Negotiating Leadership in Interdisciplinary Co-Productive Research: A Case Study of An International Community-Based Project Between Collaborators From South Africa and the United Kingdom

In the absence of empirical and conceptual considerations of the negotiation of leadership in teams doing community-based research, this article adds to the leadership literature by offering a critical reflection on positioning and collaborative teams in the context of one interdisciplinary, co-productive, cross-generational and international research project. The project focused on youth and community resilience to drought in South Africa. Fourteen co-researchers reflected on their experiences of leadership within the project, using a collectively developed questionnaire. Findings uniquely highlight wider ethical considerations when youth and novice researchers are included in research teams. A strong emphasis on cultural responsiveness was found; with local and culturally led leadership seen to positively influence both processes and outcomes. Reflections suggest collaboration may be approached as an “ethos” and aided by transformational leadership theories and methodologies. Findings may be especially relevant to research teams, funders, and ethical bodies.


Introduction
The impetus for the case being reported in this article (i.e., the empirical review of leadership), arose from a desire to reflect on and improve the leadership practices within research teams. In this article, 14 diverse members of an international, intergenerational, and interdisciplinary team reflect on their coproductive experience of leadership in a research project. Team reflections are contextualised in relation to contemporary research leadership debates, identifying nine issues of pertinence to research team leadership. This article concludes by suggesting how research team leadership in such complex settings can best be supported.
Globally, there is an increased push from policy and funding bodies toward interdisciplinary and cross-cultural research (Clark et al., 2017;Owusu et al., 2017;Sugden & energy scarcity, or the ever-changing schemas of the urbanrural continuum, all require collaborative research positions (Gross, 2012;Rockström et al., 2009). It is thus increasingly common to see nonacademics included as full-time members of research teams. These can include education, health, and social care practitioners, as well as wider community members (Jean-Louis & Lomas, 2003). Reflections on the complexities involved in such research are now commonplace in the literature (Abma et al., 2009;Garnett et al., 2009;Telford & Faulkner, 2004;Trivedi & Wykes, 2002). However, less attention has been paid to the inclusion and contribution of young and/or less-experienced researchers in teams (Åkerström & Brunnberg, 2013;Liebenberg et al., 2017), with a noticeable gap in the literature whereby all these complexities are considered. Thus, research leadership guidance in the context of research team diversity is fragmented, limited, and potentially incomplete.
Hence, this article offers the first reflective and empirical account of research leadership in the context of a co-produced, international, intergenerational, and interdisciplinary research project. The empirical element of the article focuses on a project that explored youth and community resilience to drought in Leandra, a drought-challenged area in South Africa. The research team (N = 81) included academic researchers and students from the United Kingdom, South Africa and Canada, working with community organisations, project managers, and novice youth co-researchers from the United Kingdom and South Africa. Most team members had not met prior to the project's initiation. Everybody who was trained in data collection was identified as a researcher (e.g., academics) and we used the term "co-researcher" to describe community partner and student team members throughout. The youth members of the team were identified as youth coresearchers. Academics were referred to in the study as "lead researchers" as they all had previous research experience and led the development of the training input.
This article reports on the reflections by researchers and co-researchers on the leadership role they played within a wider study on the resilience of youth living in a drought affected context. Within the wider study, 50 young (15-to 24-year-old) South African youth co-researchers were recruited and supported by a local community organisation, 43 of whom participated in the research. These youth coresearchers were trained in research methods, including artsbased data collection approaches (see www.brighton.ac.uk/ leandra for more details). Training was undertaken by 15 Masters and Honours student co-researchers from the University of Pretoria in South Africa and three young British co-researchers. The latter, British co-researchers had personal experience of overcoming complex life challenges, and of undertaking research on resilience (themselves recruited through a UK community organisation, Boingboing Resilience CIC). All co-researchers were supervised by academics from South Africa and the United Kingdom. The South African youth co-researchers contributed their personal perspectives on resilience to drought and investigated the perspectives of elders in their community. They were subsequently supported in communicating their findings to policy makers, through the production of a short film and a richly illustrated policy briefing document (www.brighton. ac.uk/leandra).
The academics most involved came from five universities (three in South Africa and two in the United Kingdom) and from disciplines spanning education, social science, health science, visual arts, theatre studies, physical geography, and climate science. Funding was secured from November 2016 to July 2017 from a UK research council that permitted 30% of the total budget to be shared with partners outside the United Kingdom. Some of the South African and UK academics had funded research time allocated. All community partners (including youth co-researchers and project managers from third-sector groups) were recompensed for their contributions. Ethical constraints resulted in having to "pay" youth co-researchers with gift vouchers rather than cash. By nature of the funding requirements, the lead investigator from the UK team had ultimate responsibility for project management and accountability. However, in operational terms, the research design, management, and accountability practices were shared.

Research Method
To reflect on leadership practices, a novel approach to research was adopted whereby reflections and literature were considered in tandem. When complex and perhaps fragmented subject areas are to be considered, Arksey and O'Malley's (2005) scoping search methodology suggests literature reviews include consultation exercises. Following this approach, team members were asked to supply literature relevant to informing their practices. At the same time, they were sent an electronic questionnaire. The questionnaire was collectively designed by the team to consider reflective practice. The scoping review and survey occurred concurrently. Findings on leadership themes that resulted from the scoping review were used for deductive analysis of the survey data.
Co-production of knowledge is synonymous with uneven spaces of power. Reflexivity can go some way to counter existing power dynamics. Not only is reflexivity well established as a strategy to enhance dependability of evidence (Guba & Lincoln, 1981;Richardson & Adams St. Pierre, 2005;Seale, 1999), it is also suitable in research spaces where the crisis of representation (Fine et al., 2000) comes in to play and where democratising research is an objective (Vaillancourt, 2006). Reflexivity assists researchers to reduce claims to authorial presence. However, it remains that, whomever scripts the results of analysis retains power to sort, classify, and order how the many voices present in the data are presented. In this study, co-researcher reflexivity was gauged using questionnaires and assisted in auditing how leadership positioning unfolded in a research team. Investigating reflective accounts enabled systematic capturing of relational understandings of the self, others, and the context (Chambers, 2010).
Fourteen team members (including researchers, youth coresearchers, and co-researchers) completed the questionnaire (N = 14), with roughly half providing literature for consideration. Demographically, participants comprised nine women and five men, of whom seven were academics (researchers), three project managers (co-researchers), three youth coresearchers, and one a community artist; seven were from the United Kingdom and seven from South Africa. All 14 were involved in various leadership roles in the wider study. For the purpose of this article, it was important to understand their perceptions of leadership and what informs those roles. However, only half provided literature that informed their leadership style.
The process of questionnaire development and completion highlighted several barriers to inclusive and meaningful involvement. Some were related to conflicting demands and lack of capacity for follow up. For example, youth co-researchers living in Leandra were not involved, due to a lack of resources on the part of the relevant lead researchers to followup with them face to face beyond the initial project timeline. Other barriers were more systemic and highlighted inequalities across the project's youth team. Virtual participation, an option largely taken for granted in the United Kingdom, was not possible in Leandra. Not only is the supply of electricity in South Africa not stable as a whole, and specifically in more rural spaces, but youth co-researchers generally had no funds for data, nor access to email or other internet-mediated platforms. They were, however, included in the co-production of a co-produced policy briefing paper and film (www.brighton. ac.uk/leandra). The South African student co-researchers were also excluded, due to academic restrictions privileging exam and dissertation requirements above that of research at the time of the survey. In retrospect, it may have been pertinent to include the students in at least part of this reflective process (e.g., in the completion of the questionnaire). However, to have done so would have limited their contributions to that of "participant," rather than "co-researcher", which was contradictory to the wider study's methodological underpinnings and inconsistent with ethical agreements. To go some way toward including student voice (although regrettably not one from South Africa), a UK-based doctoral student, who was not a team member, contributed to the article by sourcing literature and taking part in the write up and analysis. An important learning from this is that research design and funding applications need to include sufficient time and resource allocations post data collection to enable collective researcher reflexivity.
The questionnaire covered the following issues: 1. Whether team members saw themselves as leaders in this project. 2. Any critical incidences that happened on this project in relation to leadership that illustrate the kind of leadership influences at play. 3. Any critical incidences that happened on this project in relation to individual leadership of other people that illustrate the kind of leadership influences at play.

Project leadership in relation to intended outputs and outcomes.
Each person independently completed the questionnaire electronically and returned it to the lead author along with recommended literature. This process, along with subsequent hand searching, identified contextually relevant literature that considered leadership in regard to research teams, community-university partnerships, engaging youth co-researchers, and exploring cross-cultural and interdisciplinary research practices. A review of the literature was undertaken prior to analysing responses from questionnaires. Nine themes were identified within the literature. An overview of these themes will be given in the following section, after they are briefly contextualised in relation to the literature.
Responses from questionnaires were coded by the lead author using Thematic Analysis (Braun & Clarke, 2006;Terry et al., 2017). It began with a deductive analysis based on the nine themes extracted from the literature (see "Literature Review" section). This deductive analysis was reviewed by a second team member, after which discrepancies were discussed. As per Nieuwenhuis (2007), this was followed by an inductive process via wider team member discussions. Here, the nuanced differences between literature and team responses were collectively considered. The findings from the analysis were structured using the themes identified in the literature, followed by collectively drawn conclusions and recommendations.

Contemporary Debates in Research Leadership and Collaborative Team Working
The issue of leadership in conventional academic research teams has been widely addressed in the literature, both in the Global North (Hanganu-Opatz et al., 2015;Omar & Ahmad, 2014) and South (Ferreira et al., 2015;Steelman et al., 2015). Empirical studies of leadership in conventional, hierarchical academic teams, are concerned primarily with leadership behaviours by those deemed to be "in charge" of the team (Kolb, 1992). Here, leadership is often discussed in relation to responsibility. For example, Hemlin and Olsson (2011) considered critical incident responses from interviews with members of research teams in academia and industry. They identified leadership behaviours such as coordinating processes, assigning tasks, motivating members and coordinating rewards, promoting group cohesion, and liaising with external contacts. They argued that leadership in academia additionally involved providing "emotional and cognitive support" to developing researchers (Hemlin and Olsson 2011, p. 52). Compassionate leadership theories may be especially relevant here.
Compassionate leadership theory is promoted by the UK National Health Service and necessitates a greater focus on mutual and reciprocal care between team members (West et al., 2017). It appears to build on transformational leadership theories more frequently mentioned in relation to cross-disciplinary teams. Considering research within the field of radiography, Lee et al. (2004) propose transformational leadership as a potential model to improve clinical innovation (p. 72). By "transformational," the implication is that a greater degree of ownership by other team members of both process and outcome should lead to faster changes (Clark et al., 2017). Thus, transformational and compassionate theories of leadership appear relevant to guiding interdisciplinary teams. Yet their relevance in guiding teams that consider wider researcher identities (i.e., diversity in relation to culture, community, generational knowledge, level of expertise within a subject area, etc.) appears untested.
A further level of complexity occurs where research teams incorporate community members, as is often the case in studies on global challenges that require complex solutions. It is increasingly common to see practitioners, patients, and representatives from third-sector groups as full members of research teams (Jean-Louis & Lomas, 2003). This is often to increase research impact and improve service provision (Allsop et al., 2004;Boyle & Harris, 2009;Clark et al., 2017), while also promoting innovation across disciplinary boundaries (West et al., 2017). Reflections on the complexities involved in such research are now commonplace in the literature (Abma et al., 2009;Garnett et al., 2009;Telford & Faulkner, 2004;Trivedi & Wykes, 2002). Recurrent themes include power, pay, technical expertise, training and development resourcing, acknowledgment of contributions, the importance of building trust, relational agency, and ethical considerations. Much of this literature has focused on team dynamics. However, leadership of complex co-productive research teams involving community members has received significantly less attention (Bolden et al., 2008;Brookman-Frazee et al., 2012;Hart & Church, 2011;Overton & Burkhardt, 1999).
Globally, one of the most common reflections on collaborative research initiated by academics is that leadership practices must provide opportunities for academics to share control over the research processes with community members (Trivedi & Wykes, 2002). This is certainly the case in South Africa, where the use of collaborative approaches is common in studies that need to accommodate diversity and associated power dynamics (De Lange et al., 2007;Ferreira & Ebersöhn, 2012;Ferreira et al., 2015;Theron, 2016). However, the level of control over the research process and outputs is not simply determined by methodological concerns, and could present itself differently at different stages of the research process. For example, funding guidelines might dictate where research accountability lies, and limit the level of financial control that other members of the team have at planning and operational levels as a result. One prominent theme in the literature regarding teams that cross academic and community settings is just how difficult they are to manage and support. An example is from a study by Machimana et al. (2018), of the experiences of multiple higher community engagement partners on a decade-long partnership. They found that a common hindrance related to relationships and communication (i.e., unclear scope, inconsistent feedback and conflicting expectations).
With regard to cross-cultural issues, Elkins and Keller (2003) reviewed literature associated with leadership in research and development teams, where members often come from diverse disciplinary backgrounds. They argued for a more transformational approach to research leadership, with additional tasks and behavioural activities relating to the improvement of relationships between members and the beyond-team environment (Elkins and Keller, 2003, p. 587). It is unclear from their study whether the literature they considered included teams comprising community members. However, they reported that "leader roles were more important when subordinates (sic) were disadvantaged in some way" (Elkins and Keller, 2003, p. 594).
Diverse disciplinary backgrounds also provide meaningful leadership and research insights. In relation to youth specifically, a handful of research projects have explored creative ways to include youth in the development of research innovation (see, for example, Åkerström & Brunnberg, 2013, discussing youth research circles; Liebenberg et al., 2017, discussing youth in data and dissemination practices). These highlight the need for youth to have flexibility of control over their participation at various stages of the research process.
The review of these contemporary debates suggests nine leadership themes in interdisciplinary, cross-cultural research teams: 1. Enables shared leadership (including shared decision-making). 2. Supports the flexibility for team members to lead on different aspects of the research, or at different time points. 3. Aims to produce mutually beneficial knowledge. 4. Promotes knowledge-sharing across cultures/disciplines and organisational boundaries. 5. Identifies an overall leader for the purposes of accountability. 6. Encourages innovation and creativity. 7. Considers relational and emotional needs of team members. 8. Supports the inclusion and leadership development of "newcomers" to the research arena. 9. Considers ethical accountability in terms of both financial and knowledge exchange processes.
These themes were not used to develop the questionnaire. Rather, the nine leadership themes were used as an a priori framework to guide deductive analysis of the questionnaire data.

Case Study Results of Researcher Reflections
This section considers reflections from the research team in relation to the nine leadership components extracted from the literature. In instances where the reflections supporting a theme are insubstantial, this paucity is noted and theorised. This is in keeping with the spirit of deductive analyses (Nieuwenhuis, 2007).

Theme 1: Shared Leadership
Shared-leadership responsibilities were evident in written researcher reflections from our project. The majority of responses suggested that team members considered themselves as "co-leaders" with regard to conceptualisation, data collection and interpretation, and dissemination processes. One South African academic, for example, observed: I do see myself as co-responsible to generate solutions, and think of and implement strategies when there are hiccups in the field . . . keeping up good spirits and assisting in discourses that support movement towards trust, solutions and shared aspirations.
However, reflections additionally suggested this represents a novel way of working for research teams whose members are geographically and culturally diverse. The South African community artist was pleased that . . . despite some trepidation, researchers stepped-back and let others occupy principal positions in research processes and activities-at least from my perspective and in my experience. This is a structure of rotating leadership that many in activist fields strive for, so that no one person always carries a load, even when they hold a position of "principal" throughout.
There were also differences between team members, according to their specific discipline and their perceived distance from the "core" of the interdisciplinary project. A UK academic, for example, made the following observation: While I see myself as a co-leader on the project, in that I am involved in key decision-making, my real leadership is primarily within the area of the project for which I am responsible. This is very much in constructing the "climatic backdrop" for the research, rather than as a core player in the community-based research by young people.
Collectively, responses suggest that shared leadership was promoted in the context of this project by what could be described as attempts toward an intrinsically embedded nonhierarchical ethos. Unsurprisingly, this aspiration was not always achieved. For example, when a co-researcher took on a facilitation role without clearing it with the lead researchers. The data does illustrates a mutual respect for individual expertise. Here, expertise appears to relate equally to disciplinary/role specific knowledge as well as culturally and contextually based knowledge.

Theme 2: The Flexibility for Team Members to Lead on Different Aspects of the Research or at Different Time Points
In addition to the final comment above, team reflections suggested that many project members emerged as leaders in relation to specific research activities at various stages of the research. Six reflections illustrate this: I see myself as an academic leader in the project more so because I have 7 students writing their mini-dissertation under this project. (South African academic) I am a leader in this project because I worked with the research team-when they came to my province, I helped them to get the right people in the programme. (South African community partner and project manager) I respect the fact that when a researcher takes the lead in conceptualising and coordinating the writing they can take the leading position in authorship. (South African academic) The UK community partner and project manager, for example, appreciated that he . . . felt able to take charge around some of the detail of ensuring that co-researchers were properly recompensed for their involvement.
The South African community artist also reflected: I lead a process of creative expression and as such, facilitated a creative space, both in the field and post-performance, crossing over into film documentation. This is a shared space and not expressed by a top-down, give and receive orders, hierarchical type of system of leadership but a mediated process of back-andforth between director, co-directors, other research participantsnoting that all of us occupy the role of researcher in various, intersecting ways . . . Finally, a UK youth co-researcher explained her position: Whilst the academics took the lead on coding and analysis, because this is a technical skill partly, it was very appropriate that a number of more junior team members were involved in it, including community partners. That way we really are learning about all aspects of the research process and can lead on more things next time.
Responses suggest that a flexible approach to leadership was applied to both the conduct of research processes and the development of research outputs. This appears to be enabled via the aspirations toward a nonhierarchical leadership structure. Thus, shared leadership and flexibility appear interrelated. However, unique to these reflections, local individuals (in our case South African nationals) appear to be the most appropriate leaders for location-specific tasks. This is because there may be specific customs and cultural practices to consider, which, if ignored, could potentially be detrimental to the research process as well as to its outputs.

Theme 3: Mutually Beneficial Knowledge
As discussed earlier, contemporary debates suggest that academic institutions have a responsibility to support mutually beneficial knowledge production. Researchers working on this project appeared to embed this responsibility, as exemplified below: An important part of my role has been to try to bring people together-as much virtually as face to face-across disciplines and to make spaces for us all to understand, appreciate and draw on each other's contributions to the project. We've been aiming for an interdisciplinary project rather than simply a multidisciplinary project. (UK academic and project lead) Central to me is maintaining the integrity, the ethos of how and why we want to generate knowledge on resilience WHILE we address hiccups as we generate this knowledge. (South African academic) I appreciate the triad of transparency, scientific focus and rigour, and social responsibility enmeshed in the vision for outputs and outcomes. (South African academic) Considering Themes 1 to 3 collectively, and in relation to the literature, it appears that "transformational leadership" theories (see, for example Lee et al., 2004) may be relevant.

Theme 4: Promotes Knowledge-Sharing Across Cultures/Disciplines and Organisational Boundaries
Team reflections echoed discussions that suggest having clear role-descriptors has the potential to improve cross-cultural working and acknowledged that the ability to lead could be linked directly to access to insider knowledge. The thoughts of a UK youth co-researcher encountering a new culture were particularly pertinent in this regard: When working in a different cultural space, it is important for me to remember that different leadership styles will be employed. It was an interesting transition coming from the UK . . . If at all possible, having early clarity around roles is very helpful . . . Because I did not understand the social etiquette, I left feeling a bit slighted, as did South African counterparts, because we interpreted each other's actions differently from how they were intended.
Another UK youth co-researcher observed: Leadership and local knowledge have a strong association. One particular moment, which sticks out for me is when I suggested that the co-researchers could attempt a little more eye contact with elders so that the elders know they are being listened to and respected. In the UK, strong, yet non-threatening eye contact is always appreciated. They informed me that strong eye contact has the opposing meaning of disrespect in their culture. This was clearly a crucial learning point about leadership for this co-researcher, who added: . . . it is important that people on home turf get to lead, because they have a direct and thorough understanding of cultural, political and social dynamics and customs. This perspective was echoed by a South African academic: In many ways, I think the SA team has a leadership role in the sense that we are familiar with the SA context and its challenges and resources and so have an innate sense of what will work/not [but of course this familiarity has its limitations too].
There were particular challenges about working between disciplines. The South African community artist felt this keenly, noting that there were . . . glitches between understanding one another's disciplines in the face of formal means of measuring experience and qualifications to lead and "hold" a research space . . . The miscommunications-saying the same things but in different ways-were very evident but resolved in the sigh and sound of an elongated "Oh, I understand." This was not simply a challenge of an ability to lead . . . but rather a type of mistrust and/or uncertainty on what leadership and facilitation might mean from an arts and theatre perspective.
A UK academic reflected on the way they dealt with this challenge: I am very much used to working in an interdisciplinary context. For me, the most important thing in any project is to step back, respect one another's specialism, acknowledge that I am not an all-round expert, and allow people who are leaders in their field to lead.
Researcher reflexivity, and transparent and open communication appear to be key. Cultural responsiveness emerges as essential in regard to producing relevant outputs. Locationbased leadership and high respect for insider knowledge appears paramount within research processes. A lack of cultural understanding may be especially unnerving for youth co-researchers from different countries, who have the added responsibility of learning research and cultural practices simultaneously.

Theme 5: Identifies a Leader for the Purposes of Accountability
Contemporary research practices often define the principal investigator as the identified leader within a research team.
Researcher reflections supported this and suggested the need for an identified leader to consider the wider picture of the research process. In our case, unsurprisingly, a number of more experienced academics showed themselves to be constantly mindful of the bigger picture, but less experienced team members also played their part in learning to think strategically about the project. As one youth co-researcher from the United Kingdom stated: During the first day, I was assigned a role to take notes/ reflections on the activities. . . I felt that it would be more appropriate for me to engage in a co-facilitator role. . . this impinged on the designated facilitator, and could have undermined the programme of activities as a whole. I learnt from this to remember to clear any changes with project leaders.
Reflections also added new insights. First, they suggest that a leadership hierarchy has the capacity to be helpful in instances where the identified leader cannot be contacted: There is a hierarchy in place, if I can't get hold of [the lead researcher] I contact [one of the other academic researchers] and they can either make the decision or will help to get the lead researcher to address the issue. (South African community partner) Second, in contradiction to literature that suggests the principal investigator is the only person with financial accountability in the research process, our project revealed a more complex picture. Researcher reflections suggested this accountability was subdivided between members of the team and further supported by various members of their organisations: I am responsible for large parts of the SA budget and so that would be a form of financial leadership, I suppose. (South African academic) I knew the size of my section of the project budget before I travelled to South Africa, and took responsibility to ensure that every item of expenditure was accounted for. (UK academic) Responses suggest that devolved budgeting may have been helpful in enabling shared leadership and that leadership hierarchy may be helpful at times in this regard. However, it is further suggested that identifying others who can be relied upon when one team member is not available was helpful for accountability purposes.

Theme 6: Encourages Innovation and Creativity
Team reflections did not touch on the subject of innovation in an explicit fashion. Evidence of encouraging creativity, however, was more explicitly discussed: There is the role of steering the ship and producing a creative process . . . there is the facilitator role in which I mediate how the youth co-researchers embody, speak of and relate characteristics of resilience, and work to produce creative storytellings of those resilient characteristics. (South African artist) It could be suggested that all research projects aim to develop innovation in terms of knowledge production, and thus, no direct questions regarding innovation were asked. This, along with the implied purpose of research, perhaps explains why reflections were limited around the concept of innovation. Due to the prevalence of this theme within the literature, future studies may wish to explicitly consider this dimension.

Theme 7: Considers Relational and Emotional Needs of Team Members
It has been suggested in the literature that leadership additionally includes pastoral elements. Team reflections (academics, project managers, and young people) highlighted that team members provided emotional support to one another and to themselves. Reflections from community and youth researchers in particular highlighted that, at times, they felt "stressed" and "uncomfortable." This is exemplified by the view of one UK youth co-researcher: It must be said that co-production can be messy. Self-leadership is important to ensure that you can be an effective follower. It sounds simple, but trying to ensure that you get enough sleep is really important. Stress-levels can rise due to many factors, and trying to cope with that on a lack of sleep can lead to more stress. I really wanted to do a good job and was so excited to be a part of this project that I sometimes was unable to sleep, which led to more stress overall.
Both academics and project managers were acutely aware of the need to provide emotional support. The UK project manager reflected: Some of our co-researchers were very anxious about travelling to such a different cultural context and we needed to give people a lot of support around practical issues like packing. At times, I wondered whether we had done the right thing involving so many people (including myself) with limited experience of this kind of work.
A South African academic shared a similar experience in relation to students: Incidences of my leadership as an academic promoter were evident . . . I found myself dealing with emotions of students, I had to make them focus on the bigger project and to trust the process. I also make them aware that in most cases, "no data" is a significant contribution. I have to focus them on the bigger project and not minimise their significant role in the project.
As evidenced in the literature, the requirement to provide emotional and relational support has the potential to increase when working with students and researchers new to their roles. Linking to Theme 4, relational support and cultural competencies appear interrelated.

Theme 8: Supports the Inclusion and Leadership Development of "Newcomers" to the Research Arena
Co-researcher reflections suggested that they did not feel like leaders within the project at first, rather that they were developing leaders: No, I do not see myself as a leader in this project. This is because I'm not an academic member of staff that is directly publishing academic outputs . . . this project is a great method of practical learning about how to be an effective follower, which I see as the first step to becoming a leader on projects like these in the future. (UK youth co-researcher) Toward the end of the data collection phase, however, youth researchers had become more aware of their own potential leadership roles. One, for example, suggested that she had played a role in championing the perspectives of other youth researchers. Team reflections from both academics and community partners corroborated this, suggesting that less experienced researchers required support in developing leadership roles: This process of having postgraduate students made me aware that the success of the project is dependent on the collaboration and focus needed to ensure that everyone has a significant role to play. (South African academic) As supervisor of a number of the Masters and Honours students, I have at times felt the need to champion their concerns (e.g., that we do not expect more from them in terms of data transcription/analyses than would be fair). (South African academic) After all the hard work involved, a South African community partner and project manager was excited that he . . . saw a lot of development in terms of leadership of the youth participating in the programme, of becoming researchers themselves. That gave them a role. It made them realise what qualities they possess as individuals . . . participating in their community to conduct the research themselves. That for me was development, that was empowerment and I think that influenced their leadership role in relation to this project.
Linking to the finding that youth co-researchers required heightened emotional and relational support within crosscultural projects, it may be recommended that additional resources be allocated for developmental activities. This feeds into wider ethical considerations.

Theme 9: Considers Ethical Accountability in Terms of Financial and Knowledge Transfer Processes
As identified in contemporary debates, team reflections suggested the need for further discussions around research integrity and ethics, suggesting ethical considerations are interwoven with financial and knowledge exchange accountability processes. One South African academic reflected that she was uncomfortable with agreeing any actions over and above those explicitly stated in the ethics application approved by her own institution. Furthermore, making any amendments to the ethics application was not possible within the operational timeframe. The UK project lead understood these limitations yet lamented: It is almost impossible to pre-empt everything that needs to happen in a co-researched project like this to achieve full inclusion of all co-researchers. I worry that our ethics procedures precluded some South African co-researchers getting as involved as they might have liked. With hindsight, we should have included an option for them to join in with wider dissemination, including writing this paper, although our budget would have been hard-pressed to manage this.
Thus overall, reflections from the team suggest that anticipating resourcing needs was challenging, and equitable and equal contributions were not consistently achievable. Yet, they do point to a need for flexibility in regard to research design, processes and resourcing. A lack of resources appeared to limit leadership and development opportunities for involved youth co-researchers. Indeed, to some extent, this issue was a tension in the project throughout. For example, members of the project team debated at length the risks and benefits related to cost, practicality, ethics, and research output delivery of youth co-researchers from the research site traveling to Pretoria to take part in a 2-day research event. Further tensions arose because our ethics application had not included the facility to return original drawings to youth co-researchers, and some team members felt strongly that youth co-researchers should all receive their drawings back rather than them being kept as research data. Research plans and ethical applications may require scope for flexibility. These findings may be of particular interest to research funders when allocating resources and in debates relating to how ethics procedures and processes are engaged with for co-produced research.

Discussion
Despite the lack of research on leadership in research teams involving nonacademics, many of the issues raised in conventional research teams overlap with the nine themes identified in the literature. The analysis of our researcher reflections lends weight to the need for transformational forms of leadership in cross-cultural, intergenerational, and interdisciplinary research teams. Conventional hierarchical leadership, which is more typical of academic research, is not the way forward here. Given the diversity of identities and needs, initiating a discussion with research team members on how they see the research as "helpful" to them and to others, can aid in identifying a shared vision and opportunities where leadership roles and responsibilities can be shared. Findings suggest that a recognition of local and cultural competencies may aid in determining who leads on what part of the project.
Leadership is a critical issue in co-productive projects, given the power differentials and ethical issues that can arise, particularly when researchers come from different countries whereby historical tensions have existed. The need to consider historical and structural factors (e.g., past and present colonisation), and their influence on research teams within, and beyond, higher education, is also important from an inequalities perspective (Hardy et al., 2016;Hart & Church, 2011;Townsend et al., 2015). These might require a shift in thinking toward the adoption of methodologies and processes that seek to widen participation and build capacity. Providing structures and resources to engage community members (including youth) in commissioning and funding research at various time points is also important. Other essential factors for consideration include the conceptualisation of boundary-spanning accountability structures, the provision of tools and technologies to facilitate collaborations, and the consideration of how best to acknowledge community contributors in outputs.
Understanding what team members see as rewards enables self-leadership and is more associated with empowering rather than transformational management approaches (Pearce et al., 2003, p. 300), but is nevertheless relevant to this article. Furthermore, as Hemlin and Olsson (2011) argue, identifying and providing rewards and positive feedback to team members is one way in which research leaders can increase group cohesion and support the innovation process (p. 56). Reflecting back on our findings, we suggest that these kinds of reward structures can be particularly helpful for those new to research. Considering researcher development, Evans (2012) argues that research leaders should not underestimate, "the importance of reaching and winning over hearts and minds" (p. 432).
Most importantly, as was found in the literature and in our team reflections, emotionality cannot be separated from the research process. The identification of such human dimensions further highlights the need for researcher reflexivity to be openly discussed within collaborative teams. For many team members-both academic and nonacademic-such reflexivity may not be an embedded aspect of their practice, and senior research team members might not be skilled facilitators of such processes. We suggest that the nine themes we have considered in relation to our own research might be useful to initiate such conversations.
As we have seen in both the literature and in our own practice, reflection is especially critical for researchers working within new cultural contexts. Without it, mutually beneficial knowledge is unlikely to be developed and shared. Ozano and Khatri (2017) argue that shared reflexivity aids in deconstructing embedded factors which may play a pivotal role in research, in addition to understanding positionality and power (p. 13). Therefore, it is important that leaders first understand the subjective context of individual team members, the community under investigation, and the wider environment, before they can lead effectively. This necessitates reflective processes to be collective, supportive, and include the wider community. Edwards' (2011) framework for contextualising distributed leadership may provide further guidance in this area, as might Hart & Aumann's (2007) model of collaboration between community partners and academics. In our study, as raised above in our reflections, the UK researchers held a cultural awareness training session. One of our youth co-researchers led on this using the "Political, Economic, Social and Technological" (PEST) model. For more details, including the challenges of this, a blog written by the youth facilitator is available at: "www.boingboing. org.uk/cultural-awareness-training-blog." The need for open discussions surrounding leadership also pertains to various tasks, some of which might require different leadership approaches to facilitate their success. One practical suggestion to improve collaborative research processes is developing clear role-descriptors for community members and others working with academics in research teams. This may identify opportunities for leaders to emerge in relation to delegated tasks. Abma et al. (2009) use roledescriptions informed by Arnstein's Ladder of Participation (Arnstein, 1969) to express the level of control service-users have in research teams; with categories such as "patient advisors," "patient interviewers/moderators," "patient researchers," and "patient research principles" (p. 404). Role distinctions have also been outlined by a number of methods, which seek to provide a platform to include nonacademics in research activities led by academics. Such methods include practice-based resources identified by community members in literature reviews (see, for example, Arksey & O'Malley, 2005, scoping study methodology), involving community members in the overseeing of research activities via communities of practice (Wenger & Snyder, 2000), and including them in data collection activities as peer researchers (Davey & Goudie, 2009;Garnett et al., 2009;Iseke & Moore, 2011).
Whether, and if so how, team members are recompensed in these complex teams is also an issue that needs diligent attention from the project's outset. Brookman-Frazee et al. (2012) reported that members of a research collaboration donated their time in the absence of funding. For academics this is a familiar story. Indeed, in our project, the principal investigator's time was offered as match funding, so that scarce resources could be shared more widely, and other academics worked way beyond their time allocation. However, community organisations, community members and junior academics might not have these in-kind resources to share, which is an important issue in terms of ethical accountability. Therefore, it could be suggested that a team member's lack of financial power can result in an inability to exercise their knowledge power within a research partnership.
In our project, we did ensure that community partners were recompensed and that student participation was directly relevant to their courses. However, regrettably, remuneration for community partners was restricted to the data collection and early policy development phases, due to constraints on project duration imposed by the funding body. In practice, this meant that at least one employee of the UK community organisation effectively contributed 12 days of unpaid time in the dissemination phase of the project, since he continued to work on both the policy briefing paper and later stages of the film production. In relation to project planning in such scenarios, it should be acknowledged that the salaries of community sector workers are invariably less than those of senior academics. Hence, any additional unpaid work potentially comes at a proportionately higher personal and/or organisational cost.
Leadership approaches to promote the inclusion of youth and new researchers in the research process are virtually nonexistent in reviewed literature. However, Hemlin and Olsson (2011) identified that leaders supporting developing researchers are additionally called on to provide emotional support. This is consistent with our team reflections and supports our recommendation that research teams including youth and developing researchers require additional timeresources. This is necessary to reduce the stress burden on these valuable team members and the academics who support them. In collaborative research teams, additional academic and administration support can be required right from the grant-writing stage to prevent community partners experiencing "research fatigue" (Liebenberg et al., 2017, p. 5).
While our study focused largely on within-team leadership dynamics, it is also vital to consider beyond-team contexts and issue, as they set the stage for internal dynamics. Bucic et al. (2010), for example, suggest that the adopted leadership style of a research team leader is often a reflection of the leadership style of their higher-education institute or wider organisation. As such, those suggesting academic institutions promote community engagement from the strategic level (Bolden et al., 2008;Hart & Church, 2011;Overton & Burkhardt, 1999) may see a cascading impact on leadership within teams that is more collectivist. To support this practice, West et al. (2017) recommends leaders build a "collaborative leadership culture" suitable for cross-boundary working, which always begins with the question "how can we help you?" (p. 15). Certainly, in our case, the research took place, albeit with varying degrees, within the context of wider institutional commitments to community-university collaboration.
The practical implications of our findings suggest that simply providing methodological guidance is insufficient in terms of supporting research team leaders working across cultures and disciplines. This is also evident in the literature. Reflecting on an international study of leadership within schools, Swaffield et al. (2005), for example, recommend cross-cultural research teams begin with a discussion on concepts and terminologies, suggesting that methodological positions may be understood differently within different cultural contexts (pp. 30-32). Finally, our reflections suggest that involving young people in research can require additional academic time-resources in terms of mentoring and "looking out for" those new to research. It also suggests that including youth in the research process may be experienced, at times and by all involved, as stressful. Efforts should be made to provide pretraining and inspire self-leadership in youth prior to engaging them.

Conclusion and Recommendations
This article used case study data as a point of departure to craft a framework with nine distinct leadership areas to develop collaborative work. The article has offered critical reflections on the negotiation of leadership in the context of complex interdisciplinary, co-productive, intergenerational, and international research projects led by academics. We have demonstrated, from both the literature and our own team reflections, that this is very complex territory. This type of research can be "expensive," in terms of finances, time, and emotions, and requires reflective and flexible leadership by all team members. For academics planning to act as principal investigators for such projects, the flowchart in Figure 1 may assist in determining at the design stage if/in what ways nonacademics might be involved.
Once it has been identified that community members should be included within such projects, we recommend that the principal investigator and wider team consider the areas we highlighted above from the literature as a starting point in relation to broader leadership. Important issues to consider include the following: 1. Promote shared leadership (including shared decision-making). 2. Aim to produce mutually beneficial knowledge. 3. Provide flexibility for team members to lead on different aspect of the research or at different time points. 4. Promote knowledge-sharing across cultures/disciplines and organisational boundaries. 5. Identify the principal investigator as the leader for the purposes of accountability (and identify others whom can be relied on when they are not available). 6. Encourage innovation and creativity. 7. Consider and support relational and emotional needs of team members. 8. Support the inclusion and leadership development of "newcomers" to the research arena by considering: (a) Which relevant community partners should be included to support youth inclusion? (b) What areas of the research process would benefit from youth inclusion and what methods and methodologies would support this? (c) Is it possible to design the study to allow youth flexibility in participation at various stages of the process, and how can continuous consent processes allow for this? (d) What additional emotional, cognitive, financial, and time resources will be required to support young people on this project? (e) Right from the start, how can youth be credited for their work so that their participation is awarded its rightful place, rather than academic researchers receiving all the credit?