While the four case studies presented used different methodological approaches to address diverse research aims among marginalised young people, we identified several common opportunities and challenges. These are outlined below to share observations and lessons to inform future studies.
Meet often to build trusting relationships
The combination of various methods (see Table 1) coupled with multiple personal contacts and repeated meetings with the participating young people and practitioners working with (or for) them, allowed researchers to raise awareness about the project and to build trusting relationships.
Frequent and ongoing interactions with marginalised young people were key to gaining their trust and engaging them in the research. In Cambodia, the prison system in itself creates mistrust of the authorities, and the researchers took extra efforts to overcome these barriers, such as taking senior prisoners into their confidence to facilitate the data collection process, a peer facilitated model . In India, repeated group meetings and workshops with the researchers and the young people allowed peer friendships to develop between participants and with the research team. With growing relationships, the participating young people started to engage in dialogue, which in combination with meeting in smaller groups (e.g. of three to five people), seemed to contribute to more open and authentic discussions. Similar experiences emerged in the Zambian context, where multiple interactions during the photo-taking stage led to young people feeling comfortable about discussing experiences and perceptions related to sex and contraceptive use. This sense of safety was also aided by the exclusion of teachers and other school staff from the meetings, an emphasis on confidentiality of information and by respecting individual opinions shared or discussed.
We identified that particularly for young people who were highly marginalised and hard to reach, it was important to build relationships with stakeholders who could be identified as gatekeepers. Experiences from the Cambodian, Zambian and Swedish case studies illustrated how, similar to Russell , a long-term approach might not only require frequent and ongoing interactions with the young people but necessitate repeatedly consulting various (adult) stakeholders before youth participants could even be identified or approached. In Cambodia, where the participating young people were severely disempowered due to their incarceration, the process of giving them voice and the opportunity to engage in the research meant that the prison directorate first had to be convinced of the value of the study. Building on existing collaborations and previous research , through multiple meetings with the administrators, who were initially sceptical and therefore reluctant to initiate research in the prison setting, the team eventually gained the approval and contacts necessary to approach the young people in prison. In the Zambian case study, similar experiences came across, where the research team was only allowed to proceed with PEI after the school authorities were convinced that the research would not teach ‘inappropriate’ lessons on sexuality to the young people. Young people were also hard to reach in the Swedish case study without the help of adult gatekeepers. Here, strategies such as several site visits and repeated personal contacts with stakeholders allowed the team not only to familiarise itself with the setting but to identify professionals who, in turn, could provide access to young people.
In all four cases, it was important not to rush or force the building of relationships. Instead, time and resources (both funding and skills) were needed, particularly to gain the trust of young people. This need may have been necessary because all four cases included young people who had experienced structural disadvantages linked to their place of residence, incarceration or socio-economic status, which may make their barriers to participation especially high . In addition, young people on the margins of society are often distrustful of authorities and reluctant to engage with institutions due to various oppressive structures, meaning that a focus on developing trusting relationships is key to promoting their participation in research .
Use creative and context-sensitive research methods
The four case studies pointed to the value of research that used creative and context-sensitive approaches. Specifically, they illustrate how practical, playful and peer-based methods that move beyond 'traditional' ones, such as semi-structured interviews, can be a pragmatic and ethical way of conducting research, especially in vulnerable communities [7, 9, 38].
In India, young people co-facilitated FGDs and worked as data collectors among peers in interactive workshops. Participants shared their experiences by telling stories or drawing pictures. Data collection in Zambia, in turn, involved young people being invited to take, select and discuss photographs to depict their perceptions, experiences and unique discourses on sexuality. In line with youth study scholars who have argued for the need to adopt task-based approaches that are youth-led, fun and informal [9, 15], the use of such methods appeared to enhance the young people’s enthusiasm for and engagement in the research. Innovative techniques comprising visual illustrations were also used to depict questionnaire responses from young prisoners in Cambodia, who were not only disempowered because of their incarceration but also had low literacy. While this context-sensitive approach addressed the needs of young people limited by their ability to comprehend and respond to written questions, the use of a software in the third phase of the research in Sweden was insufficient to engage rural young people in the research. This means that while public discourses often depict young people as ‘digital natives’ , the mere adaptation to online solutions was not sufficient to promote their participation.
When situated within a participatory framework, practical, playful and peer-based approaches can stimulate dialogue and a frank sharing of opinions that are not censured or interpreted by adults while also having the potential to reduce unequal power differentials [16, 40, 41]. In this regard, the peer-to-peer interviews conducted in India seemed to expand the participating young women’s worldview of more equal gender relations, leaving them feeing excited and enthused to negotiate for greater freedom of movement in their families. The use of photo elicitation in Zambia also allowed the researchers to move from discussing the pictures, which typically represented something familiar and fun, to more serious and sensitive questions and probes into sexuality. The benefits of using pictures in this way align with research that shows how the use of photos can help to reduce awkwardness experienced during traditional interviews [42, 43]. When conducting face-to-face fieldwork in Sweden, the researchers also disclosed aspects of themselves, using more ‘informal language’ in dialogue with the young people. Similar to Conolly , this approach was adopted to set the discussions off on a more equal footing and to allow for in-depth and sometimes sensitive questions to be asked about their lives.
While our case studies indicate that creative and context-sensitive methodologies can positively impact on the young people’s research experience while improving the outcomes of the research by making the process more equitable and purposeful, such approaches do not offer a solution per se to their marginalisation. In fact, transitions towards a more inclusive and equitable society will entail moving beyond a downstream focus on research participation in to an ‘upstream’ policy focus on reducing social, economic and health-related inequalities, where collective responses and actions for justice that seek broader societal change can contribute to youth development more generally .
Flatten typical power hierarchies
In addition to the opportunities and limitations of engaging in research with marginalised young people as presented above, this last section presents a number of challenges that constrained our attempts to create knowledge with young people for young people through collaborative forms of youth participation [11, 12].
Although co-learning whereby young people are involved already in the design of the project and then consistently throughout the research is at the heart of participatory youth research , in all our cases the decision-making was not shared between the (adult) researchers and (young) participants. Instead, following the standard routes of research and constraints imposed by traditional academic systems and structures [8, 11], the study protocols were developed by the researchers alone or in collaboration with stakeholders. This meant that adults and academics framed the research questions, chose the methods and controlled the analysis. In Sweden, young people’s hesitancy to participate in the face-to-face workshop and online questionnaire might reflect a resistance to the adultism  of adult- and academia-centric approaches . In India, where the research was much more collaborative and ‘youth-centred’, it is more likely that the problems related to engaging young people in later stages of the research stemmed from difficulties in sustaining trusting relations to foster authentic youth–adult partnerships. The regimental structures and ambience of the Cambodian prison system, in which the young people were severely disempowered, limited the opportunities for collaborative participation and also acted as barriers to the participants fully expressing themselves . Considering the value and possibilities of approaching young people directly when the research is conducted in a non-institutional context (see e.g. ), the Swedish case study could also have benefitted from moving beyond a standardised recruitment process involving gatekeeping to a more informal one drawing upon local youth networks to more effectively build rapport, develop trust and flatten power relations.
Against this backdrop, pointing to the constraints that shaped our attempts to conduct participatory youth research, we agree with Rodríguez and Brown that the challenge for marginalised young people in shaping and influencing the process ‘hinges not on a lack of voice but a lack of power’ (, p. 32). While the approaches employed in the case studies may have moved us, at least partly, away from problematic notions of ‘helping’ young people, our sincere attempts to give voice to marginalised young people have most likely not resulted in immediate emancipation and empowerment . However, following the discussions of both Fox  and Teixeira et al. , a first step towards developing youth–adult partnerships that have the potential to challenge oppressive systems in academia and society at large is critical self-reflection. As scholars interested in doing research with young people that is transformatory and truly grounded in their lives and lived experiences , we recognise the concurrent and future need to address the adultism shaping our work through continuous discussions of how young people’s involvement may be constrained by structures or simply reflect our assumptions about what is appropriate and possible [8, 11].