and dangerous than in the United States, because many of the respondents were unfamiliar with Tijuana or Mexico before deportation. Despite the lack of programs for return migrants and deportees globally In addition to drug treatment, we learned from the women that they could benefit from services to help them cope with their emotional distress and physical and financial insecurity following deportation. Participants in a qualitative study of traumatized refugees in Switzerland held positive views of medical treatment and expected that they would benefit from services to help them cope with distress . Women in our sample also lacked formal economic opportunities and social connections following deportation, highlighting a need for social reentry programs for return migrants. For example, Mexican migrant and deportee programs could systematically provide deported women with assistance in locating affordable housing and obtaining identification cards necessary for securing formal employment. Access to legal employment opportunities might be particularly useful for deported drug-using women because migrants in Tijuana and other regions of Mexico have reported substantial migration-and drug-related stigma and exclusion from communities into which they enter . Finally, programs for deported drug-using women could also include Spanish-language education, assistance navigating financial and legal institutions in Mexico, and the development of positive social networks . Our data illustrate multiple, overlapping social and structural vulnerabilities to HIV and other physical and mental health concerns experienced by deported women. Overall, we found that women in our sample experienced heightened vulnerability to HIV in their postdeportation risk environments because of their disadvantaged status as parolees, involuntary return migrants , IDUs with significant drug dependence, and women who trade sex for money or drugs. An additional source of HIV vulnerability might result from the release of deportees in an international border region, which Rhodes and coauthors described as including increased population movement, drug trade and transport, social disruption created by language and economic difficulties, exploitation, fear of authority, and mixing of economically and socially disadvantaged populations . For women, being released late in the evening in a dynamic border region might result in physical and sexual violence that further contributes to their vulnerable status and marginalization over the long term. Furthermore, although women in our sample did not specifically remark on the recent drug violence in Mexico, more than 34,000 people have died as a result of Mexico's drug conflict since December, 2006 , suggesting important impacts of violence on Mexican communities. The social and structural factors of the postdeportation risk environment warrant additional research so that public health resources can be efficiently directed toward services that serve vulnerable deportees and other return migrants. Our exploratory study findings must be interpreted in light of the following limitations. First, our data on migration, deportation, and drug use were based on self-report, and could have been affected by recall bias. We asked participants to focus on their most recent deportation because many women had lengthy migration histories. Second, we selected participants based on self-reported deportation, and cannot confirm actual U.S. removal events. Third, the shame and fear regarding certain experiences could cause social desirability bias, or a reluctance to report certain events. Fourth, we did not specifically inquire about alcohol use, and believe that additional research is needed. Finally, the small number of women in the prospective study who reported deportation and could be located limited our sample size to 12 women. Although theoretically data saturation can be achieved within 12 interviews , the objective of our exploratory study was to describe the range of experiences of a highly vulnerable and understudied population. We believe that the 12 women's stories provide a powerful basis for future investigation of risk environments within migration and deportation trajectories. Although we give examples of potential services for return migrants in Mexico-U.S. border cities, additional research with a larger sample is needed to develop concrete policy and programmatic recommendations. --- Conclusion We found that deported, drug-using women in Tijuana, Mexico have complex migration and drug use trajectories. Special efforts are needed to help these women avoid or cope with the negative health and social consequences of involuntary return migration. A multitude of factors contribute to the unique postdeportation risk environments of deported drug-using women, including their injection drug use and drug dependence, release from incarceration, lack of social and economic resources, unfamiliarity with the postdeportation society, stigma and policing practices, gender, and engagement in sex trade. Our study findings indicate that binational coordination is needed to help deported women resettle in Mexico, especially when they emigrated involuntarily , are unfamiliar with the country to which they are returning, and are struggling with drug dependence and relapse. Women could benefit from assistance in accessing quality drug treatment services, locating safe and affordable housing, reestablishing citizenship, and securing employment. Additional research is urgently needed to help understand how such programs could help reduce vulnerabilities to HIV while improving the general health of deported migrants and their sexual and drug using partners in the U.S.-Mexico border region, and the communities in which they reside.

Deportation from the United States for drug offenses is common, yet the consequences of deportation for women drug users are poorly documented. In 2008, in Tijuana, Mexico, we conducted an exploratory qualitative study of migration, deportation, and drug abuse by interviewing 12 Mexican injection-drug-using women reporting U.S. deportation. Women reported heavy drug use before and after deportation, but greater financial instability and physical danger following deportation than when in the United States. We identified an unmet need for health and social services among deported drug-using women, including HIV prevention, drug treatment, physical and mental health services, and vocational training. Binational coordination is needed to help deported women resettle in Mexico.

Introduction Groups associated with the extreme right have maintained an online presence for some time [1 ,2], where dedicated websites have been employed for the purposes of content dissemination and member recruitment. Recent years have seen increased activity by these groups in social media websites, given the potential to access a far wider audience than was previously possible. In this paper, we present exploratory analysis of the activity of a selection of these groups on Twitter, where the focus is upon groups of a fascist, racist, supremacist, extreme nationalist or neo-Nazi nature, or some combination of these. Twitter's features enable extreme right groups to disseminate hate content with relative ease, while also facilitating the formation of communities of users around variants of extreme right ideology. Message posts by such groups, to which access is usually unrestricted, are often used to redirect users to content hosted on external websites, for example, dedicated websites managed by particular groups, or content sharing websites such as YouTube. These posts also assist in the mobilization of participants for subsequent interactions such as offline demonstrations or other offline activities. Our investigation has found that communities of users associated with extreme right groups are indeed present on Twitter. The primary objective of this analysis is the detection of such communities within individual countries, using network representations of user interactions. In this context, an interaction is defined as one user "mentioning" another within a tweet, where reciprocal mentions between users can be considered as a dialogue [3], potentially indicating the presence of a stronger relationship. For the purpose of this exploratory work, we have retrieved Twitter data, including profile information and posted tweets, for a selection of eight countries, using a sampling method that requires the identification of sets of core users who are considered to be highly relevant. A network representation is extracted for each of these country data sets, from which communities of related users are detected. Having ranked these communities based on their stability, we can describe them using the hashtags contained within tweets posted by the member users. Each such description can then be used in conjunction with manual analysis of the user profiles, tweets and external websites to provide an interpretation of the underlying community ideology. Our secondary objective is the identification of international relationships between certain groups that transcend geopolitical boundaries. This involves the analysis of two network representations of the interactions between the core users from the eight country sets, using the follower relationships and mentions dialogues. It appears that a certain amount of international awareness exists between users based on the follower relationship, while mentions interactions indicate stronger relationships where linguistic and geographical proximity are highly influential. In Section 2, we provide a description of related work based on the online activities of extremist groups. The generation of the Twitter data sets, using profile data and tweets for users from each country of interest, is then discussed in Section 3. Next, in Section 4, we describe the detection of local extreme right communities within individual countries, including the methodology used for network derivation, community detection, stability ranking and description generation. In this exploratory analysis, we focus on two case studies, using the USA and Germany networks, where we offer an interpretation of a selection of these communities. Due to the sensitivity of the subject matter, and in the interests of privacy, individual users are not identified; instead, we restrict discussion to known extreme right groups and their affiliates. Analysis of the international relationships between extreme right groups from the eight countries is presented in Section 5. Finally, the overall conclusions are discussed in Section 6, and some suggestions for future work are made. --- Related Work The online activities of different varieties of extremist groups including those associated with the extreme right have been the subject of a number of studies. Burris et al. [2] used social network analysis to study a network based on the links between a selection of white supremacist websites. They found this network to be relatively decentralized with multiple centres of influence, while also appearing to be mostly undivided along doctrinal lines. Similar decentralization and multiple communities were found by Chau and Xu [4] in their study of networks built from users contributing to hate group and racist blogs. They also found that some of these groups exhibited transnational characteristics. The potential for online radicalization through exposure to jihadi video content on YouTube was investigated by Conway and McInerney [5]. A continuation of this work by Bermingham et al. [6] included network analysis of the associated YouTube users, where it was suggested that a potentially increased online leadership role may be attributed to users claiming to be women, according to centrality, network density and average speed of communication. Sureka et al. [7] also studied the activity of extremist users within YouTube, investigating content properties along with hidden network communities. In a similar approach to that of Burris et al. , Tateo [8] analyzed groups associated with the Italian extreme right, using networks based on links between group websites. A combination of social network analysis and content analysis methods were used. Caiani and Wagemann [9] studied similar Italian groups along with those from the German extreme right. They found the German network to be structurally centralized to a greater extent than that of the Italian groups. A related study by Caiani and Parenti [10] concluded that Spanish extreme right groups do not exploit the Internet for the purposes of communication and mobilization to the same degree as other countries. The contents of websites belonging to central nodes within Russian extreme right networks were analyzed by Zuev [11]. In their review of the conservative movement in the USA, Blee and Creasap [12] discuss the engagement in online activity as part of an overall mobilization strategy by the more extremist groups within it. McNamee et al. [13] analyzed the messages from a number of hate group websites, characterizing them using four themes; education, participation, invocation and indictment. Simi [14] suggested that the apparent decentralization of white supremacist groups according to, among other aspects, their online activity, should not be confused with disorganization or irrelevance. As the majority of this related work involved the study of dedicated websites managed by extreme right groups, we felt that an analysis of their activity in social media would complement this by providing additional insight into the overall online presence of these groups. --- Data Retrieval with User Curation Twitter data was collected to facilitate the analysis of contemporary activity by extreme right groups. As the hypothesis was that extreme right communities within social media would tend to be relatively smaller than mainstream communities, a form of snowball sampling was applied rather than using a random sampling approach. One of the objectives of this analysis was the detection of extreme right communities within individual countries. Given this, we used the curation method as proposed by Greene et al. [15] to generate core sets of relevant user accounts, one set per country of interest. This involved the initial identification of country seed user sets, where the user accounts within these seed sets were chosen based on a number of criteria . Prior knowledge of extreme right groups informed the selection of countries of interest. Following the bootstrapping phase, the core sets were expanded over a number of iterations based on manual analysis of the users proposed in the recommendation phases . A selection of recommended users were added to the core sets using the following criteria:  Relevant profiles; for example, those containing references to known groups or employing extreme right symbols.  Recent tweet activity .  Similar YouTube user account profiles.  Follower relationships with known relevant users.  Users with self-curated lists containing relevant users.  Extreme right media accounts; for example, record labels, online music stores, radio stations, concert organisers etc. A number of recommended users were also ignored, such as inactive users, or those that were not deemed to be related to the extreme right. These included traditional conservative users, non-conformists/anti-establishment users considered to be left-wing, and conspiracy theorists. As the focus of the analysis was on extreme right groups, higher-profile politicians or political parties were ignored for the most part, with a minor number of these users included where it was felt that there was a close association with relevant users. Although expansion of the core sets using this criteria required a certain amount of subjectivity, it is felt that this was acceptable given that the requirement was to identify a relatively small number of relevant user accounts, and relevance is often immediately evident from manual profile analysis. A major obstacle was the language barrier, as the use of online translation tools were not always helpful in the case of ambiguous profiles. In such cases where the relevance of a user profile was inconclusive, the user was ignored. Although this leads to a disparity in core set size, for example, Spain and Germany have a relatively larger number of users, core set size does not necessarily influence the size of the communities detected within the generated networks, as demonstrated later in the case study sections. Profile data including followers, friends, tweets and list memberships were retrieved for each of the core users, as limited by the current Twitter API restrictions. The bootstrap process also retrieves a selection of profile data of those users having a follower relationship with a core user. In order to address possible incompleteness, additional profile data was also retrieved for any users having a reciprocal follower relationship with more than one core user, to produce the final data sets for analysis. As the Twitter follower relationship tends to exhibit lower reciprocity than other social networking sites [16], the understanding was that this action would be largely isolated to those users having a relatively stronger relationship with core users. Having retrieved the user data from Twitter, the interactions between users within the individual country sets were analyzed with the objective of detecting communities of related users. At the country level, an interaction is defined as one user "mentioning" another with the inclusion of "@<username>" within a tweet, where reciprocal mentions between users can be considered as a dialogue [3], thus potentially indicating the presence of a stronger relationship. A network is created with n nodes representing users, and undirected weighted edges representing reciprocal mentions between pairs of users, with weights based on the number of mentions occurrences. Currently, all mentions occurrences are considered rather than selecting those from a specific time period. Any connected components of size < 10 are filtered at this point. We use the method of Greene et al. [17], which is a variant of the recent work by Lancichinetti & Fortunato [18], to generate a set of stable consensus communities from a reciprocal mentions network, where 100 runs of the OSLOM algorithm [19] are used to generate the consensus communities. In this analysis, we are primarily concerned with the detection of the most "significant" communities with the strongest signals in the network. Therefore, the consensus communities are ranked based on the stability of their members with respect to the corresponding consensus matrix. We employ the widely-used adjustment technique introduced by [20] to correct for chance agreement: CorrectedStability= Stability-ExpectedStability 1-ExpectedStability A value close to 1 will indicate that C is a highly-stable community. As higher values of the threshold parameter τ used with the consensus method result in sparser consensus networks, and having tested with values of τ in the range [ 0 . 1 , 0 . 8] , we selected τ = 0 .5 as a compromise between node retention and more stable communities. Finally, community descriptions are generated using a TF-IDF vector for each community, where the terms are hashtags contained within tweets posted by users within the community. A description consists of the top ten hashtag terms ranked using their TF-IDF values. --- Case Study: USA A reciprocal mentions network was created for the USA core set and associated data, consisting of 835 nodes and 2501 edges, where 29 of the original 32 core set nodes were present. The consensus communities were then generated using the methodology described earlier. The consensus network generated with threshold τ = 0 .5 consisted of 672 nodes and 6876 edges. Fifty-five communities were detected, and a selection of relevant communities having high stability scores can be found in Table 2. Community A would appear to be national socialist/white power in nature, with the appearance of hashtags such as "aryan", "thewhiterace", "wpww" , "masterrace" and "nazi". An analysis of the users and associated profiles finds references to the American Nazi Party, along with other related terms such as 14 , and 88 in user names. There are also references to skinhead groups, including a website where related media and merchandise can be found for sale. Users appear to be mostly from the USA, although a small number of European users are also present. References to the London riots of 2011 are also made within tweets, where blame has been apportioned to nonwhites. "Pipa" refers to the "PROTECT IP Act", which appears to be a shared concern among such groups with respect to potential curbs to civil liberties. Some similar white power themes appear in community B, whose description also includes "wpww", although they would appear to be more subtle than those of community A. An analysis of the user profiles finds a certain number of North American white power users with some similar references as community A, albeit on a smaller scale. One of the core users promoting a radio station website could be considered influential here. The "trayvontruth" and "treyvon" hashtags refer to the recent fatal shooting in Florida of Trayvon Martin, a 17-year-old African-American teenager. It appears that this story is being used to propagate a message of alleged persecution against whites. References to "tcot" and "gop" can also be seen, indicating the presence of more traditional conservative users. However, this does not necessarily point to any official link between these groups. The appearance of the "sopa" hashtag is most likely analogous to that of "pipa" in community A's description. The users in community C form quite a large North American community which is primarily white power oriented, as can be seen from the frequent occurrences of "wpww" among the description hashtags, along with "rahowa" . There are many national socialist references within the user profiles, and as with community A, many user names containing 14 and 88. In a similar fashion to community B, an external white power radio station website seems to be influential here, which correlates with some of the hashtags such as "wpradio", "contest" and "wpwwgiveaway". Most of the users in community D appear to be connected with a number of white rabbit websites, which allege the existence of "white genocide". Here, "antiracist" refers to the associated slogan "antiracist is just a code word for anti-white". This community also contains one liberal user with a high-level of posting activity who has interacted with a certain number of the other white power users. Communities E and F are interesting as neither community contains a single user from the USA core set. However, an analysis of the users in both communities shows that reciprocal follower relationships with USA users are common. The users in community E appear to be white South Africans, with some profiles containing racist and national socialist references. Many tweets from these users relate to perceived cultural threats from black South Africans that are often retweeted by international users. Almost all of the description hashtags are related to South Africa, such as "anc" and "afrikaners" . Community F consists mostly of users associated with the English Defence League , a group opposed to the alleged spread of militant islamism within the UK. Hashtags such as "rochdale", "brighton", "luton" and "dewsbury" are UK locations, most likely indicating various EDL subgroups, while "uaf" refers to the Unite Against Fascism group; a staunch opponent of the EDL. Although both communities E and F would appear to have a certain amount of awareness and interaction with other nationalities as indicated by the follower relationships, the majority of interaction seems to be at the local level if mentions interactions are assumed to represent stronger relationships. A reciprocal mentions network was created for the Germany core set and associated data, consisting of 247 nodes and 646 edges, where 46 of the original 53 core set nodes were present. The consensus communities were then generated using the methodology described earlier. The consensus network generated with threshold τ = 0 .5 consisted of 167 nodes and 799 edges. Eighteen communities were detected, and a selection of relevant communities having high stability scores can be found in Table 3. An analysis of the users in community A finds them to be primarily associated with the town of Geithain, near Leipzig in Sachsen. This can also be seen from hashtags such as "geithainer", "geithain" and "gha", along with the Leipzig-related hashtag "lvz" . Users belonging to various extreme right groupings are present, such as Freies Netz and the Junge Nationaldemokraten . References to Aktionsbüros are also made. The "bollywood" hashtag refers to a Geithain business that has been subjected to repeated attacks by neo-Nazis, while "unsterblichen" refers to anti-democratic flashmob marches that have been occurring sporadically throughout Germany. These protests are linked to Spreelichter, an extreme right group from Südbrandenburg that was recently banned by the local authorities. 1 They used social media to propagate national socialist-related material, including professional-quality videos of the marches themselves. --- Case Study: Germany 1 http://blog.zeit.de/stoerungsmelder/2012/06/19/das-ende-der-nazi-masken-show_8923 Like community A, the users in community B appear to be mostly related to a geographical location, namely the federal state of Thüringen, where the towns of Altenburg, Gera and Saalfeld can be found. Most of the users are associated with Freies Netz, with a number from other parts of Germany. The "rfd" hashtag refers to "Rock für Deutschland" -a concert organised by the NPD in Thüringen which will take place in July 2012. Other relevant hashtags include "apw" and "13februar", a reference to the bombing of Dresden which began on February 13, 1945. This event is usually commemorated by extreme right groups each year. Also relevant is "volkstod" , which refers to the perceived destruction of German people and traditions since World War II, often mentioned in Spreelichter-related content. The binding theme of community C is related to the "130abschaffen" hashtag, which refers to demands for the abolition of a paragraph in the German penal code associated with the criminalization of incitement to hatred, along with denial and/or justification of the Holocaust and national socialist rule. Although other relevant hashtags such as "apw", "volkstod" and "spreelichter" are present, some of the users have been inactive for some time, and it appears that these may have been replaced with new accounts. Other interesting hashtags are "mobilisierungsvideo" and "unibrennt", where the latter refers to a student protest movement. This movement appears to be unrelated, but the use of this hashtag by known extreme right users may be strategic as it could ensure the propagation of tweets to a wider audience. Community D is the largest of the selected consensus communities, and contains a wide range of users from groups such as Aktionsbüros/bündnis, Freies Netz, Junge Nationaldemokraten and Spreelichter/Unsterblichen. These users would appear to be quite active, with many "apw"related tweets, along with tweets containing URLs linking to content hosted on external websites such as YouTube or other dedicated websites. The "altermedia" hashtag refers to a collective of politically-incorrect/nationalist-oriented news websites, where the German website contains content such as Spreelichter articles and a section related to the NPD. "Guben" is a town in the state of Brandenburg on the border between Germany and Poland, and its many occurrences may indicate a geographical connection with Brandenburg, particularly as Spreelichter was also based in this state. It also appears that some of the users here are replacement accounts for older versions that are members of community C. Finally, as with the USA case study, a community without any core set users has been selected for analysis. In this case, all of the users in community E appear to be related to the NPD. The hashtags "bamberg" and "ruhrgebiet" refer to locations within Germany, and the presence of other hashtags such as "flugblatt" and "rat" along with separate analysis of the tweet content may indicate mobilization prior to elections. Separately, the "türke" hashtag refers to people of Turkish descent, and is often used in tweets from community E users alleging the involvement of such people in criminal activity; a common accusation by German extreme right groups. Similarly, "stolberg" refers to the town where a German teenager was killed by non-Germans in 2008, which is also the focus of an annual extreme right commemoration. We also analyzed the international relationships between the various groups within the data sets, based on the interactions between the core users from the eight country sets . Two types of undirected network were generated; a followers network consisting of user nodes and unweighted edges representing follower links between users, and a mentions network where edges were created for each instance of a user "mentioning" another with the inclusion of "@<username>" within a tweet. As with the country-based networks, only reciprocal edges were used in order to capture stronger relationships, all stored follower and mentions instances were included, and connected components of size < 10 were filtered. --- International Follower Awareness The international followers network can be seen in Fig. 3. As might be expected, most of the follower relationships are between users from the same country, although a certain number of international relationships are identifiable. It would appear that linguistic and geographical proximity is influential here, for example, there appears to be a close relationship between the Spanish and Italian users, with strong connections also between the UK and USA. Similar behaviour with respect to social ties in Twitter has been identified by Takhteyev et al. [21] and Kulshrestha et al. [22]. However, there appear to be some exceptions to the influence of geographical proximity, most notably, Swedish and Italian users that are not co-located with their respective country nodes. In both cases, the majority of tweets from these users are in English, which presumably ensures a wider audience. The former user is a Swedish representative of a pan-Scandinavian group espousing national socialist ideals, who appears to be interacting with many international users, particularly from the USA. This user has been increasingly active in recent months. The Italian user is a national socialist whose tweets often contain URLs to music or video content hosted on external websites, but it is unclear if a direct association exists with any particular group. From an analysis of other central nodes in the network , it would seem that those involved in the dissemination of material with the use of external URLs, or media representatives such as extreme right news websites and radio stations, are attempting to raise awareness amongst a variety of international followers. This is especially the case when the English language is used. --- International Dialogue The follower-based relationship between international users could be considered as passive when compared with that of the mentions-based networks, where such interactions can represent actual dialogue between users. The mentions network in Fig. 4 can be seen to be somewhat smaller than the corresponding followers network in Fig. 3; for example, none of the Greek core users are present. Apart from this, the network has a similar structure to that of the followers network, in that most interaction occurs within individual country-based communities. Connections between these communities do exist, but are fewer than in the followers network. The influence of linguistic proximity appears to take precedence here, with the use of English playing a major role as mentioned in the previous section. For example, a relatively large number of connections remain present between the UK and USA users. In the case of the German community, while the followers network contains a number of connections with other international users, this has now been reduced to a single connection with a user acting as an English-language Twitter channel for a Swedish nationalist group. Similarly, the Swedish user co-located with the USA community is the same user as that in the followers network, who appears to be involved in many Englishbased dialogues with international users. It should also be emphasized at this point that this analysis does not necessarily provide extensive coverage of the international relationships between all extreme right groups that are active on Twitter. As it is possible that the data retrieved from Twitter is incomplete, the objective is to demonstrate the existence of these relationships by means of an exploratory analysis. This caveat also applies to the country-based community analysis. --- Conclusions and Future Work Extreme right groups have become increasingly active in social media websites such as Twitter in recent years. We have presented an exploratory analysis of the activity of a selection of such groups using network representations based on reciprocal follower and mentions interactions. The existence of stable communities of associated users within individual countries has been demonstrated, and we have also identified international relationships between certain groups across geopolitical boundaries. Although a certain awareness exists between users based on the follower relationship, it would appear that mentions interactions indicate stronger relationships where linguistic and geographical proximity are highly influential, in particular, the use of the English language. In relation to this, media user accounts such as those associated with extreme right news websites and radio stations, along with external websites hosting content such as music or video, are a popular mechanism for the dissemination of ideals among users from a variety of disparate groups. Although a certain number of the detected communities can be associated with a specific extreme right group or ideology, this is more ambiguous in other cases where communities appear to contain members from a variety of known groups. This may be a consequence of incompleteness in the data sets retrieved for this analysis. It may also be related to variances in Twitter usage patterns in different countries, for example, the use of Twitter tends to be more prevalent in the UK and the USA than in other countries such as Germany2 . The laws of different countries should also be taken into consideration, as an opinion that may be legally voiced in one country may not be permitted in another, particularly within the context of extreme right ideals. However, it may also be the case that social media websites are merely used by such groups to disseminate related material to a wider audience, with the majority of subsequent interaction occurring elsewhere. In future work, we will address the issue of incompleteness in the data sets, including the current disparity in core set sizes. Local community analysis of countries other than the USA and Germany will be performed, and we will collect data beyond the eight countries used in this analysis. We also plan to study the temporal properties of these networks which will provide insight into the evolution of extreme right communities over time.

Many extreme right groups have had an online presence for some time through the use of dedicated websites. This has been accompanied by increased activity in social media websites in recent years, which may enable the dissemination of extreme right content to a wider audience. In this paper, we present exploratory analysis of the activity of a selection of such groups on Twitter, using network representations based on reciprocal follower and mentions interactions. We find that stable communities of related users are present within individual country networks, where these communities are usually associated with variants of extreme right ideology. Furthermore, we also identify the presence of international relationships between certain groups across geopolitical boundaries.

Background Weight gain during pregnancy is one of the main markers of physiological and somatic changes that reflect fetal development and growth [1,2]. However, insufficient or excessive gestational weight gain can threaten maternal and child health [3,4], contributing to the presence of unfavorable gestational outcomes, particularly in women classified as having high-risk pregnancies [5]. As determinants of maternal health, sociodemographic conditions, such as limited access to health care [6][7][8], also play an important role during pregnancy [9]. Unfavorable aspects, such as less education, the absence of paid work and a low family income, are described in the literature as social determinants associated with insufficient or excessive GWG [2,9]. Additionally, the evidence supports the hypothesis that household food insecurity can be a particularly important barrier to maintaining health during pregnancy [10]. Studies performed to assess the effects of HFI on weight gain during pregnancy observed that HFI was related to both the possibility of insufficient weight gain [11] and metabolic adaptations that increase the risk of obesity [12]. HFI is defined as the lack of secure access to sufficient amounts of safe and nutritious food for normal growth and development and an active and healthy life [13]. Psychometric scales contribute to the assessment of HFI, which can be measured based on concerns about a lack of food to the restriction and experience of hunger in cases of more severe limitations [14,15]. It is important to note that women are more vulnerable to HFI. The global assessment of the impact of HFI showed that women were approximately 27% more likely to be exposed to more severe HFI than men [16]. In Brazil, data from the last Family Budget Survey revealed a vulnerability and limited access to food in households headed by women, approximately 52% of whom were affected by severe HFI [17]. A similar finding was reported in the study by Lignani et al. , which evaluated a model of HFI determinants. According to the authors, when the head of the family was female, there was a greater probability of having occupations with lower income levels [18]. The relationship between women and HFI is the result of unequal access to and control of financial resources and unstable incomes affecting the acquisition of food [19]. Additionally, the main causes of HFI for women have been revealed by different sources, such as poverty, low education, being out of the labor market, a large family, the number of children under five, low food diversity and a low frequency of daily meals [16]. During pregnancy, HFI can contribute to nutritional problems related to deficiencies in food intake during a stage of a woman's life with a higher demand for nutrients for fetal growth and development. HFI is a potential factor associated with the health risks of pregnant women and can contribute to the aggravation of disorders associated with gestational risk, such as anemia and gestational diabetes mellitus [10,12]. A systematic literature review showed that the associations between HFI and the health of pregnant women were related to poor dietary quality, dietary diversity and inadequate nutrient intake; moreover, HFI was related to higher risks of depression and anxiety [20]. Although HFI is an important measure to be evaluated during pregnancy, few studies have explored its effects on the adequacy of GWG in at-risk women during pregnancy. The investigation of this relationship may assist in the identification of pregnant women exposed to unfavorable social health conditions and development of interventions to reduce the pregnancy risk related to insufficient GWG and excessive GWG. This study aimed to verify the association of HFI and unfavorable sociodemographic characteristics with total GWG inadequacy in high-risk pregnant women. --- Methods --- Study design This prospective cohort study was carried out with pregnant women from a maternity and prenatal unit of a public university hospital located in the metropolitan region of Rio de Janeiro . The UH has a coverage area comprising seven municipalities in the metropolitan region of the state of Rio de Janeiro [21]. Comparing the characteristics of these municipalities, it is possible to perceive different sociodemographic conditions. In 2013, the evaluation of the municipal human development index , for example, indicated that in the municipality where the UH is located, the IDHM was 0.837, the highest index among the seven municipalities. On the other hand, among the municipalities served by the UH, the lowest IDHM rate was 0.654. The IDHM is a number from 0 to 1, where 1 represents greater human development in the locality, and the measure is composed of three dimensions: longevity, income and education. In the municipality where the UH is located, life expectancy at birth was 76.2 years; the average per capita monthly income in real was R$2.000 ; and among the population over 25, 2.6% were illiterate, and 33.7% had the highest level of education. In contrast, the municipality with the lowest IDHM had a life expectancy at birth of 72.6 years, with an average per capita income of R$ 440 ; in addition, 11.1% of the population was illiterate, and only 3.7% had the highest level of education [22]. This information indicates the differences and inequalities among pregnant women attending UH. In Brazil, all pregnant women are entitled to access prenatal care in the Unified Health System by the Primary Health Care Service . At the PHC, pregnant women are monitored periodically, and when any pregnancy risk is identified and diagnosed, they are referred by the National Regulation System 1 to a more complex prenatal service. The classification of gestational risk is a dynamic process that identifies pregnant women who need interventions that require greater assistance and technological resources and more complex care according to their risk potential, health problems or degree of suffering [6]. The gestational risk factors considered for referral to high-risk prenatal care may be related to previous clinical conditions , factors related to their previous reproductive history and factors related to the current pregnancy [6]. The UH prenatal and maternity service specifically serves pregnant women referred and classified as having some gestational risk. Pregnant women are referred to a screening consultation to confirm the diagnosis of pregnancy risk. Nonconfirmation indicates the counterreference, that is, the return of these pregnant women to the primary PHC, whereas confirmation of gestational risk indicates continuity of care in the most complex prenatal care units. In this study, prenatal and maternity ambulatory care was performed at the UH. All of the referred pregnant women who had a proven gestational risk and remained in the UH were approached, and those who consented to participate in our study were included in the sample. Figure 1 shows the flow of care of these pregnant women from the identification of gestational risk in PHC to the possible outcomes in prenatal care regarding the diagnosis of gestational risk until delivery at the UH maternity ward. The data collection for this study was carried out between August 2017 and October 2019 for the sample selection in addition to the pregnant women presenting with some confirmed gestational risk. The eligibility criteria were age ≥18 years, a single pregnancy and an expected delivery through October 2019. This study complies with the Declaration of Helsinki [23]. This study was approved by the Research Ethics Committee of the Federal University of Rio de Janeiro . --- Variables included in this study --- Gestational weight gain The total GWG was calculated as the difference between the final weight and the initial weight. The GWG is a measure for assessing and monitoring the health of pregnant women carried out by health professionals in several prenatal units, hospitals and maternity units [6]. The recommendation of the total GWG of the Institute of Medicine was adopted, so it was necessary to calculate the initial body mass index by the weight and height ratio [6,24]. Height and weight measurements at the beginning of pregnancy were collected from the pregnant woman's health booklet, a document used in prenatal care for monitoring and providing health guidelines during pregnancy. These were measured during the prenatal consultation at the PHC of origin of the pregnant women. Most of the pregnant women included in this sample had these measures reported in the first gestational trimester , an opportune time to check their initial weight [6]. The first prenatal consultation of the UH for approximately 76% of the pregnant women occurred after the first gestational trimester, a characteristic that was expected, since some of the risk factors appear or are diagnosed only after the first gestational trimester [6]. In this study, we chose to use the weight measured in the first trimester of pregnancy and recorded in the pregnant woman's health booklet as the initial weight. Thus, even if the consultation of the pregnant woman at UH took place after the first trimester of pregnancy, we considered the initial weight measurements measured at PHC until the 13th week of pregnancy. initial BMI During the first prenatal consultation at UH, an interview was carried out to apply the data collection questionnaire. At this moment, the pregnant woman's health booklet was requested to verify the initial anthropometric measurements of height and weight, and they were considered to estimate the initial BMI [6]. For pregnant women who started prenatal care at the PHC after the first gestational trimester and without data about the initial weight registered in the pregnant woman's health booklet , selfreported initial anthropometric measurements were requested from them and considered in the analyses. This assessment considers the premise that the measurement of the initial weight represents the weight measured in a period of up to two months before pregnancy and during pregnancy at the limit of 13 weeks of gestation [6]. 1 SISREG is a virtual management system of the Ministry of Health that manages from the basic and specialized outpatient network to hospital and hospitalization, at the levels: municipal, state, federal, including private services of the Unified Health System and university students, seeking greater control flow and optimization of the use of resources, in order to guarantee the humanization of services. To determine the final weight, the medical records of the pregnant woman in the maternity ward were consulted. The final weight was considered to be the weight measurement at the last prenatal consultation at UH, two weeks before the date of delivery. In the absence of information about this measure in the medical record, the women were asked about it in the maternity ward after delivery, and the self-reported final weight was considered . The initial BMI was categorized as follows: low weight ; normal weight ; overweight ; and obese . The total GWG was classified into three categories of weight gain according to the ranges established for the initial BMI: low weight -12.5-18.0 kg of total GWG; normal -11.0-16.00 kg of total GWG; overweight -7.0-11.5 kg of total GWG; and obese -5.0-9.0 kg of total GWG [17]. The total GWG was adjusted according to the gestational age as a measure of the final weight. Household Food insecurity HFI was estimated based on the Brazilian Household Food Insecurity Measurement Scale . The EBIA is a validated instrument for the Brazilian population that investigates the dimensions of access to food through the perception of a family member regarding the lack of financial resources available for the purchase of food [14]. This psychometric scale comprises 14 items and categorizes households according to HFI exposure, ranging from concern about the lack of food at home and insufficient quality of meals due to a lack of food and hunger. The HFI severity levels are measured in the EBIA by the sum of affirmative responses to each item on the scale . The absence of positive responses characterizes the condition of household food security , which indicates that the respondent has no concern or perception of income limitations regarding food purchase and access. The cutoff points classify the HFI Fig. 1 Flowchart for selecting the sample of pregnant women at risk in a university hospital . Metropolitan area of Rio de Janeiro, Brazil, 2019 into three levels: mild HFI; moderate HFI; and severe HFI. More details about EBIA have been described in the literature [14,[25][26][27]. In this study, the levels of moderate HFI + severe HFI were considered in the same category. Sociodemographic and health for pregnancy variables Sociodemographic and gestational characteristics were included to describe the sample and to investigate social vulnerabilities that could contribute to health inequalities and negatively affect GWG. Maternal age in years was investigated and categorized , considering that pregnant women aged > 35 years tend to have more unfavorable gestational outcomes [6]. The level of education was classified into three categories: < 9 years ; 9-15 years ; and > 15 years . Despite the lowest level of education being a risk factor for pregnant women [6], the categories established in this study considered the highest level of education of the sample of pregnant women, and we classified them by using a category of less education as incomplete primary education . Schooling was described in association with health conditions and GWG [28,29]. Unsafe marital relationships also compromise the health of pregnant women [6]. The variable marital status considers the impact of living alone and the difficulties faced by single women and mothers in accessing material and financial resources due to gender inequality [17]. In this evaluation, the categories were established [28]. Drinking water consumption was studied as a gestational risk factor reflecting the unfavorable environmental conditions to which pregnant women are exposed [6]. Data on the race/ethnicity variable were self-declared and evaluated in the categories of white or black/mixed race, considering the vulnerability that racial inequality confers on the health of populations [30,31]. Family income was investigated in relation to the minimum wage, and two categories were studied . The minimum wage is the minimum amount regulated by law for paid work [32]. Taking 2018 as a reference, the annual value for the minimum wage in Brazil in the current currency was R$954, which corresponds in this period to USD 261.4 [33,34] . Family income is an important measure to investigate access to resources such as the purchase of food by families [35]. In relation to health, the following information was investigated as determinants of GWG: The number of prenatal consultations was considered in categories established based on the recommendation of 8 or more consultations for the monitoring and health care of pregnant women and babies, including the identification of risk factors and monitoring of GWG [6]. The number of previous pregnancies was obtained, as studies indicate the highest risk of excessive GWG in multiparous pregnant women [28,30,36]. Gestational age at the time of delivery was also obtained. Initial BMI [24] was assessed in three categories ; the decision to sum the categories of overweight BMI and obese BMI was due to the sample size. The consumption of alcoholic beverages and smoking during pregnancy were also investigated in relation to GWG [6,36]. With the exception of the variable total number of prenatal consultations and gestational age at delivery, which were investigated at the maternity hospital, all of the other data were obtained at the first consultation of the pregnant woman at the UH. --- Statistical analysis We estimated the proportions and confidence intervals to describe the study variables. In the first step, the chi-square test was used to compare sociodemographic, HFI and pregnancy characteristics between GWG categories . The variables with a level of significance in this first analysis, defined as p < 0.20, were included in the multivariate analysis. The decision to adopt a more conservative level of statistical significance followed the recommendations used in the literature [37,38]. In the second step, a multinomial regression model was used to verify the variables associated with GWG. We assigned the adequacy GWG category as a reference in relation to the inadequate GWG categories. The data are expressed as odds ratios and 95% CIs. Variables that in the bivariate analysis were associated with GWG, with a significance level of up to 20% , were considered in the adjusted model. In the final model, the level of significance was defined as a p value <0.05. All data processing and analysis were performed using the Stata statistical package [39]. --- Results During the reference period of this study, 286 pregnant women were referred to their first consultation in the obstetrics sector of the UH prenatal clinic, and 117 of them were ineligible . The final sample included 169 women with an average age of 30.2 years, and their race/ethnicity was predominantly black/mixedrace. Most were married or lived in a stable relationship, had completed high school, and had a family income ≤ twice the minimum wage. Regarding the gestational profile, most women attended at least eight prenatal consultations and did not consume alcohol or smoke during pregnancy. More than half of the women included in this study were exposed to some level of HFI. Considering their initial BMI, there was a high proportion of pregnant women with obesity. Inadequate total GWG was found in most women; 47% of women had excessive GWG, and approximately 28% of pregnant women had insufficient GWG . The categories of gestational weight gain were correlated with the following variables: maternal age, marital status, educational level, smoking and alcohol consumption during pregnancy, number of prenatal visits, weeks' gestation at delivery and household FI . Insufficient GWG was more common in women with a lower educational level , married marital status , age up to 35 years and those who reported alcohol beverages consumption and cigarettes . In turn, excessive GWG was more common in women over 35 years old , with a single marital status and with a higher educational level . Additionally, more common were the number of prenatal consultations ≤ 8 and a gestational age at delivery greater than 40 weeks' gestation for women with excessive GWG. Regarding FI, exposure to moderate/severe FI was more common for women with both insufficient GWG and excessive GWG . Table 3 presents the unadjusted and adjusted multinomial logistic regression. In the bivariate analysis, more years of schooling and the woman's age were considered protective factors for insufficient GWG. Living without a partner, being a current smoker and having <8 total prenatal consultations were considered factors that increased the OR of insufficient GWG. An increased risk for insufficient GWG and excessive GWG was observed with an increase in the number of gestational weeks at delivery . HFI was not related to GWG. After adjusting the model, only the educational level of the pregnant woman was significantly and inversely associated with insufficient GWG; thus, the higher the education level of the pregnant woman was, the lower the risk of inadequate GWG was . --- Discussion Our study showed that inadequate GWG was not associated with exposure to FI during pregnancy in this sample of women with high-risk pregnancies. However, after assessing the sociodemographic and health conditions of pregnant women, a higher level of education was a protective factor against insufficient GWG. This study demonstrated a higher prevalence of women with excessive GWG. An analysis of secondary data from multicenter studies in the United States involving 8293 women showed that 73% had excessive GWG, corresponding to approximately three in four women [40]. Zhao et al. [4] identified approximately 50% of pregnant women with excessive GWG in a study conducted in China, and 15.2% had insufficient GWG. In Brazil, Campos et al. [41] evaluated the adequacy of GWG among pregnant women in the northern region of the country and observed a prevalence of excessive GWG in almost half of the sample and insufficient GWG in 19% of women. A similar result was observed for pregnant women in Rio de Janeiro; according to the authors, almost 50% of women evaluated had excessive GWG, and less than 30% had adequate GWG [42]. A direct determinant of GWG is the initial BMI [14]. In this study, no associations were observed between these variables. However, we observed that more than half of the pregnant women had an initial BMI corresponding to overweight and obesity. In contrast, a low initial BMI was observed in only 2.6% of the sample, a proportion close to the results of the Brazilian Food and Nutrition Surveillance System during 2019, when 2.5% of adult Brazilian women were found to have a low BMI. Data from the Brazilian SISVAN also revealed an increase in the prevalence of overweight BMI among women of reproductive age; 34.2% were considered overweight, and 29.7% were obese based on BMI [43]. In this study, it was also found that more than half of the mothers were exposed to some level of HFI during pregnancy. Oliveira et al. [44] investigated HFI in a group of pregnant women attending PHC in northeastern Brazil and identified a prevalence of 42.7% of HFI in the sample, a value lower than the result in this study. In addition, in the present study, we opted for the sum of the most severe levels of HFI . Thus, it was observed that 12.3% of women had limited access to food in terms of quantity and quality and possibly continued to experience hunger; this proportion was lower than that reported by Marano et al. [45] when investigating HFI in pregnant women in two cities in Rio de Janeiro . Other studies carried out in Brazil evaluating HFI during pregnancy showed significant variations in the prevalence estimated by the EBIA, depending on the region of the country. In studies carried out in the northeastern and southeastern regions of Brazil, the authors reported a higher prevalence of HFI among pregnant women, ranging from 59% to 71.6% [46,47] and 37.8% in southeastern Brazil [48]. These differences in prevalence are consistent with the distribution of severe HFI in Brazil. Although a higher prevalence of HFI was observed among the women in this study, no relationship with GWG adequacy was found. Few studies have evaluated the relationship between GWG and HFI. A meta-analysis identified that pregnant women exposed to home HFI had an increased prevalence of discrepant weight gain, both excessive and inadequate weight gain [49]. Other studies have reported a lack of an association between GWG and HFI, as observed by Laraia et al. [50]. According to these authors, no significant association was found between HFI and GWG among pregnant women in the United States. However, the authors observed that the average weight of women with HFI in the study was higher, as well as the GWG adequacy rate, suggesting HFI as an indicator for excessive GWG [50]. In Brazil, a study of pregnant women in the northeastern region of the country to identify the association among different factors, including HFI status and GWG, showed no significant association with HFI and the GWG outcome; this finding was similar to that of this study [44]. The main finding of this study indicates a reduction in the risk of insufficient GWG for women with more years of schooling, since pregnant women with access to secondary and higher education were more protected from this level of GWG than women with less than nine years of schooling. In the literature, a lower risk of insufficient GWG was also associated with higher levels of schooling [51]. On the other hand, a higher level of education was related to a higher risk of excessive GWG [8,45]. In addition, higher education has been described as being associated with a greater likelihood of GWG adequacy [28] and a lower likelihood of excessive GWG [52]. Pregnant women with less than four years of education had a risk of excessive GWG that was approximately five times greater than that of other pregnant women [53]. These findings indicate that the increase in years of schooling may act as a protective factor for GWG. In addition, the risk associated with low education and a greater propensity for excessive GWG [54,55] and insufficient GWG [56] point to the importance of education as a social determinant of health [57]. Thus, understanding that inadequate GWG is a modifiable risk factor and that both excessive and insufficient weight gain during pregnancy can threaten the health of women and their children [58] is important for strengthening health services and factors related to the environment, education and health promotion [59]. This study has some limitations. For the pregnant woman's health booklet and height measurements, when these measures were not available, the self-reported data of the pregnant women were considered. In such cases, the possibility of biases attributed to overestimation or underestimation of measures must be considered [60,61]. Researchers have corroborated the quality of the measures of weight and height measured by anthropometry versus those obtained by self-report in populationbased studies. For example, Conde et al. compared the measured and reported weight and height measurements of the Brazilian population and highlighted that the BMI estimates, measured or reported, were relatively close. In addition, a validation study of self-reported measures found that 84% of women who reported their weight and height measurements were categorized appropriately in regards to their BMI classification [62]. Health behaviors and interventions such as nutritional monitoring, possible interactions of the diet to control gestational weight gain and physical exercise during pregnancy were not evaluated in this study. However, there is some evidence regarding the low impact of these measurements on weight gain during pregnancy [63]. The investigation of family income in terms of minimum wages limited the assessment of income in relation to the number of family members . --- Conclusion In this population, an important factor for GWG is maternal education, among the other determinants evaluated and widely discussed in the literature as predictors of GWG. Additionally, a lower level of education of the pregnant women evaluated may have preceded the effects of HFI on its relationship with the GWG of these women with high risks during pregnancy. Given the risk of inadequate gain during a high-risk pregnancy and given that GWG is a modifiable factor, the importance of additional support and health counseling is highlighted, particularly nutrition education interventions, such as a health-promoting tool for pregnant women and fetuses, particularly for women who are more socially vulnerable and who have low levels of formal education. Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Educational level < 9 Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [ Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [ Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Weeks' gestation at delivery < 40 weeks Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [1] Ref. [ --- --- Abbreviations BMI: body mass index; CI: confidence intervals; EBIA: Brazilian Household Food Insecurity Measurement Scale ; GWG: gestational weight gain; HFI: household food insecurity; HFS: household food security; IDHM: municipal human development index ; IOM: Institute of Medicine; OR: odds ratio; PHC: primary health care; R: Real; SISREG: National Regulation System ; SISVAN: Brazilian System of Food and Nutritional Surveillance ; UH: university hospital; USD: United States dollar Authors' contributions AAR participated in the coordination, collected and interpreted the data. RSC and AAR contribute to the study design, statistical analysis and writing of the manuscript. ALPA conceived the idea of the study, participated in the coordination and helped in the writing of the manuscript. All authors participated in the study design, read and approved the final manuscript. --- Authors' information AAR is a student of the Graduate Program in Nutrition, Institute of Nutrition Josué de Castro, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil. ALPA professor at the Department of Clinical Nutrition, Faculty of Nutrition Emília de Jesus Ferreiro, Fluminense Federal University, Rio de Janeiro, Brazil. RSC professor at the Department of Social and Applied Nutrition, Institute of Nutrition Josué de Castro, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil. --- --- --- --- Competing interests The authors declare that they have no competing interests. ---

Background: Inadequate gestational weight gain (GWG) is associated with adverse outcomes in maternal and child health and can be enhanced by social inequalities, such as lower education and household food insecurity (HFI). Women are more vulnerable to HFI, which has been associated with negative health effects for pregnant women during the prenatal and puerperal periods, particularly in regard to the aggravation of pregnancy risks. This study investigated the association between sociodemographic characteristics and HFI with respect to adequacy of total GWG among women with high-risk pregnancies. Methods: This was a prospective cohort study that evaluated the total GWG of 169 pregnant women. The women were seen at a public university hospital in the metropolitan region of Rio de Janeiro (Brazil). Their sociodemographic and gestational characteristics and the Brazilian Scale of Domestic Food Insecurity were investigated. To estimate the total GWG, the difference between the patient weight at the last prenatal visit and the initial patient weight was verified, with both collected from the medical records of the pregnant women. The classification of the total GWG considered the recommendations of the Institute of Medicine (IOM) (2009). A multinomial logistic regression model assessed the risk (odds ratio; OR) and confidence intervals (CI 95%)) of insufficient and excessive GWG with exposure to HFI and other covariates (p value <0.05). Results: Insufficient and excessive GWG were observed in 27.8% and 47.9% of the pregnant women, respectively. More than half of the women (74.6%) had a high education level. Exposure to mild HFI occurred in 44.2% of the women. After adjustment, the HFI was not associated with insufficient or excessive GWG. The educational level of women was the only variable significantly associated with a lower risk of GWG insufficiency (OR: 0.10; 95% CI: 0.01-0.89). Conclusions: In this population, higher maternal education was a protective factor against insufficient GWG. We highlight the importance of additional health support and counseling for women in the most vulnerable social conditions, considering the importance of access to information for reducing health risks.

Introduction Quality of life is one of the emerging concepts in the world. It consists of new ideas about the state of the environment, happiness, housing, work, material satisfaction, and total well-being of the population living in the area . It is an indicator of good housing and the economic conditions of the human population. The most important development factors that influence the quality of life include education, health, income, employment, and environment. The term high quality of life refers to a high standard of living where people avail themselves of all the basic facilities of life in the form of education, health, and a peaceful environment. It is sustainable when they have a high standard of living, a state or condition in a society having no need of other help, and capable of fulfilling the needs of the poor . Furthermore, Seligman argued that a good quality of life enables people to perform the daily activities of life and attend to physical, psychological, and social well-being. It is the capacity to perform different tasks in connection with a person his/her role in the society. The Physical environment includes quality accommodation and access to public services, whereas the social environment contains social networks and level of social support from family members. Further, the socio-economic factors are the standard of living, means, and size of wealth. At the same time, cultural and health factors include age, gender, class, religious tendencies, general health, and physical mobility of a person. The world population of women depends directly or indirectly on men, and their quality of life is less compared to men. They avail fewer opportunities in all fields of life, such as education, land property, credit, health, and paid jobs, due to their low status in the world. Around two billion people in the world are considered poor, and twothirds of this population are women. Women have more work hours as compared to men but earn less because 60% of this is unpaid. Only 1% of total world assets are in women's names . Additionally, Women are also far behind men in the field of education. Approximately ten million girls are out of school compared to boys. Two-thirds of the world's women are illiterate. Due to poor economic status, less education and ignorance made women the victims of severe health hazards. Half a million women die during pregnancy and related diseases. Early marriages further lower their life quality in terms of income, education, and awareness. Violent behavior is the other big social problem for women as they face physical, sexual, and domestic violence. Gender representation is low in almost all fields of life, especially in political institutions. It has further restricted women from making required legislation for women's protection, equality, and better socio-economic status. All these made it extremely difficult for women to excel in the male status and improve the environment for a healthy life . Moreover, the situation of quality of life in the context of health in tribal women has again been affected due to the lack of access to basic health facilities. There are few professional doctors and lady health workers, which is insufficient for the rest of the Federally Administered Tribal Areas . Due to the extreme level of discrimination against women in tribal society, the health facilities are not being accessed and fulfilled properly. Women cannot make decisions without the consent of men. Many times, the quality of female life is compromised due to a lack of basic health awareness. The burden of female household duties combined with reproductive one results in both physical health problems and acute physical problems. They also face cultural taboos in the case of consulting doctors. They are based on false traditions and religious beliefs. Consequently, high mantle anxiety prevailed in Pakistan, especially among women, mainly due to bad economic conditions and lack of social support. The United Nations Food and Agriculture Organization reported that women have low calories and protein in the border area of Pakistan with Afghanistan. 56% of the population has food insecurity and health threats .According to the Government of Pakistan Housing Census , Federally Administered Tribal Areas face worse hygienic conditions. Only 10% of the population has adequate access to sanitation in the shape of toilets, solid waste disposal, and drainage. Livestock and Agriculture are the main income sources of livelihood in tribal areas of Pakistan. Women's involvement in both sectors is more than men's, but they always depend upon the male members while doing so. Women have to do a lot of work both at home and in the fields. However, this is unpaid and unrecognized. They have the capacity to sell milk, eggs, homemade products, preserved vegetables, and fruits in the market but could not do so due to purdah and cultural values. Furthermore, Federally Administered Tribal Areas are dominating society. The majority of women have been victims of patriarchy and feudalism. A female child often grows up with learning to obey and serve the males. The unequal gender distribution of relations hinders the process of development in FATA. In fact, the social structure and institutions, along with different norms and values systems, compel women to follow patriarchy. Women are not allowed to make decisions regarding their basic amenities of life, such as education, marriage, health, and other basic rights. Women are totally excluded from Jirga and don't participate in decision-making. Males, on the other hand, avail all the facilities in Jirga . Women in FATA are in worse conditions than the average Pakistani female. One main problem which is common for women in FATA is the security and political situation. In fact, women in tribal areas have become the victims of double discrimination since the beginning of the colonial era of British rule to successive governments after the creation of Pakistan in 1947. Under the Frontier Crime Regulation , women suffered both from a patriarchy-based social system and a law and order situation, which made FATA a safe haven for insurgents and criminals. Women's mobility and basic human rights have been further restricted due to deteriorated law and order situation . All decisions in FATA are made by the older family members in the joint family system. They look after all the affairs of human life, including the income sources that are jointly distributed. The joint family system, on the other hand, made restrictions on young women in all the important daily affairs, which proved to be a disaster for their mobility, control over income, and participation in decision-making . Moreover, the majority of tribal people were of the opinion that they were in favor of women's education and access to better health services. However, in practice, they are found against women's participation in the force. The cultural and religious beliefs are extremely strong for women in FATA. They are considered to be the center of honor, which ultimately affects women's mobility, human rights, and quality of life . The women in tribal areas want immediate actions to change the fate of their lives. They want constitutional rights that have already been given by the Pakistan government to the rest of the country's women. The honor and rights of FATA women cannot be improved under the strict FCR and Jirga system. Women in FATA not only suffer from a lack of protection under the law but also the victims of militancy and security operations. Forced marriages, honor killing, and the exchange of women between tribes and marriages have made women dependent on men. Many studies reported these as the worst for women in the whole region . --- Research Question 1. Does the unavailability of a strict culture system deteriorate the female quality of life? 2. Does the amalgamation of religious systems with culture affect the female quality of life? --- Methodology Universe of Study Kurram Agency was the universe of the study. It had three administrative divisions, namely Lower Kurram, Central Kurram, and Upper Kurrram. The major tribes living in the Kurram agency were Turi, Bangash, Mangel, Maqbal, Ghiligi, Maso Zai, and Ali Shirzai. The Agency covered a total number of 3,380 sq. km with total population and household of 448,310 and 39435 respectively . It was the third largest Agency of FATA. Due to the size of the area and population, it was not possible to collect data from the entire Agency. So, the data for the present study were collected from the tribe of Ali Shirzai. According to the 1998 Population Census Report, the total households of the Ali Shirzai tribe were 4495 households. However, for the present study, data was collected from 4 randomly selected villages, namely Tindo, Wrasta, Mirbagh, and Tandoori. The selected villages in Kuram Agency not only represented the pure Pashtun culture but also had been the victim of insurgency in FATA, which ultimately affected the quality of female life particularly. --- Sampling Procedure and Sample size All the households belonging to the Ali Shirzai tribe were the potential respondents of the study. According to the Tribal Areas 1998 Population Census Report, there were 4495 households in the selected villages. Though a large family size gives better results due to time, financial constraints, and human resources constraints, data were collected from 279 households. The sample size was drawn by keeping in the view table designed by Sekaran . A proportionate sampling procedure was used to select the sample size from each of the selected villages, while a lottery method of simple random sampling method was used to pick the samples from each of the selected villages (For more details, see --- Methods of Data Collection An interview schedule was designed in the light of the objectives of the study. It was pre-tested in the study area, and necessary changes were made to the interview schedule in light of the pilot survey. For the collection of information from the potential respondents, three females were trained and approached directly respondents of the study) in order to make it possible to collect data from them while keeping all ethical considerations in the prior view of the tribe. --- Result and Discussion Univariate Analysis Culture is a manmade part of the environment. Pakistan is a dominant society, and all the powers are in the hands of males. Most of the tribal people believe in tribal traditions, and they consider their tribal culture superior to religion. The wrong interpretation of religion and the mixture of culture into religion had a strong effect on women's quality of life. The Table shows that 52.0% of the respondents were in favor of the statement that women have no access to doctors due to restrictions imposed by tribal traditions, while 42.7% negated the statement. The remaining 5.0% were uncertain about the statement. Similarly, 69.9% of the respondents disagreed with this statement that male family members feel shy to visit doctors with women, while 30.1% of respondents were in favor of the statement that male family members feel shy to visit doctors with women. For the outside movement, 60.2% of the respondents responded that it is not fear that women in their family never travel beyond 25 square kilometers from outside the home for treatment purposes, while 39.4% of the respondents were in favor of the above statement. 68.5% of the respondents stated that their family wants to observe purdah, whereas 28.8% of the respondents disagreed with the statement. Additionally, 68.1% of the respondents further reported that women are considered a symbol of honor in the family, while 19.7 and 12.2% of the respondents disagreed or were uncertain about the abovementioned statement. Regarding the participation of women in jobs, 70.3% of the respondents reported that it is against basic Pashtun ethics, while 28.3% were in favor of women's participation in jobs. 49.1% of the respondents were again of the opinion that religion in tribal society is interpreted through the eyes of Pashtun culture, whereas 36.9% of the respondents disagreed, and 14.0% of respondents had no information about that statement. 47.7% of the respondents further responded that the mixture of Pashtun and religious culture is no threat to women's basic rights, while 47.7% of the respondents reported that it is a threat to women's basic rights whereas 9.3% of respondents were not sure about the above statement. As far as the share of women in family inheritance is concerned, 67.4% of the respondents reported that women could not dare to demand their share in the family inheritance, whereas 19.7% of the respondents agreed with the statement whereas 12.9% of the respondents were uncertain about the family inheritance. Moreover, 49.8% of the respondents were against the statement that tribal culture has nothing to do with religion, while 40.9% of the respondents were in favor of the statement. The remaining 9.3% of the respondents were uncertain about the statement. Similarly, 45.9% of the respondents were against the statement that more inclination toward culture made Islam partial towards female basic human rights, while 40.1% of the respondents were in favor of the statement and 14.0% of the respondents were uncertain. 66.7% of the respondents reported that they are not allowed to enjoy legal rights, while 29.4% of respondents agreed with the statement, while 3.9% of respondents had no information about the legal rights of women. Furthermore, the Table as a whole reveals that culture and religion played an important role in the improvement of female quality of life. These two were interrelated concepts, and the wrong interpretation of religion and the mixture of culture into religion had made women deprived of their basic human rights. There was a dire need to raise voices for the improvement of quality of life and challenge all cultural barriers in the way of women's development. Similarly, Ibrahim & Farina stated that our culture is intermixed with our religion, and their combination does not permit women to enjoy any legal right given by our Islam and the constitution. Pakistan is an Islamic country, but the different interpretation of Islamic law by our culture has deprived women of the basic rights of human life. This ultimately affects the female's quality of life. The cultural and religious beliefs are extremely strong for women in FATA. Shinwari reported that they are considered to be the center of honor, which ultimately affects women's mobility, human rights, and quality of life. Muhammad et al. reported that they also face cultural taboos in the case of consulting doctors. They are based on false traditions and religious beliefs. Consequently, high mental anxiety prevailed in Pakistan, especially among women, mainly due to bad economic conditions and lack of social support. --- Bivariate Analysis Association between Cultural and Religious Beliefs and Quality of Life The Table reports that in connection with the association between the quality of life and cultural and religious beliefs, a significant association was found with the statements calling religion in tribal society is interpreted through the eyes of Pashtun culture , women are a symbol of honor in the family , it is extremely difficult for women to do a job , it is not easy for women in the tribal area to enjoy legal right , for the outside movement women want to observe purdah and tribal women cannot dare to demand their share in the family inheritance . The statements found non-significant included the mixture of Pashtun and religious culture as a threat to women's basic rights , women not accessing doctors due to restrictions imposed by tribal tradition , male family members falling shyness to visit doctors with women , most of the women in the family never traveled beyond 25 square kilometers from outside home for treatment purposes , tribal culture has nothing to do with religion and more inclination towards culture made Islam partial towards female basic human right . The Table as a whole reveals a strong domination of Pashtun culture and religion. In fact, religion seems to be through the goggles of culture. Women in the study area were the symbol of honor and observed purdah; participation in the workforce is a deviation from basic Pashtun ethics, no share in family inheritance, and no legal rights. All these affected female quality of life in one way or another way. Ibrahim , in this regard, states that the mixture of culture into religion in tribal areas does not permit women to enjoy legal rights which is given by Islam and the constitution of Pakistan. The wrong interpretation of Islamic law has also deprived women of their basic rights to human life. This ultimately affects the female's quality of life severely in tribal areas. Most of the tribal people believe that strong cultural and religious belief systems prevail in tribal areas. Similarly, Shinwari reported that they are considered to be the center of honor, which ultimately affects women's mobility, human rights, and quality of life. --- Conclusion Quality of life is one of the emerging concepts in the world. A number of incentives and gender-based policies have been made to improve female quality of life in Pakistan. However, the study reveals that low female literacy, strong seclusion system, a dominant society, and deteriorated law and order situations of tribal people were the major causes of low female quality of life in FATA. Most of the tribal people strongly believed in cultural and religious belief systems. Further, the misinterpretation of culture and amalgamation of cultural traits into religion, consider of women as a symbol of honor, not being allowed to participate in jobs, strong purdah system, and share in family inheritance are another major factors that further affected the female quality of life in Federally Administered Tribal Area . --- Recommendations The study suggests following recommendations on the basis of the findings of the study. 1. Education is the key agent of promoting awareness and sensitization of local people's lifestyles. A quality life could be ensured in the selected area by providing them with adequate education. The government, in this regard, emphasized the promotion of female education. 2. Community organizations can strengthen the voice of women; as a result, women being organized can initiate an advocacy and lobbying campaign to discourage those socio-cultural practices and beliefs that might condense the role of women, which ultimately leads to their poor quality of life. 3. Patriarchal social structure in the selected area is the prominent hurdles in the way of promoting women's quality of life. So, effective programs and policy interventions should be initiated to provide space for women's better lifestyles. 4. Health facilitation for the citizens is the prime responsibility of every welfare state; therefore, the government of Pakistan should give special attention to maternal health issues in the selected area of the study.

The current study was formulated to probe cultural and religious belief systems that affect the female quality of life in the Feudally Administrated Tribal Area (FATA). The Data for the study was collected through an interview schedule from 279 sample respondents belonging to four villages of Kurram Agency. The statistical test, i.e., Chi-square, was used to find out the association between the dependent and independent variables. The results of religious and cultural beliefs show a significant association between females were not allowed to move outside the home and strictly observing seclusion (P=0.002), female participation in a job is considered a violation of Pakhtun ethics (P=0.000), females not dare to demand their share in the family inheritance (P=0.004), cultural values of Pashtun society consider female as a symbol of honor (P=0.000) with female quality of life. The study recommends the female quality of life in FATA can be improved through the merger of FATA into the mainstream of Khyber Pakhtunkhwa province, replacing FCR with constitutional rights, providing all the basic facilities, and discouraging all norms and practices developed on the basis of patriarchy, misinterpretation of religion and culture by using clergyman, NGOs and other government functionaries.