Challenges and opportunities in coproduction: reflections on working with young people to develop an intervention to prevent violence in informal settlements in South Africa

The project: Siyaphambili Youth

Siyaphambili Youth (‘Youth Moving Forward’) is a 3-year research project aimed at addressing the social contextual factors that create a syndemic of high rates of IPV, HIV and poor mental health among young people living in informal settlements in KwaZulu-Natal (KZN) province in South Africa. The project is a partnership between the South African Medical Research Council (SAMRC), a South Africa non-governmental organisation (NGO) called Project Empower (PE), University College London (UCL) and a team of 17 young peer research assistants (YPRAs) living in urban informal settlements and rural communities in KZN. The principal investigators, coinvestigators and research staff from SAMRC, PE and UCL are referred to jointly as the ‘research team’ throughout this article as a means of discussing the relationship between the research team and the YPRAs as a central focus.

YPRAs were recruited to the project through an iterative hiring process designed to maximise possibilities for vulnerable youth to participate. Preliminary selection criteria included youth who were not currently in school or formal work, and were between 18 and 29 years old. The majority of those recruited had completed grade 12 but had not gone on to higher education. Three had stopped education after grade 11, and one had only completed grade 8. An informal introduction to the project was held in both rural and urban locations, in addition to proactive recruitment of young men by approaching them in places where they were seen to congregate. The project responsibilities, ethical considerations and payment terms were discussed in detail over the course of several days with potential YPRAs before they were asked to give their consent and were officially hired by PE. Following this process, five urban women, four urban men, four rural women and four rural men joined the Siyaphambili Youth Project as YPRAs.

Since November 2020, the YPRAs have participated in four phases of development of an intervention (figure 1). The first phase focused on understanding the syndemics of HIV, IPV and poor mental health risks (ie, how social contexts create overlapping risks characterised by poverty, violence, mental health vulnerabilities and gender inequalities), and young people’s agency in the face of such risks. YPRAs individually took photographs representing their daily lives (referred to as ‘artefacts’) and participated in in-depth interviews to discuss their meanings. In the second phase, YPRAs were involved in a series of workshops where they drew on their personal experiences to develop fictional characters, which were then used for facilitated discussions on young people’s daily lives.

Figure 1

Siyaphambili youth intervention development study phases. YPRA, young peer research assistant.

The third phase built on the previous two phases to develop a theory of change for the intervention. YPRAs did this by identifying risk and protective factors for HIV and IPV as part of a problem tree participatory workshop activity. They were asked to visualise risk factors for HIV and IPV including the underlying causes (tree roots), the more proximal causes such as behaviours (tree trunk) and the effects on young people’s lives (tree branches). A similar tree was created for the ‘solutions’ or actions that YPRAs thought would prevent HIV and IPV. As part of this process, the research team shared analyses from other research activities conducted as part of the project with the YPRAs (ie, a quantitative analysis on alcohol use in similar communities, and qualitative analysis of artefact interviews with YPRAs) to explore new ideas or alternative ways of seeing the problem and its potential solutions. YPRAs were then asked to imagine an ideal world where the problems they had identified did not exist and to think of potential interventions that could help them to achieve this ideal world. During this activity, the YPRAs created visual maps outlining the problems, their intervention ideas and intended outcomes (ideal-world scenario). These maps were then shared with the broader research team to refine the intervention pathways based on existing evidence.

The fourth and final phase involved introducing the YPRAs to an evidence-based intervention to reduce HIV and IPV. The theory of change process pointed to the need for a gender transformative intervention with varying degrees of livelihood creation, mental health support and broader structural change for each of the four groups (men vs women and urban vs rural). We used the Stepping Stones and Creating Futures intervention,36 which has been shown to effectively reduce men’s perpetration of violence in the South African context,37 as a starting point. We adapted the original intervention to incorporate broader mental health support and to be more aligned with the theory of change developed with the YPRAs. The YPRAs were closely involved in adapting the overall content of the interventions and the individual activities to suit their context and unique needs as identified in the theory of change process.

Patient and public engagement

The form of codevelopment discussed in this article (ie, codeveloping an intervention with potential end users) is a form of patient and public engagement that aims to engage the public in the research process itself. The ‘public’ in the Siyaphambili Youth project were the 17 YPRAs who were integrally involved in all four phases of the intervention development process as mentioned. In addition, YPRAs were involved in other aspects of the study, including recruitment of participants from their peer groups for in-depth qualitative research on topics such as rioting, emotional dysregulation and sexual life histories.

Building trust: involving potential end users as researchers

The direct involvement of YPRAs provided opportunities for understanding their everyday lives, which was one of the main objectives of the project. This was greatly facilitated by both the length of the project (eg, being able to build up trust over several years instead of hours or days), and the use of participatory methods to open up different conversational starting points.

The use of YPRAs’ photographs as a starting point for an in-depth interview (ie, the artefact interviews) provided an opportunity to explore complex social dynamics, including the meaning of relationships, dating violence and how food insecurity plays out in young people’s lives. As a specific example, several of the young women shared photos of pizza, and it was only through later discussions with the research team that they explained its importance as a status symbol in their lives. One woman in an urban setting explained a personal strategy of posting pictures of pizza bought by her friends’ boyfriend on social media so that her own boyfriend would also feel pressure to provide. Another took photos of yoghurt, which was also a representation of the expectation that love is expressed through men providing ‘luxury’ food items for their girlfriends. A third photo was of a group of young men and women at the beach, which for the young woman was evidence of how she had had to invite her female friends to join her because she was worried about being alone with her boyfriend and his male friends.

The use of photo elicitation for interviews created a conversational starting point for discussions with the research team about the YPRAs’ everyday lives. However, the YPRAs were still hesitant to discuss aspects of their lives that they thought the research team might not understand, like drug use or violence in their relationships. As paid employees, the YPRAs were concerned about how illegal and negative behaviours would impact on their involvement in the project. YPRAs were equally cautious of sharing personal information with one another, and it took months for them to openly discuss their personal lives in a group.

The challenges related to YPRAs’ willingness to share personal information were overcome to a certain extent through constant interaction and trust-building exercises carried out by a small group of expert facilitators on the research team. These expert facilitators used a range of techniques, including formal methods such as community mapping, character creation and photography, as well as informal techniques to create a non-judgemental and trusting space. Informally, the facilitators achieved a sense of comradery and group belonging through sharing examples from their own personal lives and pulling the YPRAs aside for private group ‘chats’ away from the rest of the research team and the recording equipment. In addition to the participatory methods, these informal techniques were instrumental in building trust with and between the YPRAs over time.

While there were interruptions in the research schedule as a result of COVID-19 lockdown measures, flooding and an attempted insurrection, these also provided a means of extending the project timeline, which gave the research team additional time to build trust with the YPRAs. After 2 years of regular meetings (one to two per month) between the research team and the YPRAs (for 1 year more than originally planned), trust has been established to some degree, and more details about the YPRAs lives have come to the surface. However, like any employer–employee relationship, there continues to be a hesitation by the YPRAs in sharing the intimate details of their lives.

Giving interpretive ownership over data

In trying to create a space for meaningful codevelopment, we selected methods to facilitate more open conversations about the YPRAs’ lived experiences, as well as strengthening capacity in understanding research and how it operates. The research team did this through introducing results from our own analyses during workshops with the YPRAs, particularly during the codevelopment of the theory of change. As part of this process, the research team explained key research concepts and discussed data interpretation. For example, differences between causation and correlation were raised in relation to COVID-19, with YPRAs discussing how 5G towers were said to be a key cause of the pandemic. The research team also introduced some key findings about the rates of violence in South Africa’s urban informal settlements.

These activities demonstrated that YPRAs understood the structural barriers that create disadvantage for themselves and their families, including barriers to engaging in South Africa’s democratic processes. They engaged with aspects of the data conceptually; for example, the rural YPRAs openly challenged how the research team had equated food insecurity and poverty in a previous study.38 The YPRAs asserted that poverty was not having the resources to lead an ideal life rather than just whether or not they went to bed hungry. They also openly disagreed with certain results, for example, they felt that the violence they had personally witnessed within their families was higher than that found in previous studies.

While the research team was excited about many of the insights the YPRAs brought to the table, there were also roadblocks in the YPRAs’ ability to engage in the research process. For example, the YPRAs found it challenging to develop fictional characters that were separate from themselves and their own stories, the fictional characters we had hoped would enable them to speak about issues without ‘exposing’ themselves to comments from others. Moreover, when asked to identify potential intervention ideas or strategies for achieving an ideal world, the young women from urban informal settlements in particular found it nearly impossible to come up with ideas. They had never imagined a life for themselves that was different from what they were currently living. While the coproduction process was aligned with YPRAs’ lived experiences in the case of urban men and the two rural groups, coproducing an intervention for the young women was challenging simply because they were unable to conceptualise potential solutions to the problems they were facing.

Embracing conflict: uncomfortable conversations

The YPRAs openly disagreed with the research team at several points in the project, and at times, this was challenging for the research team. For example, the young men living in the rural area openly condoned sexual harassment, including rape, if women dressed or acted in a particular way. Similarly, the young women living in urban areas wanted to stop abortion in their communities as a priority and perceived the problem as women not conducting themselves in ‘appropriate’ or ‘dignified’ ways. Both of these viewpoints made it difficult for the research team to come to terms with the YPRAs’ perspectives because it went against the evidence for preventing HIV and IPV risk and the research team’s own personal and political views. The young rural men’s assumptions of gender roles and the need to control women’s lives posed significant challenges for their involvement in an intervention aiming to prevent HIV and IPV.

The research team tried to overcome these conflicts during the codevelopment process by revisiting activities multiple times from different angles and with different questions. For example, when first asked to identify solutions to HIV and IPV risk, a common solution suggested by YPRAs was rehabilitation centres for alcoholics. It was only through further conversations that YPRAs were able to reflect more on their own lived experiences and to identify more practical and reflective ideas. In developing a theory of change for the intervention (phase III), the research team redid the activities multiple times to explore different approaches to the task. In the last rendition of these activities, rather than asking the YPRAs what they thought the problem and solutions were to HIV and IPV, the team posed questions about potential solutions to the problems the YPRAs had themselves already identified (eg, pressure from others to provide financially when they cannot) and how these were linked to more structural issues such as poverty. This helped to shift the conversation away from individualised interventions that were unlikely to work (eg, institutionalising alcoholics into rehabilitation programmes) to strategies that were both more reflective about the social structures young people face and more pragmatic about what can be done to address them, for example, providing mental health support, helping to build healthy relationships, strengthening communications skills and targeting small enterprise development.

Challenging research epistemologies

The conflict that arose during the coproduction process between YPRAs and the research team as described contributed to an increased reliance by the research team on their own knowledge of methodology and the evidence base as part of the coproduction process. This brought about many lengthy discussions about methods and epistemology (ie, our assumptions about what constitutes evidence in different research paradigms) by our multidisciplinarily team of social psychologists, public health researchers, economists, ethnographers, statisticians, qualitative researchers, intervention specialists and social epidemiologists.

In searching for the perfect method, these discussions often obscured attention to the components that were most important for the coproduction process to work as intended. For example, the research team had a lengthy discussion about the sample size required for data collected when data quality was likely far more impacted by the trust and ownership that YPRAs felt over the research process itself. The emphasis on methodology and evidence also made it difficult for the YPRAs to engage with the research team in the ways originally planned as part of the project. We had originally hoped that the YPRAs would join the monthly meetings with the broad research team. However, involving the YPRAs in what were essentially non-participatory online monthly meetings with a group of academics with mixed social, economic and racial identities would have further emphasised the social marginalisation that the YPRAs already felt in comparison to the research team. As the project progressed, we increasingly worried that their involvement in discussions about sample size and methodology would make the YPRAs feel less secure in the importance of their knowledge to the project and reify rather than break down existing relations of power. Instead, we structured meetings between the YPRAs and individual academics, which helped address these barriers and provided lively spaces of conversation but did not facilitate YPRAs’ full involvement in project activities.

One consequence of this division between the research team and the YPRAs was that these two groups often had different interpretations of the data and, therefore, different interpretations of what was most important to address as part of the intervention. For example, the young women living in urban informal settlements described how they often did laundry for their boyfriends and cooked them food. Some members of the research team interpreted this as an example of gendered social inequalities, whereby young women provide free labour to their male partners to allow them to in turn gain money through formal employment. This then led to the suggestion that interventions should address the gendered inequalities within relationships, have young women stand up to their boyfriends and assert their ability to support themselves through their own livelihood strategies. However, an alternative perspective, and the one that the young women themselves presented, is that the process of doing laundry for their boyfriends provided a means of expressing domesticity and love in a context where relationships are often fleeting, not monogamous and characterised by conflict. The practice of doing laundry was not perceived as exploitation but rather a means of overcoming the equally oppressive social expectation that their relationship would fail. In contrast to a gender transformative intervention, this led to the suggestion that women should be supported in nurturing their relationships as a means of providing stability and love in their everyday lives, rather than being forced into formal paid employment. This example illustrates how we sometimes get it wrong as researchers, even when working in close collaboration with potential end users on a topic of mutual interest.